Speeduino 2T EFI Project

[TZ350]

Ecotrons sells a two stroke EFI kit, and I have been using one, the problem with their kit is that it only allows for the Alpha-N tuning control protocol. Making power has been easy but drive-ability has been my stumbling block for a long time and I did not know what was making the tuning so difficult. Now I know a Alpha-N map on its own is not enough and that I also needed a VE map to get drive-ability out of the corners.

Moving on from Ecotrons, and Ecotrons is past history now I am moving forward with a Speeduino for my 2T two stroke EFI project.

https://speeduino.com/wiki/index.php/Speeduino

The mission is to find a way of taking a MAP value from the varying pressures in my two stroke crankcase so I can use a VE table for fueling control.

The Arduino based EFI development project, "Speeduino" sells add on boards to convert the Arduino Mega 2560 to a EFI CPU and they supply them with open source firmware that can be modified by the end user, using the Arduino or other C++ code editors.

338117 The Speeduino is an EFI shield for the Arduino Mega 2560 338118

As well as all the EFI and Ignition IO the Speeduino board also has a development area where you can add your own IO and change the Arduino software code to suit, so very flexible.

The Speeduino is aimed at cars but it also has a 2T option and I like it because I can change things and should be able to modify the open source software to get a timed crankcase pressure reading for a sample and hold MAP value. It is what I need for the next step in my 2T EFI project.

https://speeduino.com/wiki/index.php/Speeduino

.

[TZ350]

What I have learned about 2T EFI so far .... A

Use the smallest injector possible. This is important because it gives you the widest range of tune-ability possible.

A small injector delivers enough fuel but as RPM goes up a small injector runs out of time to deliver the fuel required and a bigger injector is needed to deliver the same amount of fuel in the shorter time available. For high revving two strokes staged injection is must.

338131 It is all about the available time to get stuff done.

When a EFI 2T engine gets the hurry up, it becomes all about the diminishing amount of time to get the good stuff in there.


338132

Blue - RPM, Yellow - maximum injection time, Green - small injector, Red - big injector

Starts with the small injector, swaps to the big injector then runs both at the same time.

A small injector for small TP throttle position and a big one for big RPM. Because there is a minimum on time for an injector a big injector can not be turned down enough to fuel properly at small throttle settings. An injector big enough to get the job done at 12,000 rpm floods the engine at lower rpm or throttle position.

The cross over for staged 2T injection is about 9 or 10K rpm. Over that and you need two different sized injectors.

There are physical injectors and logical injectors.

A physical injector is just a physical injector and a logical injector is two or more physical injectors treated as one by the EFI's CPU and fired together.

338057

My 13,000 rpm NSR110 cylinder needs staged injection and has three physical injectors but only two logical injectors as the B port injectors are fired together as one.

338056 B port injection. Flettner was injecting 2T's this way before KTM did.

The YZ250 cylinder does not need staged injection as its RPM ceiling is less than 9,000rpm and one size of injector can cover the tuning range required. But it has two physical injectors for symmetry and they are treated as single logical injector by the EFI's CPU.

...... the timing of the Injection end point is critical .....

For me it was BDC where the pipe is sucking its hardest, Flettner prefers just before transfer port closing, but whatever, your engine will have a sweet spot and it pays dividends when you find it.

[TZ350]

What I have learned about 2T EFI so far .... B

For simple EFI there are two basic tuning protocols, Alpha-N, and VE, volumetric efficiency.

Alpha-N is TPS vis RPM and VE is RPM vis MAP.

TPS indicates changes in the size of the hole through the throttle, changes in the size of the hole don't necessarily indicate actual changes in air flow through the hole but MAP, manifold absolute pressure indicates these changes.

Alpha-N by itself maybe works Ok on basic industrial and low powered 2T's, and anything that does not rely heavily on its pipe for power. And Alpha-N is great for power and all positions where the air flow is predictable but totally useless for drive-ability on a highly strung two stroke with fussy resonant tuning.

Alpha-N by itself will maybe work Ok up to about 9 bar BMEP after that something like VE is needed too for drivability. Highly strung 2T's are typically 11-14 bar BMEP. The problem with Alpha-N is that it does not allow for possible changes in air flow at the same TPS vis RPM setting.

On a high performance two stroke the wave action in the pipe collapses when the throttle is closed. When the throttle is opened again the noise comes back immediately but the strength of the wave action in the pipe and therefore air flow through the motor takes some moments to build up again.

338068

Same Alpha-N TPS vis RPM position but different airflow's through the motor due to the different strengths of the wave action in the pipe. Because it can't adjust itself for the collapsed air flow, Alpha-N will over richen the mixture or even flood the motor when you are trying to exit the corner.

Alpha-N is only really good for near wide open throttle power when the motor is singing and the airflow is consistent. For drive-ability out of the corners a VE, RPM vis MAP, manifold absolute pressure or similar protocol is required to account for changes in air flow induced by the wave action collapsing in the pipe.

Making power has been easy but drive-ability has been my stumbling block for a long time and I did not know what was making the tuning so difficult. Now I know a Alpha-N map on its own is not enough and that I also needed a VE map to get drive-ability out of the corners.

Traditional four stroke methods of measuring air flow through the motor or manifold absolute pressure don't work in a two stroke so some other procedure is required. Not much that is four stroke translates over to two strokes.

What we need is a two stroke method of seeing changes in air flow through the motor and I think this can be done by looking the changes in the difference between the peak vis low pressure cycles in the crankcase.

I have already worked out a method and an algorithm for doing this.


338126

I have been making progress with the simulated MAP idea.

I have abandoned my original idea of using a MAP sensor to see the high and low pressures because MAP sensors are too slow and instead will use a piezo element.

338127 EngMod2T simulated crankcase pressure traces.

Because all the peak pressures all line up at the same crank angle I have embraced Flettner's idea of "sample and hold" by making only one timed measurement at peak pressure and using that for the complete cycle.

338129

I have been able to use some PC based software to generate a simulated ignition trigger pulse and crankcase pressure trace for my software's test and development purposes.

338128

The code is pretty simple, an interrupt triggered by the ignition trigger pulse at which time it waits about a third of the duration between pulses and then samples the crank case pressure. The loop writes the simulated MAP output signal that then goes to the EFIs CPU.

Simple code but it took a while to get it and my test setup to work reliably over the full RPM range.

Ecotrons .... Alpha-N, by itself is not a good enough tuning methodology for a high performance 2T

[TZ350]

What I have learned about 2T EFI so far is to go here .... C

...... the timing of the Injection end point is critical .....

Follow the posts link back because there are lots more as well as video's of EFI 2T's being ridden in the field.


I have learned from Flettner, that a competition two stroke can be successfully fuel injected and that EFI can work very well on 2T's.

338125 338124

With Flettners guidance and his example of two very successful working bikes using the Link EFI unit, one on E90 and the other on pump gas I have been encouraged to have a go myself.

338123

Blue line is the Fuel Injected YZ250 (pump gas) and Red line is the Kawasaki Big Horn (E90).

The first lessons I had to absorb were that the timing of the Injection end point is critical on a two stroke and should be timed to end at transfer port closing and the fuel should be injected into the transfer ports, either counter stream like the YZ250 or across the port like the Kawasaki but definitely not straight into the inlet or the crankcase itself. To be successful the full fuel charge has to be inducted into the cylinder on each cycle otherwise you get rich/lean cycles and poor running.

As I understand it: with a carburetor, any time air moves fuel gets mixed with it, so all the inducted air in the crank case has air with fuel more or less evenly mixed throughout it.

With a fuel injection system that has discreet injection pulses and with the injector at the inlet only some of the inducted air will have fuel mixed in with it. Now inside the crank case there are patches of air with no fuel, and some over rich areas.

With inlet injection its random what air/fuel mixture strength gets sucked into the cylinder, you might get a few lean cycles then a heavily over rich one and this is what is happening when the motor is hunting. And if you richen it up so there is no lean cycles then the whole thing goes so rich that the motor will barely run at all.

As we all know, a certain amount of air requires a certain amount of fuel. But with inlet injection the fuel does not get mixed evenly into the air as it passes through the crank case so you have no chance of controlling the exact amount of fuel being inducted into the cylinder with inlet port injection.

But with transfer port injection you have a much better chance of controlling the exact amount of fuel that gets inducted into the cylinder on each cycle.

(direct injection into the cylinder is another very good approach to 2T EFI, but I cant easily get suitable fuel injection parts to try that).

YZ250 counter flow injection makes for good air/fuel mixing.

EFI Kawasaki BigHorn running transfer port injection and E85. Runs to 8,000 rpm

Some more Youtube clips of transfer port injected 2T's
http://youtu.be/eleqBGvOM4M
http://youtu.be/hOGZ5llowoU
http://youtu.be/1YG9ko8-Nwk
http://youtu.be/UEQli7nuak4
GerbilGronk is worth a Youtube search.

I have been working on a EFI transfer port injection system of my own but trying to take the concept to 12,500rpm and that means staged injectors.

https://www.kiwibiker.co.nz/forums/showthread.php/159112-Ecotrons-Engine-Management/page11


Posted this because someone asked me about the jerky response they are getting now that they are running split injection. Ie they are using a small injector for slow running and a large injector for power. I think the issue is in the size of the step on the map at the point where the ECU changes from the small injector to the larger one.

The Ecotrons Alpha-N map is not a fuel map but a map of predicted changes in engine load (torque) with respect to TPS and RPM.

338120 338121 338122

It is like a stair case and to keep the Load steps even, it is possible to cluster the rpm steps around the area of rapid changes in Load (torque). And on my dyno graph that is 7,500 to 9,500 rpm (yellow line). Above that the power goes up with rpm but the change in Load is a lot flatter. So you can have quite wide rpm steps below and above where the engine comes onto the pipe and cluster the rpm steps where the Load (torque (yellow line)) curve climbs rapidly so as to have even predicted Load steps across the complete rpm range from 3,000 to 13,000.

Ie on my map the first rpm point is at 3K rpm, the second is 6k the third is 7k then 7.5k, 7.75k, 8k, 8.5k, 9k and rpm step 9 is 9.5k step 10 is 11k then 12.250 and finally 13,250 rpm. So the spread of rpm points is chosen to get even changes in the Load steps.

My TPS steps are 0, 10, 12, 15, 18, 20, 26, 31, 36, 41, 46, 51, 61, 71, 85, 98. there is much greater changes in air flow at low TPS than there is above 70%. If you look at the dyno graph everything above 70% has pretty much the same power output.

What I think happens when split injection does not work properly and jerks all around the place is that when the small injector gets to a load cell step that is to big and would max it out the ECU swaps to the big injector which squirts to much fuel for the rpm (because the min on time for the big injector flows to much fuel) and the rpm drops back below the small injectors last effective load cell and the air/fuel clears and the rpm slams backup and the cycle violently repeats itself. Now a smaller big injector might help but a disparate size in the load steps can be the real problem.

This will not be confined to a set rpm position on the map but a patch of Load cells where the effective differences in predicted Load is to great. And I expect the ECU's hysteresis between injectors will play a part here too.

The trick is to group the RPM columns and TPS rows in such a way that the changes between adjacent Load cells in any direction are much the same.

About choosing the RPM steps for a Alpha-N map. Its worth following the post links back as there is lots more good background info there.

https://www.kiwibiker.co.nz/forums/showthread.php/159112-Ecotrons-Engine-Management/page11


The EFI lesson I learnt last month was, that to get the greatest range of tune ability I needed to use the smallest injectors possible that will still do the job. I know, pretty obvious, but there you go.

I read the Ecotrons EFI manual and saw that the map must have as many even steps as possible and I thought that meant dividing the rpm range up as evenly as possible, seemed logical, but wrong again.

The bike bucked around swapping between the staged injectors as the EFI struggled to jump from one cell to another on the steeper parts of the fueling curve.

I guess what they really meant was that the changes in load between cells needs to be as even as possible.

So I don't need even spacing between the rpm break points and can have big or small gaps between them so long as there are no big steps between the load values in adjacent cells.

308859

I can now throttle on/off pretty well and do a series of part throttle runs, the last one is at 30% TPS, it looks a bit lean.

Its not perfect but it is running a lot better.

Ok, so I can make good power with Alpha-N but in the end for me, the Alpha-N method by itself has became a dead end because it does not adjust for changes in the mass air flow induced by changes in the strength of the wave action in the pipe. This is something that becomes ever more of a problem as the BMEP increases and the engine becomes ever more reliant on fussy resonant action in the pipe for its power.

[TZ350]

What I have learned about 2T EFI so far is to go and here .... D

A lot of back story here:- https://www.kiwibiker.co.nz/forums/showthread.php/159112-Ecotrons-Engine-Management/page11


Some helpful input from Dmcca and Wob on how to use a wideband O2 Lambda sensor to tune a 2T.

Putting the Lambda in the stinger, seems to be way more stable and repeatable when logging the A/F data.

https://www.kiwibiker.co.nz/forums/images/misc/quote_icon.png Originally Posted by TZ350 https://www.kiwibiker.co.nz/forums/images/buttons/viewpost-right.png (https://www.kiwibiker.co.nz/forums/showthread.php?p=1130961626#post1130961626)
The Ecotrons EFI has barometric and air temperature sensors for automatic fueling compensation so hopefully once I am on the money with the dyno the ECU's auto compensation will give pretty much the same results as changing the jets 3 times a day.

I feel very comfortable with tuning the EFI system for maximum power, above 50% throttle position that is relatively easy.

The problem I am struggling with is at the low end, if this was a carb then the EFI bit that is challenging me now is the area covered by a carb's pilot jet and transition to the slide cutaway.

https://www.kiwibiker.co.nz/forums/images/misc/quote_icon.png Originally Posted by dmcca https://www.kiwibiker.co.nz/forums/images/buttons/viewpost-right.png (https://www.kiwibiker.co.nz/forums/showthread.php?p=1130961832#post1130961832)
May I ask why you're not using a wideband O2 to tune this area? It would be relatively straightforward with a wideband and a brake dyno.

I started out trying to get a 12.5:1 A/F everywhere on the map, as you do if you don't know any better. Things weren't going well and I gave it away when it was pointed out to me that it was the short circuiting fooling the O2 sensor into thinking the system was running lean when it was not.

It really confused things and it looked like the O2 sensor was not going to be very useful so I discarded it. But your experience encourages me and it looks like that with a bit more experience myself I might be able to get somewhere with it now I have a better idea of what is going on.

https://www.kiwibiker.co.nz/forums/images/misc/quote_icon.png Originally Posted by Frits Overmars https://www.kiwibiker.co.nz/forums/images/buttons/viewpost-right.png (https://www.kiwibiker.co.nz/forums/showthread.php?p=1130961833#post1130961833)
Why a brake dyno? Then you would be developing an engine for a set of circumstances that it will never see on a race track. Going through the revs on an inertia dyno and registering revs, torque and lambda simultaneously is an excellent way of establishing an injection map. Besides, running constant revs on a brake dyno with too little fuel or too much ignition advance may kill your engine, whereas it might survive on an inertia dyno.

https://www.kiwibiker.co.nz/forums/images/misc/quote_icon.png Originally Posted by dmcca https://www.kiwibiker.co.nz/forums/images/buttons/viewpost-right.png (https://www.kiwibiker.co.nz/forums/showthread.php?p=1130961900#post1130961900)
When i say brake dyno i mean an eddy current or similar that allows both constant rpm tests and normal 'runs'... Im sure that doing only constant rpm tests is not ideal.

Personally I do a combination of both to tune an engine and so far it has worked well for me, especially when tuning the low throttle area that TZ was talking about... in saying that ive only ever had an eddy current dyno, never an inertia-only dyno. And have developed my own methods of tuning without ever seeing first-hand anyone else's process for tuning a two stroke... maybe its a case of ignorance is bliss, however I find my process to be very quick and very reliable for a wide range of engines.

I find that the lambda sensor is a little slow to react at low revs and part throttle. Its fine at higher revs and higher throttle. By holding the engine at a certain rpm point for 3-5sec at each throttle position it gives the readings time to stabilise and gives much better, more repeatable data.

I start by doing a range of fixed rpm tests at different throttle openings, with the revs building higher each test... eg 4000, 6000, 8000, etc at 1/8, 1/4, 1/2, 3/4, full throttle. By doing this you can easily see the areas that need changing (pilot, slide, needle diameter, clip position, needle taper, main jet, etc).

If there is a serious problem with the tune you will always catch it before doing damage as you’ll see AFR's begin to lean out or detonation begin to show before you get too high in rpm. Once I’ve done basic tuning like this then i do full runs and constant load testing and fine tune if necessary, but these are normally only very minor changes.

Whether the AFR reading is showing true AFR or not doesn’t really matter, the numbers can still be used to tune an engine... around full throttle and peak torque it is fairly reliable to use ~12.5:1 and leaner readings below that, down to around 13.5-14:1 at part throttle/low rpm.

Of course you still need to use experience and 'feel' to do the fine tuning but the AFR numbers will get you close. Even if the engine is not actually seeing 14:1 at part throttle/low revs, its still a repeatable data point that can be used to tune an engine. Its certainly a lot better than guessing, especially when tuning the part throttle areas.

https://www.kiwibiker.co.nz/forums/images/misc/quote_icon.png Originally Posted by wobbly https://www.kiwibiker.co.nz/forums/images/buttons/viewpost-right.png (https://www.kiwibiker.co.nz/forums/showthread.php?p=1130961921#post1130961921)

+1 the above.Adding that putting the Lambda in the stinger, seems to be way more stable and repeatable when logging the A/F data.

Using data this way is something that you really don’t appreciate how good it is until actually doing it.
Hence TeeZees surprise when I related that every time you change a small item in the engines tune - the dyno result for sure is not telling you really what you want to know, unless you reject every time to hit a predetermined baseline.

As Jan Thiel said, most all of the pipe testing at Aprilia was a complete waste of time, when he finally realized allot of the power differences were down to how the egt reacted to that pipe. Not that the pipe was "better " for some aspect of scavenging, or trapping efficiency in making more or less power.

[TZ350]

What I have learned about 2T EFI so far is to go here .... E

338130 position of my WBO2 sensor near the stinger.

I started out trying to get a 12.5:1 A/F everywhere on the map, as you do if you don't know any better. Things weren't going well and I nearly gave it away before it was pointed out to me that air was the short circuiting fooling the O2 sensor into thinking the system was running lean when it was not.

"Transfer Efficiency" is the amount of air inducted and passed up to the cylinder.
"Trapping Efficiency" is the amount of that air that is retained in the cylinder.

So if the transfer efficiency is greater than the trapping efficiency then the spilled air will make the O2 sensor show lean even though the trapped air is at the correct air/fuel ratio. The indicated Lambda always improves as the motor gets onto the pipe and the trapping efficiency improves.

I found things worked well when the motor was off the pipe and Lambda indicated 1.2 and on the pipe it was 0.85.

An indicated Lambda of 0.85 or A/F ratio of 12:1 in a two stroke is great for power.


Some links to YouTube clips of a couple of two stroke bikes that have been successfully modified for EFI.

Yamaha YZ250.
https://youtu.be/hOGZ5llowoU
https://youtu.be/UEQli7nuak4
https://youtu.be/1YG9ko8-Nwk
Kawasaki BigHorn.
https://youtu.be/eleqBGvOM4M
https://youtu.be/ifSEql1X4R0
https://youtu.be/CnIemdISKrM




I have tried a few different places for injectors, some places seemed to be better than others.

Follow the post link back to see the variations injector position I have tried.

And this is why direct in cylinder fuel injection wont work on a high performance 2T.


Intuition is great for giving you a direction to look in and being able to put (rough) numbers to something like TeeZee does can tell you if you are on the right track or not. Numbers can save you from a lot of wasted time.

Time at 13k rpm = 4.6ms for a complete 360 deg crank revolution.

Transfer opening to closing 130 deg. 1.7ms.

From exhaust port opening to closing 200 deg. 2.6ms

From exhaust closing to ignition. (with ignition timing 18 deg BTDC). 72 deg. 0.8ms.

Pressure in the cylinder ATDC greater than the 3 bar fuel injection pressure 90 deg, 1.2ms.

So somewhere between 1.7 and 2.6ms for injection, 0.8ms for fuel homogenization.

2-3ms required to open and close the injector 2-3ms required to deliver the fuel another 2-3ms or so to homogenize it.

Total time required:- 6-9ms

Is there enough time to get everything done? I am picking not, and I am picking TeeZee doesn't think so either.


With EFI, "Time" is King. :niceone:


Injector Position

338134 338133

Central injector fires through a slot into the underside of the piston.

My favourite low speed injector position has been to fire under the piston. This seems to help enormously with fuel homogenization at lower and off pipe engine speeds.

338140 338142 338141

I have tried the injectors in lots of different places. direct through the port, across the crankcase, inlet, throttle body, bell-mouth but across the transfer ports is my favorite and when that is not physically easy to arrange then straight down the B ports and under the piston is always great for the small injector.



TZ350, we rev 250 4 strokes to 15,000rpm all day with a single injector in the throttle body. I'm comparing apples to oranges I know, but still... was shocked to think you need 3 injectors for 13k.

15,000 4T rpm is only 7,500 2T rpm. Anything 2T over 9,000 rpm is looking at staged injection

There are actually only two "logical" injectors, one big one small. The two "physical" injectors in the B ports are there for symmetry of fuel distribution and are fired together as one Logical injector. I use three physical 124g/min injectors but the EFI CPU sees only two logical injectors, a 124g/min slow speed injector and one 248g/min high speed injector.

But it is not about injector size but time available to get the injecting done. If you only have half the time, then you need an injector twice the size to get the job done. And you need a small one for tune-ability when there is plenty of time like <9k 2T rpm (equivalent to 18k 4T rpm) and a much bigger one to deliver much the same amount of fuel when time is getting short like above 10k.

The reason you can't use the big injector for slow running is because an injector must be turned on for a minimum amount of time to get it completely open and controllable. And a big injector just spills to much uncontrolled fuel while it is opening for idling speeds and small throttle opening.

15,000 4T maxed out rpm "Time" for an injection cycle is just about where my 2T is starting to come on the pipe and get going. A 4T has twice as much time to get things done so only needs an injector half the size of a 2T, ie something about the size of my slow running injector.

It is "Time" not injector size that is the real issue with EFI.


a single injector in the throttle body.

Another issue with 2T EFI and throttle body injection is that unlike a carburetor which adds fuel to the air stream all the time, regardless of whether the air is being sucked in or blown back out of the crankcase ie fuel standoff.

A fuel injector squirts fuel in discreet blobs. And is just as likely to add all the fuel to the air stream as its blown back out as it is to the air stream being sucked in. So there could easily be rpm patches where you could wind up with little fuel at all in the crankcase.

This is less of a problem with transfer port injection. Where the bulk of the fuel is concentrated in the transfer duct and not the inlet tract where it can be more easily ejected. And will be one of the reasons why 2T EFI can show improved fuel efficiency over a carburetor.

And another reason for injecting into the B transfers is that the bulk of charge short circuiting happens from the A ports and if there is little fuel in the short circuiting air then the motor is cleaner and more fuel efficient. Another win over the carburetor which would have loaded all the crankcase air with fuel. With a carburetor all blow back and short circuiting air loses fuel and pollutes the environment, with 2T EFI not so much.


If a clear suction signal to the carburetter is no longer needed, you could start opening the RV at about 20° after BDC at max.torque rpm, assuming sufficient transfer angle.area, and even earlier at lower rpm's.
Opening the RV as soon as the crankcase pressure has dropped to the same value as the pressure upstream of the RV, will prevent the case pressure from dropping much further, which would slow down and eventually reverse the transfer flow.

[ceci]

Very interesting, so you can not see or I do not see parts D and E.

[TZ350]

Very interesting, so you can not see or I do not see parts D and E.

I havent finished, maybe later today.


Links to a whole bunch of EFI projects. Mostly amature home tuners, and all great work.

EFI Kawasaki Bighorn on the dyno https://youtu.be/ifSEql1X4R0
EFI Kawasaki Bighorn Trail ride 1 https://youtu.be/eleqBGvOM4M
EFI Kawasaki Bighorn Trail Ride 2 https://youtu.be/CnIemdISKrM
EFI Kawasaki Bighorn VMX https://youtu.be/v8i6ps0PdVI

Maramarua classic forest trail ride https://youtu.be/ZqzOgOJuJ1E

EFI YZ dyno run https://youtu.be/UEQli7nuak4
EFI YZ250 look around and first start up https://youtu.be/hOGZ5llowoU
EFI YZ250 trail ride 1 https://youtu.be/1YG9ko8-Nwk

Nath88?
EFI first start up https://youtu.be/wIRF3VJUqBk
EFI ride around the block https://youtu.be/4Zsz26QmfAU

CVT EFI NOS and everything else https://youtu.be/A9d-atuQO7Y
EFI and Turbo Snow Cat https://youtu.be/U1uz7gdS-cA
SeeDoo Etec https://youtu.be/LhNHcy9si5o

1975 Suzuki GT250 road bike fuel injected start up https://youtu.be/A7fONUsJQD4
1975 Suzuki GT250 road bike fuel injected road test https://youtu.be/BePHcg5dNjw

Fuel injected 1971 Yamaha R5 https://youtu.be/qzVHgHJ9PWI

Athena Direct Injection https://youtu.be/h-F_IndOm5I

BRC 250 EFI kart engine https://youtu.be/dNat9ZwWbq8
BRC 250 EFI kart engine https://youtu.be/Kfx61bUOznc

Racing EFI Vespa https://youtu.be/_sjx2i4xFlY

EFI Banshee first start https://youtu.be/e1WHVVT6d9c
EFI Banshee run https://youtu.be/eznxX_ndpII

Direct injection https://youtu.be/IpjWi-0qTJg

Will direct injection DI save two strokes https://youtu.be/ql2-wW0LuWY

Orbital direct injection https://youtu.be/fIoB0zTeUAY

EFI Dragbike Thailand https://youtu.be/VB6Q_fQQ_PY

Mega Squirt YZ125 https://youtu.be/Di8FuHApLKA

Aprilia RSi 125 EFI Part 1 https://youtu.be/S1xvr_VXzx4
Aprilia RSi 125 EFI Part 2 https://youtu.be/j5FOUn7V3nc
Aprilia RSi 125 EFI Part 3 https://youtu.be/WBaiMJb6ntY
Aprilia RSi 125 EFI Part 4 https://youtu.be/pZl5KPwvm0s

EFI 125 kart engine https://youtu.be/4D8WRjDxan0
EFI 125 kart engine https://youtu.be/b1YbPdw00Ww
EFI 125 kart engine https://youtu.be/r5ljaDdaH_o
EFI 125 kart engine https://youtu.be/uexMJbtyZdU
EFI 125 kart engine https://youtu.be/uzbSsYh8Z9w
EFI 125 kart engine https://youtu.be/XIcX9-GsMGQ
Fuel injected Kart https://youtu.be/R5_KlcefQ8k

2017 KTM dyno https://youtu.be/8g0uSIisMtI
2017 KTM demo ride https://youtu.be/nf1tfdUtjJI

EFI CR250 https://youtu.be/VlHyRcmlLgU http://www.supermototecnica.<wbr>com/2015/09/10/sviluppi-del-<wbr>sistema-iniezione-diretta-<wbr>indiretta-due-tempi/ (http://www.supermototecnica.com/2015/09/10/sviluppi-del-sistema-iniezione-diretta-indiretta-due-tempi/)

EFI YZR500 https://youtu.be/tNAtmKzqo9A

EFI CR500 https://youtu.be/BskSxMoQrwA

2T Nitro Engine visible combustion https://youtu.be/aO5Qkzlo2Kw
2T visible combustion https://youtu.be/G7PSec7RjRw
More 2T Visible combustion https://youtu.be/BrWpF6YiSDs

Fuel Burn in a side valve 4T https://youtu.be/jdW1t8r8qYc
4T Visible fuel burn https://youtu.be/31mtOdHGbB4

[pete376403]

Did you look further into the Aviamechanica controller (or was that someone else?)

[TZ350]

Did you look further into the Aviamechanica controller (or was that someone else?)

Yes I took a look but they don't use a MAP sensor, basically it is a modified Alpha-N setup. Very good system for what they are doing.

Their system can be found here:-


We make our EFI + DC spark ignition compatible with http://tunerstudio.com/index.php/tuner-studio So it completely open for experiments.

https://www.kiwibiker.co.nz/forums/showthread.php/159112-Ecotrons-Engine-Management/page18



They use the same tuning application that Speeduino do too.

[ceci]

I haven finished, maybe later today.

Sorry for my impatience.

[TZ350]

338160 338162 338161

My Speeduino ECU kit turned up today. The Speeduino board with all the heavy current handling output, MAP and other IO is made as an Arduino shield and plugs straight onto the Arduino Mega. There is a box too with a handy row of screw terminals that makes experimentation easy.

The beginners guide to getting going with the Speeduino:- https://speeduino.com/forum/viewtopic.php?f=21&t=221

These are the Injectors I have been using, good for 13,000rpm

https://www.aliexpress.com/store/product/BRAND-new-fuel-injector-for-motorcycle-QUAD-RHINO-Dune-Buggy-ATV-Snowmobile-with-Connector/2189051_32292413156.html?spm=2114.12010612.8148356 .4.27f53d56DUcmks

Other motorcycle sized EFI parts.

https://superturbo.aliexpress.com/store/group/Parts-for-EFI/2189051_513253989.html?spm=2114.10010108.100008.2. 7d5c6ccdMKczmJ

[ceci]

This year 2018 Ktm uses the EFI and in the TPI version, but what kind of TPI they use, the one designed by Flettner, and why they use this and not other types of TPI http://www.supermototecnica.com/2015/09/10/sviluppi-del-sistema-iniezione-diretta-indiretta-due-tempi/.

The advantage of this type of injection: injected against the current and that makes the size of the drop of the spray is large quickly evaporate gasoline.

The disadvantage of the others is that the air stream transports the spray droplets instead of evaporating them

[TZ350]

http://www.supermototecnica.com/2015/09/10/sviluppi-del-sistema-iniezione-diretta-indiretta-due-tempi/

Interesting link, I notice the guy says he has good drivability. Drivability is my problem.

[ceci]

TPI there are several types, this is also TPI:
https://patents.google.com/patent/US6691649?oq=two+stroke+engine+injection+rotax

The placement of the injectors is different in the three cases mentioned, but common in the tranfer

[TZ350]

http://www.supermototecnica.com/2015/09/10/sviluppi-del-sistema-iniezione-diretta-indiretta-due-tempi/

338371 I like the way he squirts above and below the piston. Not so sure about squirting down the side of the piston. But it is 100% time available.


https://patents.google.com/patent/US6691649?oq=two+stroke+engine+injection+rotax

338368338369338370

This arrangement looks very time limited.


The disadvantage is that the air stream transports the spray droplets instead of evaporating them

Yes, I very much agree with you, Flettners idea of straight down the B ports is the best for evaporating the fuel droplets and getting a homogeneous fuel mixture.

338372

[TZ350]

338373

Got my Speeduino fuel injection CPU kit. Put it together and started to find my way around it and understand its methodology.

338374 Squeeze bottle is for the MAP sensor. Leds are Injectors and the pot is for TPS.

Interesting system. It has staged injection but instead of it being the small or large injector. Just one or the other as in the Ecotrons system. The Speeduino fires both together but varies the on time of them individually to suit.

338375

I have run into a bit of a problem. I have the Speeduino system in two stroke mode and have been sending a 100Hz ignition signal to it. This should equate to 6,000 rpm but I am getting a healthy 12,000 four stroke reading.

Not sure what is happening. No doubt I am doing something wrong, but what? ..... :scratch:

[TZ350]

http://www.supermototecnica.com/2015/09/10/sviluppi-del-sistema-iniezione-diretta-indiretta-due-tempi/.
Interesting link, I notice the guy says he has good drivability. Drivability is my problem.

People have had a lot of success with the Alpha-N methodology on bikes like the CR250 and have got good drivability.

But the 250 and bigger singles have a RPM ceiling less than 10,000 rpm and a lower than 10 bar BMEP compared to tuned 125's and 50's that typically rev to 13-15,000 rpm and develop more than 10 bar BMEP.

Over 10 bar BMEP a two stroke motor becomes very reliant on the wave action in the pipe for its power and this wave action is very finicky at smaller throttle settings and prone to collapsing with a subsequent loss of air flow through the motor. Alpha-N (TPS vis RPM) cant cope with this effectively because it cant see changes in airflow, it can only guess at it.

It seems that less than 10,000 RPM and 10 bar BMEP you only need one logical injector and an Alpha-N map will do. Over the magic 10 threshold you need staged injection and a way of seeing changes in air flow for the same RPM and Throttle position.

To get good drivability with a tuned 125 or 50 something like a VE volumetric map and MAP or MAF sensor value is required. Something that can adjust the fueling to match the changes in the finicky airflow at small throttle settings. VE is RPM vis MAP or RPM vis MAF so it can see changes in air flow and adjust the fueling accordingly.

Four stroke methods for measuring MAP or MAF do not work very well in two strokes. Another, more two stroke method has to be used. I am not sure what that is yet.

[Niels Abildgaard]

Four stroke methods for measuring MAP or MAF do not work very well in two strokes. Another, more two stroke method has to be used. I am not sure what that is yet.

Many years ago I made hot wire anemometry for some district heating caloric meters.
The signal was from .5 to 250 hz and had to pass through a layer of epoxy and stainless steel.
Without that I think a sensor in one of the transfer channels could detect amplitude using the first half of the open period .
Would that leave enough time for injecting?

[TZ350]

Many years ago I made hot wire anemometry for some district heating caloric meters.
The signal was from .5 to 250 hz and had to pass through a layer of epoxy and stainless steel.

Without that I think a sensor in one of the transfer channels could detect amplitude using the first half of the open period .
Would that leave enough time for injecting?

It is an interesting idea. I would like to know more about how they are made. The frequency range is good.

Injection is timed to finish at BDC. But the hot wire reading could be used on the next cycle.

The bulk of transfer happens well after TPO transfer port opening and just before BDC where the pipe should be sucking its hardest.

A sensor in the transfer channel would have to cope with regular doses of fire, you can see the combustion staining all the way down the transfer ducts.

338376

[ceci]

338368338369338370

This arrangement looks very time limited.




I do not understand why you say that the injection time is less.
It is injected into tranfers B in all cases

https://www.cycleworld.com/sites/cycleworld.com/files/styles/655_1x_/public/images/2017/08/ktm-two-stroke-illustration-bottom.jpg?itok=HnmNgOLy&fc=50,50http://www.arcticinsider.com/Images/Article/201701240725125.jpg

[TZ350]

338381338382

I am not sure that washing the oil off the piston is the same as the direct down the transfer duct like Flettner and KTM's arrangement.

338384

That is a great picture you posted of the injectors firing straight down the B transfer tracts.

Looks like Flettners and KTMs injector arrangement and is what I have tried to copy with my own engine. Having experimented with lots of different injector placements it looks like we are all starting to following Flettners lead. Straight down the B transfer duct is proving to be the most sucessful.

338383

[TZ350]

338393

Mass flow in the transfer duct, its easy to see why finishing the injection squirt around BDC is a good idea.

[TZ350]

.
Really Happy. I am starting to understand how it all works and have got all the gauges working. Everything is dummied up and my calibrations are crap but for the moment it is just about understanding how to wire things and set up Tuning Studio.

With my previous EFI experimentation the system completely switched between the small injector to the big one so it was all small or all big and that seemed to work Ok. I will be interested to see how the Speeduino approach of running the small injector all the time and adding the big injector as required works out in a two stroke.

I have been able to simulate a multi tooth wheel and get sensible RPM readings. And a lot of things look familiar, like the pulse width and duty cycle is the same at various RPM that I recognize from my previous work with that other EFI system. So things look very promising.

The only thing I have not simulated yet is the fuel pump. With that other EFI system at "Key On" the fuel pump would run for a short moment to prime the system then turn off and wait until the motor is cranking before turning on again. I will wire in another LED to check to see how Speeduino handles it.

Very little wiring required out to the injectors Etc so putting it all on the bike should be easy.

The engine needs a re build, may take a couple of weeks to get sorted and back into the bike, then I am looking forward to having a go with the Speeduino.

I am not sure how I am going to handle the VE mapping but I am ... https://speeduino.com/forum/images/smilies/icon_e_biggrin.gif ... Very excited to be making progress.
.
338513 338514

[TZ350]

.
I needed to figure out how to get the fuel pump switching on and was told that there is not a dedicated output for a fuel pump but I could re assign a spare MOSFET to do the job or I could use the development area of the board to add some components for switching a pump relay.

Assigning a spare MOSFET sounded easier so on the Speeduino user projects thread https://speeduino.com/forum/viewforum.php?f=19 I asked what I needed to do to re assign a MOSFET and whether I needed to go into the Arduino code to do it or should I be looking somewhere else?

This is the reply.


Just assign in TunerStudio to the desired output pin in the Accessories > Fuel Pump section. The post here (https://speeduino.com/forum/viewtopic.php?f=19&t=1569#p23096) is one that explains, also with two following posts for two methods (schematics and utils.ino) for determining the current pin assignments of specific PCB versions.

On the TS pin reassignment for Fuel Pump, note there is a slider on the side of the selection list, giving you many more options than 3 through 9 if you scroll the list down. https://speeduino.com/forum/images/smilies/icon_e_wink.gifv0.4 FP output is default on Mega pin 45 (goes to the proto area). Boost output is default on Mega pin 7.

Set FP to pin 7.
Burn.
Set Boost to pin 45.
Burn and Save tune.
New FP output is now through Q8 to J5 pin 1 and IDC pin 35 (your choice for wiring).

The output will ground when the FP should be active. It should be active when Speeduino is powered-up with 12V, for as long as the Fuel Pump Prime Duration is set (default 6 seconds), or when Speeduino detects RPM.

David

[TZ350]

.
Well that was easy, I followed the instructions and it all came together real easy. The third LED represents the fuel pump. The fuel pump itself draws to much current for the Speeduino board so a relay would be needed and we would wire the relay coil where the LED is.

338609 1-Primary Injector, 3-Secondary Injector, 35-Fuel Pump.

I have labeled everything important, wiring is real simple.

[Number-One]

Hi, do you know RusEfi?

https://www.youtube.com/watch?v=X7IXrSoFngc

[pete376403]

TZ, what platform are you running TunerStudio on? I have been trying to get it running on Lubuntu, (i've downloaded the correct versions for Linux) about to give up and go back to Windows.

Edit - never mind, I can see from screenshots that you're on W10

[TZ350]

.
I have been having a bit more success. I was using a signal generator to simulate the cranks toothed wheel but without a missing tooth Speeduino seemed to be having a bit of trouble interpreting what was going on.

I followed some good advice and got Ardu-Stim, compiled it and installed it on a Arduino UNO. Using Device Druid to configure it I was able to simulate a 24-1 tooth wheel and Speeduino seemed much happier and Tuning Studio displayed realistic results so it looks like I am getting close to the pot of gold at the end of the rainbow.

338763 It is not that obvious what to click to down load the Ardu-Stim zip file.

Arduino toothed wheel simulation code.... https://gitlab.com/libreems-suite/ardu-stim

Serial communication Library required by Device Druid to talk with the Arduino DUO.... https://www.arduinolibraries.info/libraries/serial-ui

Device Druid, used for configuring the Arduino simulation application.... https://devicedruid.com/

YouTube video clip of the TuningStudeo gauges moving about.... https://youtu.be/YeBDTIAjPys

338761 338762

The UNO is used to generate a simulated toothed wheel signal for the Speeduino EFI board. The Tuning Studio gauges all seem to be doing sensible things.

[TZ350]

338764 338765 338766 338767 338768

Now as soon as I can get this lot back together I will be able to try my Speeduino EFI CPU out in the real world.

[TZ350]

338769

My wide band O2 sensor is in the end of the reverse cone near the stinger. A position at this end of the pipe seems to give much better results and sensor life.

With two strokes, air fuel ratios and O2 sensors there are two important concepts to keep in mind.

"Transfer Efficiency" how much air/fuel mixture that has been transferred to the cylinder and at certain rpm it can easily be greater than 100%, ie VE > 100.

"Trapping Efficiency" is how much of the transferred air/fuel is trapped and retained in the cylinder to be burnt. Often much less than 100% of the transferred mixture. So a fair amount of air escapes along with wet fuel.

The escaping air reads as lean mixture where as in fact the mixture may be correct or even rich. So when using a O2 sensor one has to always bear in mind what the RPM is. Off the pipe the mixture may be correct but the motor looks real lean, Lam 1.2 or so and at peak torque the same correct mixture would look like Lam 0.85.

With Av Gas on peak torque and WOT Lam 0.85 or air/fuel ratio 12.5:1 is the go.
.

.

[TZ350]

.
Staged Injection. Currently the Injectors are simulated by 12Volt LED's.
On my single cylinder two stroke I need to use staged injection and thought I had cracked it by connecting the primary injector to pin 1 (pin 1/2) and the secondary injector to pin 40 (pin 2/2).
Ticked staged injection and both injectors would light up and then I set the cells in the staged injection map to 100% to check the pulse width on the secondary injector running on its own at maximum load but it would not light up at all, neither of them would.
At anything less than 100% both would light but at 100 when only the secondary should be alight, it was not, neither were.
If anyone has any suggestions about how I should setup staged injection properly it would be welcome.
.


...On my single cylinder two stroke I need to use staged injection and thought I had cracked it by connecting the primary injector to pin 1 (pin 1/2) and the secondary injector to pin 40 (pin 2/2).No problem. It sounds like you have both primary and secondary injectors connected to INJ1 channel. Each injector channel has two pins for output, in order to carry more current through the little pins. So looking at the pin-out here (https://speeduino.com/wiki/index.php/V0.4#Physical_Layout), pin IDC1 and pin IDC40 are both connected to injector output 1 (INJ1). Wire your primary injector to both IDC1 and IDC40, and your secondary injector to both IDC2 and IDC39.

Before wiring it up, you can just enable Staged, input your RPM signal, and watch the Speedy LEDs for output in order to verify this with no injectors attached. Set for Automatic and increase the RPM signal and one set should brighten with RPM, then at the point the primaries are maxing-out DC, the secondary channel LEDs should begin to brighten as they engage.

It worked fine for me last time I tried it, and I assume it still does; but if things are not working as you expect I would load the Speedy Base Tune and set for your trigger and Staged in order to verify INJ output LED operation as I described. Once you know how it's supposed to act, then you can reload your 2S tune and look/test for similar operation. David

I will try David's suggestions tonight.

[TZ350]

Thanks David for the heads up. Posted the issues on github and Josh is looking at it. And it appears there is a software issue. I am looking forward to it getting sorted as I am very keen to move on from bench testing to actually running the Speeduino on the bike.

In this Carb vis EFI comparison they mention that the EFI bike had much more fuel in the tank at the end of the day than the carburetor version did:- https://dirtbikemagazine.com/friday-wrap-up-2-stroke-shootout-carb-vs-injection/

The back story of KTM's 2S fuel injection efforts:- https://dirtbiketest.com/fresh-dirt/two-stroke-fuel-injection-ktms-path-to-tpi/#2sxjFlWDrXgwim9R.97

An Enduro 2S bike was successfully fuel injected here in NZ a few years before KTM:- look for GerbilGronk https://www.google.co.nz/search?rlz=1C1CHBF_enNZ743NZ743&biw=1199&bih=575&tbm=isch&sa=1&ei=JHiUW9bJCMvW-Qa38KSQAQ&q=+GerbilGronk+site%3Ayoutube.com&oq=+GerbilGronk+site%3Ayoutube.com&gs_l=img.3...54225.55405.0.56210.2.2.0.0.0.0.255.4 64.2-2.2.0....0...1c.1.64.img..0.0.0....0.f4F0uIqeHPQ

I am into small 125cc 2S road racers and are looking to push a EFI two stroke past the 10k rpm 10bar BMEP barrier and because it can be modified to suit the needs of a high performance 2S I see the Speeduino as an effective tool for doing that.

Going past the 2S 10k 10bar barrier is like climbing in the death zone on Mt Everest. It is still just climbing only a little technically trickier. Once its been done I guess it will become common place. Can't wait to get the Speeduino and Bike back on the dyno.

[TZ350]

.
Basic initial settings I used to get started.

338926 338927 338928 338934

[TZ350]

.

All the basic settings to get single cylinder two stroke staged injection working.

338930 338932 338931 338933

Now that I have got my head around Speeduinos two stroke staged injection. The next move is to get a sample and hold crankcase peak pressure measurement system working. Where the crankcase pressure is sampled at a pre determined point in the cycle and the reading is then used as the MAP value for the rest of the cycle. By using the change in the difference between maximum and minimum crankcase pressures as a MAP value should hopefully work well with a VE table.

[pete376403]

Are you using the basic (free) version of TunerStudio or did you get the paid for version? I've got a Speeduino NO2C (small form factor for want of a better description) which wil hopefully end up on the KLR. BMW1200 throttle body, havent got a pump yet and next major will be the "bumps" on the flywheel (currently only 1)

[TZ350]

.
Full paid for Tuning Studio version. And you will need a toothed flywheel. I suggest 24 or 36 with 1 missing tooth.

[TZ350]

338946338945


Hi TZ350, I don't know if you remember the Bimota V Due, a tweak RGV250/Aprilia IIRC which had "direct fuel injection" Well it was overpriced and broke Bimota, but people are revisiting this 1990s bike, and a French company have transformed it using a different ECU, and revised injector location. I saw it and thought of yours (like OEM Bimota) and the newer better system. It apparently makes the bike so much more rideable and has a wider spread of power. Pics for reference...

Thanks Panpan, great pictures of the Bimoto injectors. I have previously tried both positions and for a relativly high performance 2S like the Bimota from my experience those positions will prove inferior to the direct down the B transfer ducts with the primary through the C port. Anyway, for what its worth that has been my experience.
.

[TZ350]

.


Interestingly and to my surprise I have managed as much hp out of the original GP125 air cooled cylinder as I have recently out of the engine after it was fitted with the NSR cylinder.

You have an ancient GP125 fitted with a Honda NSR125 top end?

Yes, 1980's Suzuki GP100 commuter fitted with a NSR250 cylinder, crank de stroked for 110cc. The spacer in the middle is so the 5 speed gearbox can be replaced with a 6 speed TS125 unit.

An ancient GP with everything plus fuel injection..... 31 RWHP


338947 338949 338948

[TZ350]

338950


Staged injection looks Good. Next move is a crankcase pressure sample and hold pseudo MAP value.

[TZ350]

Hi TZ, I'm glad somebody else is trying the Speeduino 2T route. I am in the process of building my system, using a V4.3 board. It's funny how you're using a NSR250 cylinder as that's the engine I'm hoping to convert. I am trying to follow GerbilGronk's method of using transfer port injection, except I plan on using all four A&B transfers.
I have had no trouble making power with EFI and a single NSR250 cylinder for 125cc. The injectors I used were:-

Primary injector, is one only 38 g/min about 50cc/min in the C (boost) port. Very important to have a small as possible primary injector for widest range of tun-ability. Anything bigger and the injection squirt was not able to be dialed down low enough for correct air/fuel ratio at low load small throttle settings.

I found a single primary 60 g/min about 80cc injector was to big to get correct low load A/F. so for the NSR you are pretty much stuck with a single primary injector.

Secondary injector is two 128 g/min injectors in the B transfer ducts for a total of 256 g/min or about 340cc.

Primary plus secondary is about 390cc total.

With 2S and EFI, for tun-ability you must use the smallest injectors possible.

The four B plus A transfer duct injectors are a great idea but it will not be easy to put a secondary injector in the A transfer duct behind the NSR's Power Valve pulley. This would be more achievable on 250 or 350 single where there is more working space for four injectors.

I found it difficult (actually impossible;e) to weld completely around the injector plug so I just welded the easy to get to part and then used Devcon F Aluminium Putty to seal where I could not reach with weld.

338951

[TZ350]

I too have been thinking about the problem of 2T airflow measurement as realistically the engine has two distinct characteristics i.e. either in the power-band or out of it and as you say the pipe is the key.

Hi DaveyB, its great to have someone else working on this too.


I am still pondering how the air that is sucked into the pipe and then pushed back into the cylinder can be measured.

I would not worry about accounting for the spilled air that is pushed back into the cylinder because it can be accounted for with the injection table at injection time because the spilled air will be consistent. Consistent big, or consistent small . It can be accounted for by the table when the engine is on the pipe and when it sort of is and when its not.

[TZ350]

.
Measuring the Pressure in the Exhaust Pipe.

I have tried this and found that the pressure at full power is slightly less than the pressure at idle.

Easy to understand. A negative suction wave passes the measuring point and the wave is reflected back but with reduced strength as a positive pressure wave. The greater the strength of the suction wave the lower the average. But the average between big or small is not that much different.

The only way to make a measurement that can determine a meaningful difference between high or low energy waves is to make the measurement at a set time in the cycle.

For a 2S we need to be able to take an instantaneous reading at a set crank angle.

338952

Data logged with the Ecotrons software I was using at the time. Data recorded during a WOT dyno pull.

White line is TPS, Brown is RPM topping out at 12500 RPM and the Yellow line is a MAP reading. It did not matter where I measured it, in the crankcase or 100mm, 250mm, 500mm down the pipe or at the mid dwell section of the exhaust I always got the same result.

Pressure dropped slightly when I opened the throttle and started making power. In a 2S the pressures measured by a MAP sensor at idle is close to ambient and the average pressure drops slightly when the the throttle is opened. All a bit different to what you would expect from a 4S's MAP reading.

The reason the yellow line is so flat is because it is reading the average pressure. To get a useful 2S MAP reading it would have to be instantaneous and taken at a set crank angle, not over the whole cycle like it usually is in a 4S.

[TZ350]

.
Although the pipe is doing the work there is no real reason to make the measurements there. The pipe is like a hammer and all we need to do to know how effective it is being is by measuring how far the nail is driven in with each blow.

Measuring the changes in peak crankcase pressure at a set crank angle is enough to know how well the pipe is working.

[TZ350]

What kind of readings do you get between the throttle plate and the crankcase valve - instant or average?

I have not tried that myself as I expect it does not give useful results and is not at all like a 4S would be.

From the Ecotrons Manual:- ""As you know, 2-stroke engine does not have much meaningful manifold pressure to be used as air charge detection. In this case, you only need to tune the "TPS based Load Mapping" table.""

Unfortunately the "TPS based Load Mapping table." or Alpha-N is not enough either when venturing into the 2S death zone.

[TZ350]

TZ350 I might be being dumb here, but have you measured the exhaust temperature in the pipe?
I don't know if you are aware but there is a correlation with temperature, that equates to the combustion temperatures and when the pipe is working better. So if the exhaust is around 520-600ºC the pipe should be performing pretty well, below this, it isn't doing much, and as it goes towards 720ºC it is very likely in a lean burn condition, as the pipe is much shorter, and dragging fuel out of the cylinder.


So there maybe scope in using temperature sensing to activate fuelling changes when the pipe is working better, or adversely affecting things.


There are some issues with that approach like leaning out the mixture to heat up the pipe for more over reeve, and the situation where detonation heats the head and cools the pipe. If the mixture is then leaned out because of the cooler pipe the detonation just gets worse.

[TZ350]

I was hoping I made a good decision on the injectors, but if my two primaries of 107cc/min @3 bar equal 214cc/min are too big I’m stuffed. These were the smallest I could find at the time.

You might be stuffed.

38 g/min (50cc) is what I use for the primary injector. I found 60 g/min (80cc) with a min on time of 1.7ms was to big for a 125cc cylinder. The smallest injector available in the style I use is 30 g/min.


338953

[TZ350]

.
I developed a Arduino Nano to measure the crankcase pressure at a set crank angle and present that as a pseudo MAP value to the Ecotrons.

The Ecotrons could log the changing MAP values but because the Ecotrons 2S software would only run a Alpha-N table it could not do anything with the information. As much as I begged or tried to explain, Ecotrons was not interested in making the VE table available in their 2S software.

Making VE available would not have been much of an issue for Ecotrons because their 2S system was just their 4S program with the VE table turned off because the MAP value as measured in the inlet tract of a 2S did not produce meaningful information.

Now I need an enhancement for the Speeduino that allows for an instantaneous MAP measurement at a chosen crank angle.

First picture shows the variations in peak viz min crankcase pressures at different power conditions.
Second picture is the Arduino Nano. The MAP sensor is hard to see but it is plugged into the green hose on the brass crankcase plug.
Third picture shows the MAP value as recorded by the Ecotrons logging system. The yellow circle shows where the step is that we saw in the averaged MAP lines previously.

338955 338954 338956

[TZ350]

338957

The picture shows the variations in peak viz min crankcase pressures at different power conditions.With that, I can see using Speeduino instantaneous MAP, timed to ~135° (assuming pressure timing slightly advancing at lower rpm — more rpm plots needed), then using MAP multiply to modify the Alpha-N fueling calculation. Would be a place to start, anyway. 65 to 130 kPa is a good range of values to work with.


Conversely, Speedy can do minimum MAP value of each cycle, but change it to max, and use that with MM as an un-timed max value. This would promote greater fueling with higher max, and less with lower values. That's the direction you need to go when twerking the throttle :lol:, in order to recover after it has fallen off-pipe, yes? This last option would likely be simpler to mod the code for testing.


Have you tested response using Speedy with basic MM on crankcase pressure yet?


David


Hi David. Yes I will do some more plots at different RPM.

[TZ350]

.
Some simulated crankcase pressure traces with the minimum in the middle because I wanted to see if there was anything useful there, but I don't think so. Open to suggestions though.

The peak on the left looks like air that went into the crankcase then bounced out again, as it does. Or more likely, exhaust gases back flowing down the transfers into the crankcase.

This pollution from the exhaust may explain why the minimum crankcase pressure can be the highest and much the same at idle and low RPM whatever the throttle position. The exhaust pollution in the crankcase may also explain why the motor idles when the throttle is near shut in spite of the average crankcase pressure being the highest, much higher than at WOT.

A 4S motor has a throttle plate with an inlet tract attached to a fairly good piston pump so its inlet MAP is low at idle and high at WOT.

A 2S also has a throttle plate but its inlet tract is attached to something that acts more like a pulse jet engine than a pump.

A 2S crankcase MAP is high at idle and lower at WOT. So MAP behavior for a 2S is quite different to what we are used to from a 4S.

My current thinking is to take the difference between the minimum crankcase pressure seen in the middle of the graph and the maximum that is seen on the right hand side where the piston is compressing the actual air that was trapped in the crankcase and using the difference as an air flow indicator.

To me this difference seems to give the best indication of mass air flow. With the difference at 65% being about 0.33 bar, 15% 0.20 bar and 5% 0.15 bar at a rough guess.

If we do something like MAP = (Max - Min) * 3

Then 65% on the pipe would be 0.99 bar
And 15% on the pipe would be 0.60 bar
And 5% on the pipe would be 0.45 bar

They Look like useful traditional 4S MAP values that Speeduino can use and off the pipe situations would easily show up as a lower MAP values.

,338958 338959 338960

.
Like most people I guess, I just assumed that a 2S crankcases pressure behaved like a 4S inlet manifold. Low pressure when the throttle was closed and high when the throttle is open. But for a 2S that is not true, in fact its the complete reverse and another example of the mistake of trying to apply 4S thinking to a 2S.

I expect rotary valve, 24/7, reed, case reed or piston port would all have the same crankcase pressure scenarios. If anyone wanted to, they could use something like EngMod2T to check.

The lowest crankcase pressure is with the piston at TDC, and that makes sense.
The highest crankcase pressure is with the piston about 140 ATDC, sort of makes sense too.

At 140 ATDC the inlet is closed and the transfers and exhaust port are open. Some mixture is flowing out the transfers but not fast enough for all of it to escape before the piston gives the remaining mixture in the crankcase a hurry up.

The pipe should be starting to suck real hard around that point too, so you have a real push pull situation accelerating the transfer of fresh charge into the cylinder.

If the pipe is not sucking that hard you will have a higher residual pressure left in the crankcase. That explains why at idle and low throttle at low revs the crankcase pressure is higher than when the throttle is wide open and the motor is singing on the pipe. With high crankcase pressure exhaust gas pollution explains why when the throttle is closed the motor slows down to an idle.

At idle the motor is running on a very small amount of fresh air mixed with a lot of pollution that has back flowed from the exhaust. I am not sure if the crankcase is full of pollution or its full of clean mixture and only a small amount is being transferred to a cylinder full of un scavenged pollution, maybe a bit of both so the story is not completely told yet.

339048

But one thing is for sure, the crankcase pressure is higher at idle and low throttle than it is at wide open throttle with the motor making real power.

[ceci]

.


Yes, 1980's Suzuki GP100 commuter fitted with a NSR250 cylinder, crank de stroked for 110cc. The spacer in the middle is so the 5 speed gearbox can be replaced with a 6 speed TS125 unit.

An ancient GP with everything plus fuel injection..... 31 RWHP


338947 338949 338948

Tu ya estas al final del camino, está a punto de alcanzar tu éxito.

Yo al contrario esto al comienzo del camino, y me falta mucho hasta que llegue alcanzar el éxito.

Por ahora solo he realizado las dos primeras fases:
1º El estudio de que piezas, de que motores, me valen para hacer mi prototipo
2º La adquisición de las piezas principales y de mayor importancia (están tal como las he recibido).


Aun me falta:

3º El adaptarlas y unirlas entre ellas

4º transformar cada una de estas piezas para que realicen las funciones extras que realizaran cada una de ellas.

Nuestros caminos son paralelos pero distintos, distintos ya que yo voy a emplear la inyección mecánica en lugar de la electrónica.

Si he reflexionado y ya que mi prototipo tendrá un circuito de aire secundario creado por un pistón bomba auxiliar, el empleo de un mini carburador para alimentar el motor de la mezcla necesaria; esto es a modo de lo ya empleado por:

ttps://patents.google.com/patent/US4276858

Con diferencia de donde tiene que ir el carburador y como emplear esa masa de aire secundaria

[ceci]

3º El unirlas entre ellas

[TZ350]

Tu ya estas al final del camino, está a punto de alcanzar tu éxito.

Thank you for your post, I translated and read it, you have a very interesting project there. But as we are a former English Colony best to post here in English.

[Drew]

The fuck is that rooster on about?

[ceci]

Thank you for your post, I translated and read it, you have a very interesting project there. But as we are a former English Colony best to post here in English.


I do not know English, I use google translate.
When I review what I have translated, sometimes it is not what I want to say.
So if it's something important I put it in Spanish and everyone who interprets it

[Drew]

I do not know English, I use google translate.
When I review what I have translated, sometimes it is not what I want to say.
So if it's something important I put it in Spanish and everyone who interprets it
Google is how we translate anyway.

To make sure its correct, translate to English then back to Spanish. The errors will be clear.

[TZ350]

I do not know English, I use google translate.
When I review what I have translated, sometimes it is not what I want to say.
So if it's something important I put it in Spanish and everyone who interprets it

Good on you Ceci, thanks for making the effort, I enjoy your posts.

[ceci]

Good on you Ceci, thanks for making the effort, I enjoy your posts.

Thanks to you, your publications are more relevant than mine, I just try to give another approach, but always respecting yours.
The approach that I have chosen is mechanical injection, even so I will need your contributions since I will have to use some electrovalve

[ceci]

.

But one thing is for sure, the crankcase pressure is higher at idle and low throttle than it is at wide open throttle with the motor making real power.

With these data that you contribute, you make me imagine what was the result of the test of this prototype.
Which I assimilate them to an experience lived in the youth, in which an Amal carburetor was obstructed the low conduit (without air filter, exposed) throttle closed, the engine revolutionized

[ceci]

With these data that you contribute, you make me imagine what was the result of the test of this prototype.
Which I assimilate them to an experience lived in the youth, in which an Amal carburetor was obstructed the low conduit (without air filter, exposed) throttle closed, the engine revolutionized

I'm constantly discovering that you always go hiding or hiding things, in the case this prototype I begin to understand why I do not progress

[TZ350]

.
My first look at the Speeduino code and at first sight it looked very confusing.


But after figuring out the INI file I have now got a 2S MAP option displayed in Tuner Studio.

And now that I have found in the sensors code where the MAP value is read and any calculations for instantaneous, average or minimum is done. All I have to do now is develop the code needed for the 2SMAP, Diff = (Max - Min)*3 algorithm.


.339085 339086 339083

[ceci]

The injection is advancing in the 2S are joining more manufacturers




https://www.youtube.com/watch?v=abzZe-r0ams&hd=1

http://www.enduromagazine.com/wp-content/uploads/2017/11/Jotagas-motor_9085-copia.jpg

Jotagas 300 EFI ,EICMA 2018

[ceci]

The injection is advancing in the 2S are joining more manufacturers




https://www.youtube.com/watch?v=abzZe-r0ams&hd=1

http://www.enduromagazine.com/wp-content/uploads/2017/11/Jotagas-motor_9085-copia.jpg

Jotagas 300 EFI ,EICMA 2018

https://i1.wp.com/thehellteam.blog/wp-content/uploads/2015/09/img_0936.jpg?ssl=1

Vertigo and Jotagas indirect injection

[TZ350]

.
Thanks Ceci for the pictures. It is great to see 2 stroke EFI making an appearance on more bikes.

339190

Ok, so the picture is of a Beta trials bike but two stroke EFI must be heading in a good direction if the latest Jotagas 300 is going to be using fuel injection.

339191

https://ultimatemotorcycling.com/2016/01/02/2016-vertigo-combat-camo-works-exclusive-first-test/

http://vertigomotors.com/en/combat-vertical-2019-3/#pll_switcher

And I see that the latest Vertigo Combats come in 125 - 250 and 300cc all fuel injected, very impressive. Thanks again Ceci for the heads up.

[ceci]

Thanks to you, I am passionate about this subject as you .
The TPI in port C see it as a possible solution for urban 50cc

[TZ350]

.
Just about got my engine back together and ready for its Speeduino EFI system.

I have looked at the IMAP/EMAP control methodology but I think most of the problems with 2S EFI is that people try to adapt a 4S approach to their 2S project but that invariably turns into disappointment.

A 2S demands a uniquely 2S approach appropriate to the style of 2S we are working with and there are lots of different flavors of 2S. A high performance, pipe resonant unit that revs past 10k rpm needs staged injection and a way of seeing changes in air flow through its crank case.

Now my motor is coming together again the race is on to get a Beta 2S Speeduino version that measures the high/low crankcase pressure and converts the difference to a usable MAP value. I have been trying my hand at developing the code myself.

339227 339228
.

[TZ350]

.
Engine back together after getting a new rod kit and the direct B/E oiling modification. Getting the piston to head squish at 0.65mm and a compression ratio of 14:1 took a bit of time.


The 24 - 1 tooth trigger wheel ring was made from an Aprilia SR50 flywheel and pressed onto the original GP-NSR110 flywheel. Hopefully get the engine back in the bike next week and up on the dyno the following Sunday for an initial startup attempt .
.

339320 339321

[TZ350]

.
Going to try my hand at programming an Arduino Nano with an add on digital potentiometer to give me a 24-1 crank speed trigger wheel and a sign wave of sorts from the digital potentiometer representing high and low crankcase pressure.


This will give me a 24-1 crank speed trigger wheel simulation and a simulation of high and low crankcase pressure for testing my 2SMAP coding changes in the Speeduino firmware.


The digital potentiometer uses a X9C103S chip and has the normal potentiometer 3 legs on one side. On the other side is Vcc/Gnd and three other pins for increment up or down.
.
339357

[TZ350]

339471

I have now got my simulator working and it can sweep automatically from 1000 to 13000 RPM. Took two weeks but I learnt a lot after hours Googling the answers. The red line is teeth and the height of the green line sign wave that simulates crankcase pressure can be varied with a potentiometer. Works well and stays in sink all the way to 13,000 rpm. I feel very pleased with it. Now I can de-bug my modified Speeduino code. Using the simulator to get everything right sure beats trying to do it on the bike.

339469 339470

[TZ350]

339492
.
I have finally got the simulator working and the modified Speeduino code that reads the MAP sensor working properly.

I now have a fuel injection system that measures the high and low pressures in a two strokes crankcase and displays the difference as a valid MAP reading that the Speeduino can use in its other tables.

The next step is to get the Speeduino hardware installed on the bike and give it a twirl on the dyno to see if we have got rid of that pesky problem with coming back on the throttle again after shutting off from a WOT run. If this works it will be as far as I know the worlds first ever publicly accessible EFI firmware that works for high performance two strokes that run above 10,000rpm and that make more than 10bar BMEP by relying on their pipe resonance for power.
.

[TerraRoot]

super excited for this
also love the BE oiling pump setup, never would have thought to just have it at one speed,independent of engine revs, race bikes eh?

[TZ350]

super excited for this, also love the BE oiling pump setup, never would have thought to just have it at one speed,independent of engine revs, race bikes eh?

Yes I am pretty excited too, hopefully it lives up to expectations, but I am pretty confident I am on the right track. Otherwise its egg on face.

There is a potentiometer that controls the VSD's speed. I have not measured the voltages but it may be that the pot wires could be connected to a TPS and the throttle would then control the amount of extra oil.

[TZ350]

.
I have attached the simulation software, it is not very clever and the coding is a bit rough. It runs on an Arduino Nano but I guess it could be adapted to any Arduino.

It simulates a 24-1 missing tooth crank wheel and RPM from 1000rpm to 12,000rpm and high and low crankcase pressures in a two stroke crankcase. Low being at TDC and High being at BDC. To get other tooth options is just a matter of changing the number of iterations in the code that the loop is expected to do.

A short coming is that the sweep across the tachometer face gets ever quicker as the rpm increases and the time per revolution gets ever shorter. I tried to fix this with a loop that forced the crank to keep rotating for a set amount of time each rpm step but could not get it to work.

The main rpm loop uses one "for" loop for both increasing rpm and decreasing it again. The square wave pulses are feed directly to the Speeduino's crank sensor input. The sign-wave pressure trace is generated by a digital potentiometer board with a X9C103 chip for Arduino's and takes a step for each tooth. It behaves like a normal 3 leg potentiometer, with 12 steps up and 12 steps down to vary the output voltage (it is a 100 step pot so it could be as much as 50/50 steps if you wanted). I feed the sign wave pressure signal to the external map sensor in the proto area.
.

[TZ350]

339548 339549

Time to strip out the old Ecotrons ECU and fit the Speeduino unit with the modified 2SMAP sensor code.

339547

Goodby to an old friend, the Ecotrons ECU. Without meaningful support from Ecotrons, and their help dept is a nightmare to deal with, I have given up on them. After a lot of careful work I have to say that their 2S EFI system as it is, is useless for high performance two stroke fuel injection.

I would have liked to stick with it but all my requests for a simple box tick so I could use the VE table was meet with a demand for $500 USD. Goodby Ecotrons, your loss not mine.

Hello my new friend Speeduino where I can make whatever changes that my be required myself.

[TZ350]

.
Finished all the bench testing of the Speeduino and familiarizing myself with it and Tuning Studio.

339617

This is how Speeduino can be setup to run a fuel pump. Primes the pump when Speeduino is first turned on then runs the pump continuously when the Speeduino detects RPM.

339618

I tried my ignition idea tonight. I have Speeduinos Ign 1 output connected to the Ignitec crank sensor input. It would spark a few times then stop. I think the Ignitec ignition input is a very high impedance and the Speeduino output is also a very high impedance when its switched low. So things get pumped up to 5 Volts and the Ignitec does not get to see a low voltage and re set itself.

Tomorrow night I am going to tie the Speeduino's ignition output to earth with a 100k resistor. hopefully that will allow a reliable 0 - 5 - 0 Volts swing at the Speeduino's and Ignitec's ignition terminals.

[TZ350]

339620

I have now got the Speeduino and Ignitec talking to each other. I am using the Speeduino's number one ignition-out locked at 15 deg and the Ignitec base set at 15.

For things to work I had to tie the Speeduinos ignition output to earth with a 10K Ohm resistor that way the ignition pulses went to 0Volts between ignition events and the Ignitec could tell what was going on and when to fire.

339621 339622

Relevent Ignitec and Speeduino set pages.

[ceci]

Hi TZ350, sorry for filtering this here.
I want to know your opinion
https://www.facebook.com/groups/671081149638729/permalink/1972578262822338/

I believe it is a conduit for a pressure sensor and NTC.
Operation: when discovering the conduit the skirt of the piston (in each cycle) the pressures of both are equal, the sensor registers if there was variation.
Advantage: it registers the actual load of the crankcase in each cycle, in a more simplified way since it only records the maximum and not all the crankcase pressure.
Utilization: control of the powerjet for a correct use

[ceci]

https://scontent-mad1-1.xx.fbcdn.net/v/t1.0-9/46485775_1923435204419248_4239909319958593536_n.jp g?_nc_cat=103&_nc_ht=scontent-mad1-1.xx&oh=1385bd1ee782bc9eb7aaa9000af3d3ab&oe=5C6C704Bhttps://scontent-mad1-1.xx.fbcdn.net/v/t1.0-9/46480632_1923435171085918_6581101653775089664_n.jp g?_nc_cat=106&_nc_ht=scontent-mad1-1.xx&oh=a93a43067527828a332cb6b2540bdb84&oe=5C76B815

I analyze what function this conduit can have and I can only find the reason that I already mentioned (transmitting the pressure of the crankcase to a MAP sensor)

https://www.kiwibiker.co.nz/forums/attachment.php?attachmentid=338954&d=1537352479

I think having more data is not beneficial. You have to separate the valid from the rest (use of time to perform this task). It is not necessary to know all the pressure of the crankcase throughout the cycle
Using as a reference your data in which you quote that in idle the maximum pressure has the crankcase and that the piston being in TDC is the lowest for each cycle.
based on this concept the starting pressure value would be 1 and as the rpm motor goes up we would go to less

[TZ350]

I want to know your opinion
https://www.facebook.com/groups/671081149638729/permalink/1972578262822338/


Hi Ceci. I have seen that picture before on the Team ESE thread I think.

From memory it was an injector to overcome a problem with a particular rider who would crash down gears and rev the engine excessively in overrun causing seizures. The injector was to add a bit of fuel and extra oiling while the throttle was shut during overrun.

If you have the time you could search the pictures on this thread:- https://www.kiwibiker.co.nz/forums/showthread.php/86554-ESE-s-works-engine-tuner/page2066 ... a picture leads to the related post.

[ceci]

Hi Ceci. I have seen that picture before on the Team ESE thread I think.

From memory it was an injector to overcome a problem with a particular rider who would crash down gears and rev the engine excessively in overrun causing seizures. The injector was to add a bit of fuel and extra oiling while the throttle was shut during overrun.

If you have the time you could search the pictures on this thread:- https://www.kiwibiker.co.nz/forums/showthread.php/86554-ESE-s-works-engine-tuner/page2066 ... a picture leads to the related post.


At the moment I do not have my press collection of the motorcycles of past decades.
If I remember that in the report of the KTM 125 GP mentioned carburation problems in certain situations, what I do not remember is that they solved it with an injector.
A powerjet varies the mixing ratio independently of the air flow (the important thing is to know when and how it has to vary) and how do you know if you do not have data?

https://i.skyrock.net/0978/91930978/pics/3279384954_2_16_WBPKxt6f.jpg

https://marcndm.skyrock.com/3279384954-KTM-FRR-125cc-suite-2.html

These according to indeican in this page are the carburetors used, they are ELECTRONIC carburetors
The theory that I have is the most economical and simplest (no injection pump needed "is not injected only with an injector, you need a pump and many other things").
A direct injection to the crankcase to enrich the mixture I see it very complex for something that the powerjet makes you

[TZ350]

.
Here you are Ceci, I found the original post. You are right, it was instead of a power jet. Maybe my memory of an injector being used to eliminate seizures from overrun was an Aprilia GP thing.
.

Not likely; the 125 cc KTM GP-bike already had an injector in the crankcase that injected through a port below the exhaust port.
They used it instead of a powerjet and it would only have been logical to use the same setup for full EFI.
328804 328805 328806

[TZ350]

.
Got the Beast running today, just.


The MAP value varies depending on the difference between the High and Low pressures in the 2S crankcase. Close the throttle and the difference is less and the MAP value is lower. Open the throttle and the difference increases and the MAP value goes up. Just like it does with a 4T.

Hit a bit of a snag, at low speeds, not much more than idle (1000 to 3000 rpm) that I have been able to run I have found that the difference in pressure is much greater with the throttle closed than it is when the throttle is partially open.

Makes sense though, a closed throttle allows the descending piston to give the trapped mixture a good squeeze but any pressure applied to the mixture when the throttle is open just allows it to escape out the inlet and so the pressure difference is less.

[TZ350]

I find this description on Ecotron site

2-Stroke Small Engine Fuel Injection kit is designed to run 2-stroke small engines. This kit can drive 2 injectors per cylinder (one big injector, one small injector). During idle, and low part loads, ECU only opens the small injector for small fuel quantities; it will switch to the big injector for mid, or mid-high loads; and it will activate both injectors at WOT conditions. The transitions between the injectors are transparent to the driver and they are so smooth that the driver can not feel it. This setup controls the accurate AFR (lambda) from idle all the way up to 16000RPM. We have a

The whole idea of 2 injectors (one small, one big) is based on the need: 2-stroke engines have only half the time to inject fuel compared to 4-stroke (360 vs 720 degrees), and the 2-strokes usually (esp. high-end engines) have such a wide range of RPM (2000 - 16000rpm, for example). One injector is not able to cover the whole RPM range, simply because every injector has its physical limit: the fixed flow rate. If you use a big injector to cover the high end, then you will have too-rich idle. If you use a small injector, to have a good idle, then you don't have enough fuel for WOT. Given a certain inject flow rate, you can only run an engine either at low RPM range or high RPM range, but not both (from 2000 to 16000rpm).
For example, an engine runs at 16000rpm, you only have max 3.75ms time-window to inject fuel. In this short time, you have to inject enough fuel for WOT conditions. If you pick a super big injector, you will not be able to run idle (too rich even at 1ms pulse width). So what can you do? add more injectors. At low load, use one injector, and at high load/high RPM, use 2 or more injectors. That's why you see a lot of racing engines have 2,3, or even 4 injectors per cylinder. Our system is better: while others use the same size of injectors, we use one small, one big injector. Why? because at idle, you need a very small size of inject, and at WOT, you need a big size. The benefits: you can run the engine at ideal AFR over the whole wide range of RPM, with only 2 injectors (again, save cost compared to 3 or 4 injectors). The tansitions from idle, to WOT, will be: the small injector works at idle, low load; then it switches to the big injector at mid-high load (the small one shuts off); then at WOT both injector works together. And all these are made transparent to users!
This is our unique technology for 2-strokes, not even big companies have this.

I want to implement same or similar in my EFI. Is this specs rational enough or someone advice corrections?

Yes totally on the money but only half the story.

Ecotrons use the Alpha-N topology for their 2S fueling map and it works well on 2S's that are not very pipe dependent like 250cc Enduro bikes but Alpha-N does not support high performance 2S's like GP125's where they make a lot of power through clever use of their pipe.

Because Alpha-N is a fixed pre determined fuel squirt amount for every TPS vis RPM cell on the MAP. This only works half the time for a high performance GP engine because AN cant be both correct for when the engine is on the pipe and when it is not.

I use the 10/10 concept. Anything less than 10,000 RPM and 10 bar BMEP is probably possible with an Ecotrons but anything making more than 10 bar BMEP and 10,000 RPM needs something more sophisticated for their fuel mapping.

My current EFI 2S efforts are aimed at solving this and am looking for ways to see the actual changes in airflow through the crankcase. Currently I am looking at the difference between maximum and minimum crankcase pressure and how it changes when the engine gets on or comes off the pipe.

[patouille]

Hello,
I just registered in this forum as I saw this very interesting topic. I'm glad to see someone is trying and sharing his ideas.
I built my own 2S EFI for my racing Solex (a very particular 50cc here in France).
The beginning is very encouraging, take a look : https://www.youtube.com/watch?v=LW19l64nm7Y.

I wanted to share some of my feedback which is sometimes different form the TZ350 observations.
I made my own printed circuit board (still based on the arduino mega chip). The code is from 4S codes I found in the internet and adapted to my use.
The main purpose was to make it run at idle but it worked beyond my expectations. As you said staged injection is more appropriate, I was planning to use it but it turned out that my Suzuki 600GSR injector could achieve the best idle stability (~1100 µs PW at 2200 RPM) I ever had with any carburetor (which is of course way bigger than necessary) !
The second thing different from your observation is injector timing (but keep in mind it's a throttle body injected engine). The injector is triggered by just the magnet of the ignition rotor in front of a hall sensor. I added a triggering offset tuning possibility but with no noticeable effect.
My algorith is working in alpha-N because I could never have a significant MAP reading (I tried before the throttle plate, after and in the crankcase). Your idea of the sensor reading triggered by a specific crankshaft position is clever. I will probably try that.
Despite the bike achieving all of its regular RPM range in the video (~14000RPM), it was with the bike on its stand. On the road I'm currently having some trouble over 10000RPM. I tried different mappings with no effect. I'm still working on it...
For more info, you can see our facebook page (https://www.facebook.com/TeamMothersoxer/) (sorry in french).

[ceci]

.
Here you are Ceci, I found the original post. You are right, it was instead of a power jet. Maybe my memory of an injector being used to eliminate seizures from overrun was an Aprilia GP thing.
.

Thank you very much TZ350 for the information.
KTM takes pole position in 2S injection research.
Three models: crankcase, cylinder and transfer port.

[speedpro]

KTM takes pole position in 2S injection research.

KTM would like you to think so

[TZ350]

KTM takes pole position in 2S injection research. Three models: crankcase, cylinder and transfer port.

This has all been done successfully here in New Zealand well before KTM. Possibly even the inspiration for KTM with regards to how to get their overly complicated first attempts right. They certainly knew about these bikes before they abandoned what was not working for them and changed to B port injection themselves too.

Some Youtube clips of transfer port and crankcase injected 2T's None of this is my work. The Kawasaki and Yamaha were running well before KTM's EFI 2T offering.

EFI Kawasaki Bighorn on the dyno https://youtu.be/ifSEql1X4R0
EFI Kawasaki Bighorn Trail ride 1 https://youtu.be/eleqBGvOM4M
EFI Kawasaki Bighorn Trail Ride 2 https://youtu.be/CnIemdISKrM
EFI Kawasaki Bighorn VMX https://youtu.be/v8i6ps0PdVI

Maramarua classic forest trail ride https://youtu.be/ZqzOgOJuJ1E

http://youtu.be/eleqBGvOM4M
http://youtu.be/hOGZ5llowoU
http://youtu.be/1YG9ko8-Nwk
http://youtu.be/UEQli7nuak4

GerbilGronk is worth a Youtube search.

339685 339686 "B" port injected Yamaha YZ250


339683 339684 Crankcase injected Kawasaki Bighorn

[TZ350]

.
Just in case KTM try to also claim they "invented" a sleeve valve two stroke. Well you have seen pictures of a running one here well before KTM have even started talking about it.

The most powerful piston powered WW2 aero engine was the sleeve valve two stroke "Rolls-Royce Creasy".

A 2S sleeve valve motorcycle engine with port fuel injection might be something .......... remember you heard it here first...... :motu: looking at you, KTM.

[ceci]

http://thedirtbikerider.com/wp-content/uploads/2017/07/KTM20182TLaunch_49.jpg

First cylinder: Orbital Di-tech (Aprilia leonardo SR500 and did that motorcycle work?).
Second cylinder: Bimota VDue and Sherco (Bimota has the patent for that method).

http://dirtbikemagazine.com/wp-content/uploads/2014/11/RON_1228web.jpg

https://patentimages.storage.googleapis.com/d1/eb/95/4e482a67a03d4e/WO1993022545A1.pdf
https://patentimages.storage.googleapis.com/da/77/0d/5372913799c1fd/EP0591509B1.pdf

If true the third is a copy of Flettner.
These three and the GP125 to the crankcase.
From my scarce knowledge they place it in the manufacturer that more to investigated

[ceci]

https://www.kiwibiker.co.nz/forums/attachment.php?attachmentid=339008&d=1537822468

Beta also tried to inject by that method 2

[TZ350]

If true the third is a copy of Flettner. These three and the GP125 to the crankcase. From my scarce knowledge they place it in the manufacturer that more to investigated

Flettner was there first with "B" port injection as seen on KTM's which seems the most practical and so far successful system but having got the idea KTM sure has more resources to throw at an innovative approach that was first seen here in NZ before KTM abandoned what they were doing and adopted it. Resources don't necessarily equal innovation but hats off to KTM for running with a good idea when they see it. But the cheeky buggers are trying to Patent it. I hope they can't as it is prior art that is known in the public domain and not their original idea.

[ceci]

How much does KTM declare is the difference in weight between the TPI and the carburetor?

To the first cylinder besides that the injector assembly is heavier, add the weight of this other mechanism.

https://patents.google.com/patent/DE102009041733B4/

The loss of hydrocarbons due to the exhaust in the three cylinders is the same

[TZ350]

.
I put the Ignitecs pickup back to its original place and fitted a new pickup for the Speedy. Now both systems are entirely independent. The red LED was just so I would know if the Speedy is seeing the new pickup.

Made a bit of a false start with a proximity sensor. It did what it should, it saw a flywheel and turned on but was useless at seeing teeth. Found my spare ignition trigger and hooked that up to the Speedy, it worked and had no problem seeing the 24-1 flywheel teeth.

Fired the Beast up and it idled away happily. It would not take any throttle but the idle sounded sweet. Time to connect the laptop and see if we can get things hopping along a bit better. Very happy.
.
339695

[TZ350]

Hello,
I just registered in this forum as I saw this very interesting topic. I'm glad to see someone is trying and sharing his ideas.
I built my own 2S EFI for my racing Solex (a very particular 50cc here in France).
The beginning is very encouraging, take a look : https://www.youtube.com/watch?v=LW19l64nm7Y.

I love it......

<iframe width="560" height="315" src="https://www.youtube.com/embed/LW19l64nm7Y" frameborder="0" allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>

[TZ350]

The injector is triggered by just the magnet of the ignition rotor in front of a hall sensor.

That is very clever, that could be the basis for a very simple and effective EFI system..... you have got me thinking. Get rid of all the firmware clutter that usually comes with a EFI system. A simple fueling map and a way of seeing changes in airflow induced by the pipe. All on something like a Raspberry Pie or Arduino Nano and it could be a winner.

[patouille]

A simple fueling map and a way of seeing changes in airflow induced by the pipe.
That's the point. That's why your idea of reading the pressure in the crankcase only at a very precise point of the crankshaft angle is better than the alpha-N algorithm.
Next step for me is to change my development bike. Given the power delivery of my solex is very poor (due to the roller sliding on the tire), it's not the most appropriate way to have a fine tuning :laugh:

[ceci]

.


Fired the Beast up and it idled away happily. It would not take any throttle but the idle sounded sweet. Time to connect the laptop and see if we can get things hopping along a bit better. Very happy.
.
339695



I'm very happy for you, and at the same time I envy you (it's normal, everyone who has a project wants to see it finished and successfully)

[crbbt]

The raspberyy PI only has digital inputs and outputs.

analog can be achieved with an 8-bit digital to analog convertor however

I am following your progress with a keen interest!

[TZ350]

... it turned out that my Suzuki 600GSR injector could achieve the best idle stability (~1100 µs PW at 2200 RPM) I ever had with any carburetor (which is of course way bigger than necessary) !

1,1 ms is very short, what is the injectors dead time? do you know how long it takes to open your injector far enough to be in the fuel controllable region.

339700 http://www.mre-books.com/sa135/sa135_5.html

[speedpro]

I've run the Ecotrons supplied injectors down to 1.5mS without problem. It's less than ideal but works OK.

[TZ350]

I've run the Ecotrons supplied injectors down to 1.5mS without problem. It's less than ideal but works OK.

Yes, I found that 1.5 worked but 1.8 seemed better, seems silly but it just sounded better.

[patouille]

1,1 ms is very short, what is the injectors dead time? do you know how long it takes to open your injector far enough to be in the fuel controllable region.



Yes it is but it works ! I even tried ~900µs and it's precise enough to notice the influence on the engine behaviour. I couldn't find any datasheet for my injector concerning the closing/opening timing.
By the way, I'm trying to add a voltage compensation for the opening time. Do you think it would be useful ?

[ceci]

KTM would like you to think so



The research never stops, they are many people and in many places where they investigate how to evolve the 2S.
KTM will not have its own ideas, but if you investigate what you find interesting (here is the largest number of investigations).

[speedpro]

By the way, I'm trying to add a voltage compensation for the opening time. Do you think it would be useful ?

Compensation is a feature of other systems for injectors and ignitions

[TZ350]

Yes it is but it works ! I even tried ~900µs and it's precise enough to notice the influence on the engine behaviour. I couldn't find any datasheet for my injector concerning the closing/opening timing.

I think you must be operating them in the dribble zone, but if it works, it works.

339723

I would love to know more about this injector if you ever find out any more technical info. Genuine ones are not hard to get, I checked Japanese auction sites.


I'm trying to add a voltage compensation for the opening time. Do you think it would be useful ?

Not something I have dabbled with but it seems to be a standard feature of EFI systems I have seen so must be useful. Maybe useful for engines with starters and there is a voltage drop when the motor is being cranked over.

[Drew]

Not something I have dabbled with but it seems to be a standard feature of EFI systems I have seen so must be useful. Maybe useful for engines with starters and there is a voltage drop when the motor is being cranked over.
Everything modern mists the secondries while the motor is turning over on the starter as far as I know. They must only be half open or summat though, since they never flood...except stupid fucking MV. They have a counter keeping track of starter revolutions as well as crank revolutions. So when the starter clutch slips even a tiny bit, the fuckers shut the coils down and just wind over and over.

But I digress. Lekky start injected motors just run rich as fuck to get themselves to fire.

[TZ350]

.
Unlike the Ignitec and Ecotrons EFI system which could both be triggered from the same pickup. The Ignitec needs a single tooth trigger and the Speedy a multi tooth wheel. So as well as a trigger for the Ignition I needed an additional and separate trigger for the Speedy EFI system.

Space is limited but the easy answer is to cut a slot in the side cover and glue the Speedy trigger in there. The red 12 Volt diode hanging from the primary Injector plug is there so I can see if the Speedy is responding to the new pickup.

339724

Setting up the trigger angle in Tuning Studio.

339722

On my engine with the piston at TDC the crank has to rotate CCW 205 deg for the number one tooth to align with the trigger sensor. The number one tooth is the first tooth after the gap.

339721

In "Trigger Setup" 205 is entered in the "Trigger Angle (Deg)" field.

[TZ350]

As you said staged injection is more appropriate, I was planning to use it but it turned out that my Suzuki 600GSR injector could achieve the best idle stability (~1100 µs PW at 2200 RPM) I ever had with any carburetor

Despite the bike achieving all of its regular RPM range in the video (~14000RPM), it was with the bike on its stand. On the road I'm currently having some trouble over 10000RPM. I tried different mappings with no effect. I'm still working on it...
For more info, you can see our facebook page (https://www.facebook.com/TeamMothersoxer/) (sorry in french).

Although it works for idle and high rpm no load. I think your injector might be "to big" to be operating in the controlled fuel zone at idle and low rpm and low Loads. And "to small" to deliver enough fuel in the short time available at high rpm and high loads on the road. Proper staged injection may be needed. Although working in the dribble zone then the controlled zone could be pseudo staged injection, interesting idea.

339725

[Iceroadracer]

Hello from here in Finland!

This is very interesting project to follow. I was struggling with the same things and problems in my own project (Honda CR500AF EFI). For the diffence is that I have microsquirt ecu. The biggest problems are kick back at startup. Several broken parts like: crankcases, clutch, kick start lever...

Another situation is define 2T load. Now I have alpha-N method. It is not accurate enough for iceroadracing. The engine should work range 3000-10000rpm. I've been thinking about for a few years Exhaust pipe and chamber pressure difference to determine the load. Megasquirt does not just give it a good chance. I'm thinking about whether it would be possible toa separate controller make "MAP" signal from exhaust pipe and chamber pressure difference. The pressure measured on both signals at the right time window and their pressure difference is made with a 0-5v signal to the ECU map input. In addition EGT correction factor. In principle, the result should be good compared to alpha-N method.

Some project images: https://team1000.kuvat.fi/kuvat/J%C3%A4%C3%A4rata/-09+CR500AF+projekti/
And a few test video https://www.youtube.com/user/TeamTonni/videos

PS: sorry bad english:o

[TZ350]

Now I have alpha-N method. It is not accurate enough for iceroadracing. The engine should work range 3000-10000rpm. I've been thinking about for a few years Exhaust pipe and chamber pressure difference to determine the load. Megasquirt does not just give it a good chance. I'm thinking about whether it would be possible to separate controller make "MAP" signal from exhaust pipe and chamber pressure difference. The pressure measured on both signals at the right time window and their pressure difference is made with a 0-5v signal to the ECU map input. In addition EGT correction factor. In principle, the result should be good compared to alpha-N method.

If you have a ECU that will take a MAP input then Alpha-N with a bit of MAP should be good.

I have experimented with a Arduino Nano and MCP4725 I2C DAC Digital to Analogue Breakout module for the Arduino. The objective was to find a suitable pressure reading and use the Nano to present it to the EFI's CPU as a MAP value. Most Arduino's don't have a analog output. They say they do, but it is really PWM. But by using a DAC you could develop something suitable for what you want with a Arduino Nano.

339732339733

Arduino Nano on a breakout board, add on MCP4725 I2C DAC and a MAP sensor on the bikes crankcase.

Some of Iceroadracer's project images: https://team1000.kuvat.fi/kuvat/J%C3%A4%C3%A4rata/-09+CR500AF+projekti/ .. Great work, very impressive.

[TZ350]

.
Another possibility is a NO2C 3rd party Speeduino board. MAP and two injector channels, so staged injection is possible. Code is open source so you could modify it to suit your needs.

https://speeduino.com/wiki/index.php/3rd_party

339735

[patouille]

Although it works for idle and high rpm no load. I think your injector might be "to big" to be operating in the controlled fuel zone at idle and low rpm and low Loads. And "to small" to deliver enough fuel in the short time available at high rpm and high loads on the road. Proper staged injection may be needed. Although working in the dribble zone then the controlled zone could be pseudo staged injection, interesting idea.


I made an experiment on the dyno with the original carb (keihin PWK 28) measuring the fuel quantity over a given duration at full throttle. According to my calculations, it uses 43cc/min at 13300rpm.
The injector is a 200cc/min at 3bar. So it's way too big for low rpms but it should flow way enough at full throttle/high rpm. I know the theory requires staged injection but today my experiments are showing the opposite.
Anyway I'm looking for injectors with a lower flow rate but it's hard to find Bosch ev14k-like injectors with lower flow rate.

[TZ350]

I know the theory requires staged injection but today my experiments are showing the opposite.


Your results are interesting.

Any of the common EFI CPU's all have a lot of complicated car orientated complexity and Tuning HMI interface. Your project has got me thinking, it does not have to be that complicated. Something very simple using an Arduino Nano or similar could do the job for racing two strokes.

The missing piece to the puzzle, is being able to see changes in airflow as the strength of the resonant activity in the pipe changes. The answer to that, is what I am looking for now. Currently I have to work with what I have got but when we know what to do. I think I will look at making a ECU setup like yours. I think a simpler approach than a Link or any of the other car based systems could be very effective for racing two strokes.

[Drew]

What's wrong with an airflow meter?

[TZ350]

What's wrong with an airflow meter?

The vain types tend to flap around until they break off then the engine happily swallows it. The hot wire type, the engine loves to spit on the hot wire and cool it off, just for fun and to confuse the ECU.

I think some of the big bikes with large air box's have been able to use MAF meters but on a single close to the engine they are not much use.

The fallacy and problems with trying to adapt four stroke air flow tech to a two stroke has been discussed on the Ecotrons and ESE thread in some depth.

[husaberg]

The vain types tend to flap around until they break off then the engine happily swallows it. The hot wire type, the engine loves to spit on the hot wire and cool it off, just for fun and to confuse the ECU.

I think some of the big bikes with large air box's have been able to use MAF meters but on a single close to the engine they are not much use.

The fallacy and problems with trying to adapt four stroke air flow tech to a two stroke has been discussed on the Ecotrons and ESE thread in some depth.
There is a newer hot wire like system that uses a plastic film protected on one side.


Hot Film

Hot-film MAF sensors function much like a hot wire sensor, and used a centrally-heated film or metallic grid-type element. One side of the film encounters cooling airflow, while the shielded backside maintains a consistent temperature, and the current differential between the two is measured and relayed as a square-wave digital frequency output, between around 30Hz at idle and 150Hz at wide-open throttle. Hot film sensors tend to be more robust and less susceptible to contamination than hot-wire types.


I think you could move the sensor away from the inlet to the inlet of A SEALED airbox
remember how much air it uses in a .25 of a second

B​UT IF YOU WERE SOPHISTICATED ENOUGH YOU COULD CREATE A SPEED DENSITY SYSTEM USING MUILTIPLE MAP SENSORS AND TACHO OUTPUTS PLUS TPS
sorry re caps

[TZ350]

B​UT IF YOU WERE SOPHISTICATED ENOUGH YOU COULD ........

I is not sophisticated enough .......

[husaberg]

I is not sophisticated enough .......

I was meaning the software
what i are refring to is a Speed density EFI system
https://www.hotrod.com/articles/electronic-fuel-injection-2/
GM did it in the 90's
http://support.moates.net/theory-speed-density/
http://www.ecutek.com/Products/Trade/ProECU-Tuning-Suites/Subaru-Tuning-Suites/Subaru_DIT/RaceROM/Speed-Density
https://www.onallcylinders.com/2018/04/27/efi-101-mass-air-vs-speed-density-best-ride/

[TZ350]

.
Speed Density is well known. My issue is to be able to read changes in airflow through the motor. Changes that happen as the strength of the resonant action in the pipe changes and therefor the strength of the suction action of the pipe drawing mixture up from the crankcase.

There are issues, like at idle the pressure in the crankcase is higher, near atmospheric. Than at full torque where the average crankcase pressure is a little lower.

And there are more issues, like at idle there is quite a high MAP reading of the difference between high and low crankcase pressures but when you open the throttle about half way the MAP value of the difference drops to about a quarter. Easy to understand when you realize the descending piston is chasing the air out of the open inlet instead of compressing it.

[husaberg]

.
Speed Density is well known. My issue is to be able to read changes in airflow through the motor. Changes that happen as the strength of the resonant action in the pipe changes and therefor the strength of the suction action of the pipe drawing mixture up from the crankcase.

There are issues, like at idle the pressure in the crankcase is higher, near atmospheric. Than at full torque where the average crankcase pressure is a little lower.

My thoughts are to map all the variables with a carb on (as we know that maps create and responds to the variables)measuring fuel flow plus all the different pressure readings then create a truth table of how the EFI needs to respond in these circumstances.
Basically what Riley Will did but create a far more in depth map of the pressures involved as well.
the beauty is once you have this map you will have a system that will be able to be readily adapted to any two stroke.
Those pressure changes are actually what is preventing the simple alpha N system working on a High output racing 2 stroke.
Other than maybe Honda i doubt anyone as mapped out all the pressures changes with a carb and a full array of pressure transducers to attempt to get it fully right.

[TZ350]

My thoughts are to map all the variables with a carb on (as we know that maps create and responds to the variables)measuring fuel flow plus all the different pressure readings then create a truth table.

Yes that is a good idea, do you or a friend have a 2S with a carb you could do this mapping with. I would be very interested in the data.

[husaberg]

Yes that is a good idea, do you or a friend have a 2S with a carb you could do this mapping with. I would be very interested in the data.

As I am operating with one arm plus a fractured back atm so unlikely i could even start the CR500, plus i am a complete Bastard so i have no friends anyway.
Plus i like carbs, i almost understand them:laugh:

[TZ350]

.
Sorry to hear you have been hurt.

Your data mapping looks like a good plan. As you cant help at the moment it will have to be plan "B" so it can wait a bit. The plan "A" is to get the bike running as well on the new Speeduino system as it did with the old Ecotrons CPU. Then see if blending AN and VE in the troublesome areas does anything worth while.

Other wise I will need to turn to you for that plan "B" data. Get busy and make some friends...... :laugh:

[husaberg]

.
Sorry to hear you have been hurt.

Your data mapping looks like a good plan. As you cant help at the moment it will have to be plan "B" so it can wait a bit. The plan "A" is to get the bike running as well on the new Speeduino system as it did with the old Ecotrons CPU. Then see if blending AN and VE in the troublesome areas does anything worth while.

Other wise I will need to turn to you for that plan "B" data. Get busy and make some friends...... :laugh:

I think you need to use a disc valve test mule as well as they would be different than a reed.
I think mapping all the pressure data is easily doable, but the fuel flow in real time would be bloody hard , maybe with a single circuit carb like a lake injector but i have no idea how to measure fuel flow in real time accurately is fast enough?
Why not approach link as far as i can see you would need a transducer pre and post disc plus on in the pipe?

[TZ350]

I think you need to use a disc valve test mule as well as they would be different than a reed. I think mapping all the pressure data is easily doable.

I have the disk valve test mule. When you have perfected the data collection system using your reed valve bike I would love to borrow the test gear and run it on my disk valver.

339739

[TZ350]

There is a newer hot wire like system that uses a plastic film protected on one side.

It sounds interesting. Do you have anymore info? ie what uses them and is there an inexpensive source. What cars should I be looking at in the wrecker yard.

[Drew]

Surely with the injection and a disk valve, fuel cant end up splashing back past the throttle body?

[husaberg]

Surely with the injection and a disk valve, fuel cant end up splashing back past the throttle body?
Google three times the mainjet

[TZ350]

Well that was an interesting afternoon. Tried smaller injectors to try and get the VE numbers up for better resolution but after a lot of work all I had to show for it was confirmation that I had the right ones to start with. Well, bummer, and sorry if I sent you barking up a tree, but it actually tells me more than that. You either (or all four) have greater power output, or poor latency, or actual Lambda richer than expected, or blowing more out the pipe than expected. The up-side is down the road, where more efficiency and/or power is potentially hiding; though it didn't need this exercise to find it. :? Where did you end-up for config numbers? David

It was probably worth the effort to confirm the injector sizing. Actual Lambda needs to be around 12.5:1 for a 2S. Maybe that is to cope with the residual exhaust gasses left in the cylinder. Gives an oxygen a better chance of finding a fuel molecule. For config numbers, I am not sure. I finished up chasing my tail for a bit and basically have to start again.

VE is a good concept for 4T's but misleading for a 2T. For 2T's it is all about "Transfer Ratio" and "Trapping Efficiency".

My bike may have a true VE of 80% (of 110cc) but the "Transfer Ratio" of air transferred to the cylinder is maybe 120% and the "Trapping Efficiency" of the air actually trapped in the cylinder after the exhaust port closes is maybe 66% of that 120%. So there is a wasted 34% of air that also gets fuel in it but is not burnt to produce power. The better I can time the injection so less of that 34% of wasted air gets fuel in it the cleaner the emissions and the better the fuel economy but it will never be perfect.

A Carburetor adds fuel to all the air that passes through it. Whether the air is being sucked in by the engine or blown back out. Either way it gets fuel added to it. Properly timed EFI can do better than that by avoiding adding fuel to air that is being blown back out of the crankcase. So EFI can effectively reduce pollution at the inlet and exhaust sides of the engine which would improve fuel efficiency too.

I think the bigger than expected injectors are needed to cover that 34% of wasted air that passes through the engine and the richer air/fuel requirements of a two stroke.

Picture of the bike on the dyno and a picture of the last dyno graph using the Ecotrons CPU. The plan is to get back to the same level then take a look at using the Speedy to find ways of seeing the changes in air flow through the motor that are induced by changes in the resonant action going on in the pipe. Ie, we want to be able to see the air flow changes when we throttle off and on again. The Ecotrons Alpha-N methodology could not do that but I am hoping that by adding a MAP value to the mix better drivability will be possible.

1st step is to get back to where I was before swapping CPU's.

.339740 339741

[TZ350]

.


Surely with the injection and a disk valve, fuel cant end up splashing back past the throttle body?Google three times the mainjet

<iframe width="560" height="315" src="https://www.youtube.com/embed/KcNfci9OruU" frameborder="0" allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen=""></iframe> ..... https://www.kiwibiker.co.nz/forums/showthread.php/86554-ESE-s-works-engine-tuner/page367

With a carburetor fuel is added whether the air is going in or coming out. 2S EFI "B" port injection can mostly avoid adding fuel to air that is being ejected from the crankcase.

[husaberg]

It sounds interesting. Do you have anymore info? ie what uses them and is there an inexpensive source. What cars should I be looking at in the wrecker yard.
It was new to me as well
Bosche do them pretty much exclusively now so likely mini VW and BMW

The Bosche system auto heats the hot wire to 1000 degees on shutdown or start up to burn of residue too.


https://www.youtube.com/watch?v=affYSqb5i5c
all these ones here with HFM
https://www.turnermotorsport.com/c-204-bmw-air-mass-sensors-hfm
at the bottom is the years and models

or Beemer
for E46 M3, E60 M5, E63 M6, MZ3-S54, MZ4

2001-2006 E46 BMW M3
2006-2010 E60 BMW M5
2006+ E63 BMW M6
2001-2002 Z3 BMW M Roadster M Coupe with S54 engine
2003-2008 Z4 BMW Z4 M Roadster Z4 M Coupe
$189.95 this one is compact

E39 M5, Z8

ie 1999-2003 E39 BMW M5
2000-2003 E52 BMW Z8 Roadster
$139 new

https://www.turnermotorsport.com/p-389-hfmmass-air-sensor-for-e46-m3-e60-m5-e63-m6-mz3-s54-mz4/

oh look the sensor itself is dirt cheap
$32.41
https://www.turnermotorsport.com/p-207718-13628645877-genuine-bmw-part/


Its also a mint DRZ125 engine.
Paging former Beemer witch doctor.....

unconfirmed but i think the 3.8 buick v6 commodores had a film maf sensor..........

The hot film MAFs, which AC Delco introduced in '84 on the Buick turbo V6 and have since used on the 2.8, 3.0 and 3.8L V6 engines, produce a square wave variable frequency output. The frequency range varies from 30 to 150 Hz, with 30 Hz being average for idle and 150 Hz for wide open throttle.
Nah buggar the holden one didnt use maf until the VS and even then i am not sure if its hot film them.


here is a pic of the output of the GM system used in the states
https://www.aa1car.com/library/maf_digital_waveform.gif

On the GM hot film MAFs, you can also tap into the onboard computer data stream with a scan tool to read the MAF sensor output in "grams per second" (GPS). The reading might go from 3 to 5 GPS at idle up to 100 to 240 GPS at wide open throttle and 5000 RPM.https://www.aa1car.com/library/maf_sensors.htm


on another subject i found this which i never knew about Map sensors



Waveform notes

The Manifold Absolute Pressure (MAP) sensor waveform seen in the above illustration is a typical output from an analogue sensor. With the motorcycle's engine at idle, the voltage shown on the example waveform is at 1.0 volts. The throttle is then snapped open and the voltage rises to just under 5.0 volts. The voltage achieved depends on how fast the throttle is opened, as a slower reaction time results in a lower voltage. While the engine is slowing back to idle with the throttle closed, the increased vacuum within the inlet manifold results in a lower voltage than that seen at idle. This lower voltage returns to the original voltage when the speed returns to idle.


https://www.picoauto.com/images/uploads/screenshots/_sml/moto_map.png
Technical information

MAP sensors can be either integral to the motorcycle's ECM or separate units. An ECM that is fitted with an internal MAP sensor can be identified by a vacuum pipe connection.
External MAP sensors are three wire devices which have:



a 5 volt supply voltage
an earth connection
a varying analogue output
a vacuum connection to the inlet manifold

The output from the sensor shows a rise and fall in voltage depending on the vacuum.
When the engine is stationary or the throttle is wide open, zero vacuum will be recorded and a voltage approaching 5 volts will be seen; as a vacuum is applied the voltage will reduce. The example waveform shows that at idle a voltage of around 1 volt is seen, and as the throttle is opened the vacuum in the manifold drops and a higher voltage for these conditions is seen. In this case the voltage rises to almost 5 volts.
The 'hash' on the waveform is due to the vacuum change from the induction pulses as the engine is running.

mabye this is the way to make it work with alpha N?
treated like a turbo car one map and Maf either side of throttle as to detect boost which would be vacuum in a two 2t you might need one on the pipe though as well.

[Drew]

My bike may have a true VE of 80% (of 110cc) but the "Transfer Ratio" of air transferred to the cylinder is maybe 120% and the "Trapping Efficiency" of the air actually trapped in the cylinder after the exhaust port closes is maybe 66% of that 120%. So there is a wasted 34% of air that also gets fuel in it but is not burnt to produce power. The better I can time the injection so less of that 34% of wasted air gets fuel in it the cleaner the emissions and the better the fuel economy but it will never be perfect.


.339740 339741
Best written description I have ever read on this subject, I understood all of it and could picture it without going backwards and forward a dozen times.

[husaberg]

if iis hot film
its cheap
https://www.ebay.com.au/itm/MAF-Air-Flow-Meter-Sensor-for-Holden-Commodore-VS-VU-VT-VX-VY-Statesman-V6-3-8L-/162637109813

[ceci]

with which we are neighbors, the government of my country in those years impeded the importation of this jewel

https://www.motobecane-passion.fr/injection_electronique.htm

[ceci]

with which we are neighbors, the government of my country in those years impeded the importation of this jewel

https://www.motobecane-passion.fr/injection_electronique.htm



It is good to look to the past to see everything that has been advanced

[husaberg]

Hi Rob I got bored last night and started reading ypur Speeduino thread on the Speeduino forum
I havent got to the end but i see most of the stuff i mentioned you have tried or was covered.
But i did find this that looks like its available as an add on to the Speeduino.
http://speeduino.com/forum/viewtopic.php?f=12&t=238&sid=d70b4d586962f21db46854d82b3b7a7c&start=10
Which is the deceleration fuel cut. If you havent tried that yet.
The other thing i found was that most cars have idle control valves also function on normal over run high revs no throttle open to bleed in air to a four stroke.

The idle air control valve works at idle and during deceleration. In other words, any time you take your foot off the pedal. During deceleration, the computers on most fuel injected vehicles do a “fuel cut” procedure where they stop operating the fuel injectors to force the engine to lose RPMs. However, since the pistons are still moving up and down, the engine still needs an air supply. The idle air control valve opens during deceleration to provide that air. Some people think that the idle air control valve fine tunes air/fuel mixture as you’re driving. It doesn’t. In fact, if you view a scan tool while driving, you’ll see that the PCM isn’t giving ANY commands to the idle air control valve. Take your foot off the pedal and you’ll see commands start up again during deceleration and fuel cut mode. Fly by wire systems do this directly with the throttle nowdays though.
As an aside MotoGP bikes actually use this feature to help control engine braking i think its called a "kicker" it works as afar as i know really well and simpler method than a over running sprag clutch at stopping the wheel locking from compression breaking.

[husaberg]

Another respose to deceleration Surge
http://www.powercdi.com/manual/combustion/pipe_bangs/en.html
Hes expressily fobids cut and paste.

[TZ350]

Another respose to deceleration Surge
http://www.powercdi.com/manual/combustion/pipe_bangs/en.html
Hes expressily fobids cut and paste.

Very very interesting article. A great description of four or six stroking combustion in a two stroke.

He is wrong about there being low pressure in the crankcase at low throttle openings. In reality at low throttle opening there is a low migration of fresh mixture through a partially contaminated crankcase in which the pressures are relativly high. In fact near atmospheric. Counter intuitively, at low throttle you have higher average crankcase pressure than you do at open throttle. You have the lowest average crankcase pressure at full open throttle.

Crankcase pressure represented by the black line is very nearly atmospheric and at much higher rpm, much higher than the 250's he is talking about it barely makes it down to an average of 0.9 bar. At 13,000 rpm like my GP it gets down to about 0.75 bar. That is not low pressure like we would associate with the inlet of a four stroke at low throttle.

339803

[TZ350]

But i did find this that looks like its available as an add on to the Speeduino.
http://speeduino.com/forum/viewtopic.php?f=12&t=238&sid=d70b4d586962f21db46854d82b3b7a7c&start=10
Which is the deceleration fuel cut. If you haven't tried that yet.

Thank you for finding these things for me.

I have tried DFC "Decal Fuel Cut" where the fuel is cut completely on over run. When you stand beside a DFC two stroke bike on a dyno that has just shutoff from a full noise 13,000 rpm pull and it seems to take forever to wind down you start to wonder about the wisdom of cutting off its lubrication. Fuel cut or no fuel cut the problem of 1 or 2 below remains the same when you throttle on again. Anyway it was not the answer to my drive ability problem.

My problem is that at any Throttle position and RPM combination there are two distinct fueling requirements that are entirely different from each other.

In both scenarios the "average" crankcase pressure is much the same.

1:- when the pipe is sucking air through the motor.

2:- when the pipe is "NOT" sucking air through the motor.

The problem is how to tell the difference between 1 and 2 and fuel accordingly.

I think I know the answer, the "average" is not much different. But the "difference" between the highest and lowest pressure each cycle is much less for 2 than it is for 1. And its this change in difference that I think will indicate relative changes in airflow through the motor.

I think this is the answer:-

339804

When there is a big difference between max/min crankcase pressure you have big air flow and when there is little difference you have little air flow.

You heard it here first.

[husaberg]

Very very interesting article. A great description of four or six stroking combustion in a two stroke.

He is wrong about there being low pressure in the crankcase at low throttle openings. In reality at low throttle opening there is a low migration of fresh mixture through a partially contaminated crankcase in which the pressures are relativly high. In fact near atmospheric. Counter intuitively, at low throttle you have higher average crankcase pressure than you do at open throttle. You have the lowest average crankcase pressure at full open throttle.

Crankcase pressure represented by the black line is very nearly atmospheric and at much higher rpm, much higher than the 250's he is talking about it barely makes it down to an average of 0.9 bar. At 13,000 rpm like my GP it gets down to about 0.75 bar. That is not low pressure like we would associate with the inlet of a four stroke at low throttle.

339803
I agree i think he wrote the original in Italian.Still interesting he claim to be able to tune most of it out with Ignition and i suspect pv as well.


Thank you for finding these things for me.

I have tried DFC "Decal Fuel Cut" where the fuel is cut completely on over run. When you stand beside a DFC two stroke bike on a dyno that has just shutoff from a full noise 13,000 rpm pull and it seems to take forever to wind down you start to wonder about the wisdom of cutting off its lubrication. Fuel cut or no fuel cut the problem of 1 or 2 below remains the same when you throttle on again. Anyway it was not the answer to my drive ability problem.

My problem is that at any Throttle position and RPM combination there are two distinct fueling requirements that are entirely different from each other.

In both scenarios the "average" crankcase pressure is much the same.

1:- when the pipe is sucking air through the motor.

2:- when the pipe is "NOT" sucking air through the motor.

The problem is how to tell the difference between 1 and 2 and fuel accordingly.

I think I know the answer, the "average" is not much different. But the "difference" between the highest and lowest pressure each cycle is much less for 2 than it is for 1. And its this change in difference that I think will indicate relative changes in airflow through the motor.

I think this is the answer:-

339804

When there is a big difference between max/min crankcase pressure you have big air flow and when there is little difference you have little air flow.

You heard it here first.
I think total fuel cut would be fast death as the carb still feeds air and fuel even with the throtle closed through the pilot and other circuits on return
I was musing more a 10-30%. I think it needs a Map which if set with the correct perimeters can tell the difference or a Maf which directly measure real flow.
Something similar is mentioned to your last bit in one of the threads i read last night.
I found some other stuff last night including some stuff from Nath88.
http://www.msextra.com/forums/viewtopic.php?t=65104

http://www.oocities.org/tech4tdi/pages/maf.html
http://www.msextra.com/forums/viewtopic.php?t=65104
https://www.sciencedirect.com/science/article/pii/S1876610217337578



GM MAF sensors output a frequency proportional to airflow as do Toyota MAF sensors. Ford, for certain and I'm pretty sure, Bosch sensors, output a simple voltage propertional to airflow.
.
You can mount the Maf or restrict the airflow and recalibrateor get the flow table if you cant find one on the net it using aflow bench
All of this makes you appreciate how clever a carb really is

[speedpro]

I've been measuring this AFR on overrun on my scooter lately. On overrun it goes super lean, over 20:1, and as the engine drops to idle the AFR returns to whatever you have it tuned for. This was easily repeatable simply blipping the throttle or slowing down on the road. As soon as the throttle was opened even a small amount the mixture returned to normal. It was almost like a Decel fuelcut on the fuel injected FZR which I've also played with. This is a 4T engine with a genuine Mikuni TM33 carb. A 2T will work exactly the same I would expect.

[husaberg]

I've been measuring this AFR on overrun on my scooter lately. On overrun it goes super lean, over 20:1, and as the engine drops to idle the AFR returns to whatever you have it tuned for. This was easily repeatable simply blipping the throttle or slowing down on the road. As soon as the throttle was opened even a small amount the mixture returned to normal. It was almost like a Decel fuelcut on the fuel injected FZR which I've also played with. This is a 4T engine with a genuine Mikuni TM33 carb. A 2T will work exactly the same I would expect.

Would some of that be why Foul strokes need accelerator pumps while two strokes strong intake pules only need shrouds on the Needle Jet?
I mentioned the Delco ECU a heck of a lot of the 100's millions that where spent on it were in the control of vacuum advance and lean off for cruising.
The Ecu are $60-100 now not really pertable enough for bikes but for their time pretty clever and there is all sorts of free software to tune them in real time.

Rob Might remember years ago i posted about my mates Mb100 that ran with a Pumper carb of a jet ski the carb was 38mm and it worked perfect other than on long over run
I think of these carbs as a primitive fuel injection
The carb was perfect but it wasnt able to tell the difference if the throttle was open or closed it just pumped fuel based on crankcase pressure alone.
So what happened was you would open the throttle after a long run in to a tight corner like a hairpin and it would load up for a second or two then come in with a bang once the excess fuel cleared.
Years later i found a story that jennings wrote where he had the same issue and it was able to be resoled with a simple check valve.
Karts dont suffer from it as they dont really shut the throttle like a bike does.

Oil and Maf sensors


Mass Air Flow Sensor Investigations
No dealership or service provider, when contacted, has ever been able to provide us with evidence to support this "myth," and in fact, our investigations have revealed that even authorized dealerships are simply speculating and do not have the test equipment necessary to know whether the sensor has failed or why. In the last 7 years, we have had more than 300 actual sensors sent to us by consumers with documents showing dealerships claimed our product had caused them to fail. Microscopic, electronic and chemical testing revealed that none of these sensors were contaminated by K&N oil (K&N Detailed MAF Sensor Test Results). What is perhaps the single biggest clue to what is going on is that over 50% of these sensors sent to us were not broken in the first place for any reason. Click here for more information on how this may happen.

“A lot of dealerships do not have the ability to test MAF sensors,” said K&N Vice President of Product Development and Engineering Steve Williams. “If you’re going to diagnose something you should show whether it is at fault or not. What they have is what we call the plug and play process. That is where they go through and download some codes and guess. The consumer then gets hit with a bill for components that were never a problem. K&N always tells people the facts.”
K&N Engineering Research and Development Department
K&N Engineering Research and Development Department
There is no silicone in K&N oil and testing has shown that MAF sensors will operate normally after being immersed in K&N Filter oil.
https://www.knfilters.com/MAF/massair.htm

https://youtu.be/HrU-sXkN4UY
https://youtu.be/Ml8OIrMBfKU (https://www.knfilters.com/MAF/massair.htm)
Even if it needs a squirt after every ride
https://www.crc.co.nz/Mass-Air-Flow-Sensor-Cleaner/6895-4e893416-ea9f-49c3-acee-0081ef6e9f4d/
https://www.crc.co.nz/ImagesCust/imagesdb/product/9e9473eb-1345-4280-b616-a8c5bc1849ac.jpg

[TZ350]

.
Trying to apply 4 stroke thinking to solving a 2 stroke issue is as silly as trying to understand a woman's needs by applying a male point of view, both internal combustion engines are entirely different species behaviorally.
.
.
.
1:- air flow when the pipe "IS" sucking air through the motor.

2:- air flow when the pipe is "NOT" sucking air through the motor.
.
.
.
1 & 2 are the issue and there is a whole range of meaning attached to "sucking" and that is the real problem to address.
.
.
The Ecotrons could not do drive-ability well on my bike with its Alpha-N only map. It is impossible for Alpha-N to do drive-ability well for engines with high performance pipes, because of the wide variance in air flow. So some sort of modified MAP or MAF value is needed. Currently I am exploring the possibilities of using the Speedy with its firmware that I can re write to suit my needs. I am looking at a modified MAP reading to adjust the Alpha-N table. MAF is plan "B".
.
.
What I need is a sure fire and affordable way to measure crankcase pressure with < 1ms settling time and a 2.5 or 3.5 bar absolute range.

[TZ350]

Another respose to deceleration Surge
http://www.powercdi.com/manual/combustion/pipe_bangs/en.html
Hes expressily fobids cut and paste.

This is interesting because I have seen it on my MAP gauge, random high MAP values as the engine is in over run or idling. Deceleration Surge looks like a good explanation for it.

It takes about six crank revolutions at WOT and full power to total refresh the crankcase with completely new mixture. The plan is when in over run or idling to smooth the MAP value over a number of revolutions.

[TZ350]

339806

This is a Alpha-N table. Each cell delivers a "FIXED" amount of fuel for that set RPM/TPS position.

For example the fixed amount of fuel delivered for 7.5% TPS at 9,000 RPM can not be both correct for when the pipe is sucking a lot of air through the motor and when its not.

Below 10/10 ... 10,000 RPM / 10 bar BMEP you might get away with it with some engines and a few people have. But above 10/10 the air flow changes over a much greater range than delivering just a fixed amount of fuel can cope with. Alpha-N works well on WOT because air flow is predictable but fails when the air flow becomes variable due to pipe resonance collapsing or not at lower throttle settings.

[TZ350]

.
There is quite a treasure trove in the links Husaberg has posted.

[husaberg]

.
Trying to apply 4 stroke thinking to solving a 2 stroke issue is as silly as trying to understand a woman's needs by applying a male point of view, both internal combustion engines are entirely different species behaviorally.
.
.
.
1:- air flow when the pipe "IS" sucking air through the motor.

2:- air flow when the pipe is "NOT" sucking air through the motor.
.
.
.
1 & 2 are the issue and there is a whole range of meaning attached to "sucking" and that is the real problem to address.
.
.
The Ecotrons could not do drive-ability well on my bike with its Alpha-N only map. It is impossible for Alpha-N to do drive-ability well for engines with high performance pipes, because of the wide variance in air flow. So some sort of modified MAP or MAF value is needed. Currently I am exploring the possibilities of using the Speedy with its firmware that I can re write to suit my needs. I am looking at a modified MAP reading to adjust the Alpha-N table. MAF is plan "B".
.
.
What I need is a sure fire and affordable way to measure crankcase pressure with < 1ms settling time and a 2.5 or 3.5 bar absolute range.

That tallies with these readings from what i posted of a industrial 115cc 2 stroke with no decent pipe.
339807
Also with KTM the map sensor feed is the bare spigot on the back of the cylinder just above the base
339808
They call it the Vacuum connection part.
https://www.bikebandit.com/oem-parts/detail/ktm/55430023000/b4596273?m=162555&sch=1067495


Here it is here on the oil pump
KTM Crankcase Presure Sensor

https://www.bikebandit.com/oem-parts/detail/ktm/55641075000/b4596314?m=162555&sch=1067500

The flywheel is saw toothed multi pickup just like yours too.

[TZ350]

339809

The changes in air flow are reflected by changes in the height between maximum and minimum crankcase pressure. This is where I am looking to get my MAP value from. It would be interesting to know if that is what KTM are doing.

[husaberg]

339809

The changes in air flow are reflected by changes in the height between maximum and minimum crankcase pressure. This is where I am looking to get my MAP value from. It would be interesting to know if that is what KTM are doing.

Well We could ask Fletner;)
I seen some very similar values mentioned elsewhere on a thread i read somewhere with a jetski or Snowmobile.
my gut says maybe it should go at least another 1/2 bar either way on the pipe, what does the modeling say on the sim?
With that output i posted from the GM sensor it can detect deceleration vs acceleration with the output and goes lower again with a shut throttle on over run than idle. no mater what sensor output either frequency or voltage is used (at least in a 4T situation)
The KTm ecu is about $450 US
The Sensor $40
So its really quite inexpensive for new tech by KTM standards
I will only take a mater of time before someone has a peek to see what the fuel map looks like and tells the world.
From memory at the time there was a lot of the talk about the oil pump which was Japanese having a lot of the patent tech
I can recall if its keihin or Mikuni but as the sensor is mounted to the oil pump the sensor might be also from them?
The oil pump output is ECU controlled according to the blurb.


1. What are the benefits of new TPI models?
• Automatic setting for altitude and temperature
• The 2 Stroke engine oil mixer,does not need premix fuel anymore, More precise engine performance, improved linear power delivery
• Reduced fuel consumption, There is no waste of fuel from the carburettor
• Simpler use, easier engine start, no change jetting required,The TPI range does not require the fuel and oil mixture


2. Are the Husqvarna TE250i & TE300i EFI (Electronic Fuel Injection) models differ from the model KTM EXC TPI 2018 ?
No, Husqvarna and KTM use the same technology, the same drive engines.
Husqvarna comes standard with a switch to change the injection map, Husqvarna also has a different concept of air filter (Airbox) with which the engine running and feeling remains specific to Husqvarna.

7. How much is fuel economy savings on injector models?
From the conditions and driving style, the savings are about 30%-40%.

8. What fuel / oil lubrication ratio is used?
The amount of 2t of oil constantly adjusts the computer’s ECU, depending on the engine rev , and the scale itself is constantly changing.
But the average fuel usage is ,oil is 80: 1

16. Is it possible to change the ignition map and engine power settings?
There is an alternative map that is named Map Switch and programmed in the ECU engine. This other map is something of lighter features (softer engine characteristics) It is not possible to create your own ignition and injection maps.


If you watch this it also has a simple diagnostic port.

https://www.youtube.com/watch?v=db_HBWmC830

but wait
from ktm

ow using accepted and understood parts it all came down to processing the information and getting the mixtures right. Quickly. The window of time for the injector to be open and getting the large amount of fuel a two-stroke needs is the challenge, especially in even smaller displacements we are told. Plus when a two-stroke has a “flame-out” inside the combustion chamber, it can take up to 50 strokes for the engine to recover. This was a big issue in high RPM, low throttle times. For instance when you are going down a hill with the engine revving and then you give it a little gas. Making the throttle response as good as that antiquated carburetor was very difficult. There is not a lot of time for the computer to decide how much fuel and when to shoot it from one stroke to the next since the conditions inside the two-stroke engine can change drastically and very suddenly. Even small things like having the engine’s crankcase pressure sensor and fuel injectors mounted so that vibration was not affecting them were lessons learned in testing. Luckily the new 2017 engine was designed from the beginning to eventually have a non-carbureted fuel system so it would vibrate less and also be able to produce as much electrical power as would be needed.
Read more at https://dirtbiketest.com/fresh-dirt/two-stroke-fuel-injection-ktms-path-to-tpi/#mTs9BF1q5sgXMoYl.99


In the end the answer, for now, was actually quite simple. In partnership with Synerjet the ECU was developed and tuned and that has been the difficult part in getting a base system to work off of. Now more tuning and testing can take place to see where FI will go on two-strokes.
Read more at https://dirtbiketest.com/fresh-dirt/two-stroke-fuel-injection-ktms-path-to-tpi/#mTs9BF1q5sgXMoYl.99


of synerjets site
http://w.synerject.com/images/vehicule/fiche/fra-hose/308/300/img_8659.png
only 2t one i can find is this
http://w.synerject.com/en/2-3-wheeler/scooters/lpdi-system-p9-5
http://w.synerject.com/images/vehicule/fiche/tmap-30/308/300/img_8435.pngcombined temp and presure sensor 4mm direct fitting
http://w.synerject.com/en/2-3-wheeler/scooters/tmap-30-p23-5
One more piece i just found the KTM has a lambda in the pipe
339815

[ceci]

339809

The changes in air flow are reflected by changes in the height between maximum and minimum crankcase pressure. This is where I am looking to get my MAP value from. It would be interesting to know if that is what KTM are doing.

I love this graphic and what you have indicated in it.
In the one it is appreciated well where the pressure begins and it is not in the TDC (because this occurs: at that point the inertia of the incoming gaseous mass is greater, than the speed of the piston creating empty) (in the 4S that is the delay of closing of intake valves "a few degrees after the BDC").
I know I'm wrong, but I still think that taking the pressure of the crankcase through a hole located below the exhaust port and controlled by the skirt of the piston, is a simpler method that will allow us to know how much is the quantity aspirate

[TZ350]

I know I'm wrong, but I still think that taking the pressure of the crankcase through a hole located below the exhaust port and controlled by the skirt of the piston, is a simpler method that will allow us to know how much is the quantity aspirate

Maybe wrong, or maybe not. It is a good idea for a timed pressure reading and worth looking at. I will find the series of EngMod crankcase pressure simulations I ran and have a look.

339818 5% 339817 15% 339816 65% Throttle Position

At 9,000 rpm they look much the same at 0.8 bar and at 12,000 rpm they range from 0.7 to 0.85 bar. So might be able to work as an indicator of air flow.

For myself I still think the variance between maximum and minimum pressure each cycle gives more definition and is the approach I am going to try first.

[TZ350]

One more piece i just found the KTM has a lambda in the pipe 339815

That is interesting, KTM have located the O2 sensor in the header, the ideal place for a four stroke. But conventional two stroke wisdom has it in the rear cone or stinger for a more effective reading and less risk of oil contamination. In a two strokes header it is possibly going to see a diet of raw oxygen spilled from the cylinder before the exhaust port closes and then hot combustion products. A confusing mixture that can appear lean when its not.

As the response time for an O2 sensor is quite long compared to the cycle time at higher rpm the signal must become confused about what its looking at. The tail pipe position gives a better average reading and also responds quickly enough. I think KTM has fallen into the trap of trying to apply four stroke thinking to a two stroke. I believe KTM copied Flettners "B" port injection concept but he did not use an O2 sensor. So who were they going to copy for that, probably 4 strokes of course.

[husaberg]

That is interesting, KTM have located the O2 sensor in the header, the ideal place for a four stroke. But conventional two stroke wisdom has it in the rear cone or stinger for a more effective reading and less risk of oil contamination. In two strokes header it is possibly going to see a diet of raw oxygen spilled from the cylinder before the exhaust port closes and then hot combustion products. A confusing mixture. As the response time for an O2 sensor is quite long compared to the cycle time at higher rpm the signal must become confused about what its looking at. The tail pipe position gives a better average reading and also responds quickly enough. I think KTM may have fallen into the trap of trying to apply four stroke thinking to a two stroke.

Not only is it in a 4t postiion its the same sensor used in the 4ts and bloody expensive with it
It would be $679NZD based on the AUS price

<tbody>
Year
Model
Fiche Section


2017
250 EXC-F (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/250-exc-f-eu/exhaust-system#n_75041090200)


2017
350 EXC-F (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/350-exc-f-eu/exhaust-system#n_75041090200)


2017
350 EXC-F SIX DAYS (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/350-exc-f-six-days-eu/exhaust-system#n_75041090200)


2017
450 EXC-F (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/450-exc-f-eu/exhaust-system#n_75041090200)


2017
450 EXC-F SIX DAYS (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/450-exc-f-six-days-eu/exhaust-system#n_75041090200)


2017
500 EXC-F (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/500-exc-f-eu/exhaust-system#n_75041090200)


2017
500 EXC-F SIX DAYS (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/500-exc-f-six-days-eu/exhaust-system#n_75041090200)


2018
300 EXC Six Days TPI (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2018/300-exc-six-days-tpi-eu/exhaust-system#n_75041090200)


2018
300 EXC TPI (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2018/300-exc-tpi-eu/exhaust-system#n_75041090200)


2017
1090 ADVENTURE R (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/1090-adventure-r-eu/exhaust-system#n_75041090200)


2017
1290 SUPER ADV. S BLACK 17
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/1290-super-adv-s-black-17/exhaust-system#n_75041090200)

</tbody>

<tbody>
Year
Model
Fiche Section


2017
1290 SUPER ADV. S BLACK 17 (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/1290-super-adv-s-black-17-eu/exhaust-system#n_75041090200)


2017
1290 SUPER ADVENTURE R (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/1290-super-adventure-r-eu/exhaust-system#n_75041090200)


2017
1290 SUPERDUKE R BLACK 17
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/1290-superduke-r-black-17/exhaust-system#n_75041090200)


2017
1290 SUPERDUKE R BLACK 17 (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/1290-superduke-r-black-17-eu/exhaust-system#n_75041090200)


2017
1290 Super Adventure R TKC
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/1290-super-adventure-r-tkc/exhaust-system#n_75041090200)


2018
250 EXC TPI (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2018/250-exc-tpi-eu/exhaust-system#n_75041090200)


2017
1290 SUPER ADV. S ORANGE 17 (CN)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/1290-super-adv-s-orange-17-cn/exhaust-system#n_75041090200)


2017
1290 SUPERDUKE R WHITE 17
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2017/1290-superduke-r-white-17/exhaust-system#n_75041090200)


2019
250 EXC TPI (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2019/250-exc-tpi-eu/exhaust-system#n_75041090200)


2019
300 EXC TPI (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2019/300-exc-tpi-eu/exhaust-system#n_75041090200)


2019
300 EXC Six Days TPI (EU)
Exhaust System (https://www.ktmonlineparts.com.au/partFinder/fiche/ktm/2019/300-exc-six-days-tpi-eu/exhaust-system#n_75041090200)

</tbody>

[TZ350]

339819

For a two stroke this is a better position in the rear cone for a O2 sensor than KTM's position in the header. I am not sure but Wobbly may even put his O2 sensor in the Stinger itself.

[husaberg]

339819

For a two stroke this is a better position in the rear cone for a O2 sensor than KTM's position in the header. I am not sure but Wobbly may even put his O2 sensor in the Stinger itself.

I found two rules of thumbs on a net troll

i suspect ktm is only using it as a trim for euro emissions possibly only idle and and using averages for basic trm rather than a constant data input.
the main fueling will be based of a table and Map sensor


Bell uses these figures for mixtures

Starting 1:1-3
Idling 1:8-10
Low speed running 1:10-13
Light load ordinary running 1:14-16
Heavy load 1:12-14

The numbers for a four stroke are as follows: (again, just recommendations for baseline tuning, exact numbers also depend on other factors)

idle 14-15:1
light load/low rpm 14-16:1
light load/ordinary running 14-15:1
heavy load 12-13.5:1

On a four stroke, you can rely 100% on the afr for tuning. On a 2 stroke, the way the exhaust (https://thumpertalk.com/link/click/4478/) works, the air in the expansion chamber can be very turbulant at certain rpm and throttle ranges. When you use an AFR in a 2 stroke you will work off the averages rather than the exact. It is a great tool to keep an eye on where you are at, it is however 1 tool out of a whole arsenal that you can use to get the bike tune dialed in.

I trust the gauge 100%, what I cannot trust is that with all the air moving and resonating around in the chamber, the readings may not be as accurate as in the 4 stroke engine. The wideband in the 2 stroke will likely not read very accurately near idle and in low throttle situations, so care must be taken to ensure proper running in those areas. You should also verify that the afr's you are seeing are corresponding with the plug readings and cylinder wash readings to make certain the bike is running as it should.It's a matter of where the sensor is placed in the exhaust stream.

Two rules of thumb i found where

The best spot on a two-stroke is from about 90% down the length of the header to the first 5% of the divergent cone. "Off-the-pipe" readings may be affected by A/F blow through, but anywhere within the usable power range, there should be very little chance of outside air being anywhere near the sensor at that point.

To install you will need a 12v dc power source. The sensor must be heated before use, so the power source has to be able to run the wideband before the bike is running. The sensor should be places in the largest part of the expansion chamber, on the upper 3rd of the exhaust. The placement needs to be up high to keep from collecting oil.



I prefer the second as the rest of what he said makes so much sense all the suff in the first paragraphs

[TZ350]

.
O2 sensor is often associated with EFI but Wobbly uses it for dialing in carb's too.

Target Air/Fuel for a WOT power pull is 12.5:1 (lamba 0.85).

Off the pipe I usually see 17:1 (lamba 1.2) because of the excess short circuiting, in the cylinder combustion is probably closer to a proper 13 - 14:1.

[husaberg]

.
O2 sensor is often associated with EFI but Wobbly uses it for dialing in carb's too.

Target Air/Fuel for a WOT power pull is 12.5:1 (lamba 0.85).

Off the pipe I usually see 17:1 (lamba 1.2) because of the excess short circuiting, in the cylinder combustion is probably closer to a proper 13 - 14:1.
But wait these more

The KTM 250 XC-W TPI system starts with the powerful and proven counterbalanced 250cc two-stroke engine. After years of two-stroke FI development, KTM found good performance, reliability, and (significant in Europe) the ability to meet Euro 4 emission standards with the patented “Transfer Port Injection” (TPI). The system starts with a 39mm Dell’Orto throttle body equipped with an idle adjusting screw and a separate cold-start device that opens a bypass to allow in extra air. Oil is supplied from the automatic oil pump and injected through the throttle body with the incoming air. The Dell’Orto throttle body does not house a fuel injector, as KTM found mounting two fuel injectors downstream into the rear transfer ports provides excellent atomization of the fuel with upstream air. The result is highly efficient combustion and reduced losses of unburned fuel.
Knowing when and how much fuel to inject is all controlled by KTM’s engine management system (EMS). This system uses a newly developed electronic control unit along with five sensors that determine not only how much fuel but also how much oil to inject. You will recognize these sensors as: throttle position, coolant temperature, intake air temperature, intake air pressure, and ambient air pressure. Thanks to these sensors and the EMS, there is no longer the need to manually adjust the jetting for temperature or altitude conditions.




Extra air for choke has to be a typo two pressure sensors it must be a double one on the oil pump two ports one as reference

[TZ350]

... two pressure sensors it must be a double one on the oil pump two ports one as reference

"Ambient" .... to keep things in balance as you climb a mountain or the day clouds over and a cold front rolls in.

"Crankcase" .... so as to see changes in air flow.

That is my guess.

I think a pressure sensor in the inlet tract itself would be pointless as there is not the sort of pressure variance there like you see in four strokes.

[husaberg]

"Ambient" .... to keep things in balance as you climb a mountain or the day clouds over and a cold front rolls in.

"Crankcase" .... so as to see changes in air flow.

That is my guess.

I think a pressure sensor in the inlet tract itself would be pointless as there is not the sort of pressure variance there like you see in four strokes.

I wonder if all Map sensors see actually ambient pressure anyway as wouldn't they need it for a reference. inlet tract or Crankcase wouldn't they see the exact same pressure anyway other than a Disc Valve cylinder or if it was pre reed. sampling,Looking at the pics the air supply for the Map sensor is at the rear transfer position
Edit i found this

ANALOG MAP SENSORS
The MAP sensor consists of two chambers separated by a flexible diaphragm. One chamber is the "reference air" (which may be sealed or vented to the outside air), and the other is the vacuum chamber which is connected to the intake manifold on the engine by a rubber hose or direct connection. The MAP sensor may be mounted on the firewall, inner fender or intake manifold.
A pressure sensitive electronic circuit inside the MAP sensor monitors the movement of the diaphragm and generates a voltage signal that changes in proportion to pressure. This produces an analog voltage signal that typically ranges from 1 to 5 volts.
Analog MAP sensors have a three-wire connector: ground, a 5-volt reference signal from the computer and the return signal. The output voltage usually increases when the throttle is opened and vacuum drops. A MAP sensor that reads 1 or 2 volts at idle may read 4.5 volts to 5 volts at wide open throttle. Output generally changes about 0.7 to 1.0 volts for every 5 inches Hg of change in vacuum.

When the ignition key is first turned on, the powertrain control module (PCM) looks at the MAP sensor reading before the engine starts to determine the atmospheric (barometric) pressure. So in effect, the MAP sensor can serve double duty as a BARO sensor. The PCM then uses this information to adjust the air/fuel mixture to compensate for changes in air pressure due to elevation and/or weather. Some vehicles use a separate "baro" sensor for this purpose, while others use a combination sensor that measures both called a BMAP sensor.

On turbocharged and supercharged engines, the situation is a little more complicated because under boost there may actually be positive pressure in the intake manifold. But the MAP sensor doesn't care because it just monitors the absolute pressure inside the intake manifold.

A good MAP sensor should read barometric air pressure when the key is turned on before the engine starts. This value can be read on a scan tool and should be compared to the actual barometric pressure reading to see if they match. Your local weather channel or website should be able to tell you the current barometric pressure reading

MAP SENSOR TESTS
A MAP sensor can also be bench tested by applying vacuum to the vacuum port with a hand vacuum pump. With 5 volts to the reference wire, the output voltage of an analog MAP sensor should drop, and on a Ford digital MAP sensor the frequency should increase.
An analog MAP sensor's voltage can also be read directly using a voltmeter or oscilloscope. A digital MAP sensor's frequency signal can be read with a DVOM if it has a frequency function, or an oscilloscope. The leads would be connected to the signal wire and ground.
Warning: Do NOT use an ordinary voltmeter to check a Ford BP/MAP sensor because doing so can damage the electronics inside the sensor. This type of sensor can only be diagnosed with a DVOM that displays frequency, or a scope or scan tool.
Another way to check out a Ford digital MAP sensor circuit is to input a "simulated" MAP sensor signal with a tester that can generate an adjustable frequency signal. Changing the frequency of the simulated signal should trick the computer into changing the fuel mixture (look for a change in the injector pulse width signal).
No change would indicate a possible computer problem.

The Ambient pressure sensor in the KTM can get confused if you change altitude quickly with the bike not running IE trucked on ute.
To re-calibrate it needs to be stopped and started again.
Its the only foible reported i have seen.


I find the manual choke interesting though.It seems odd that given the bike has a temp input and a ECU it would need it.

KTM have reported the TPI is only really a stop gap measure rather than a total solution as Euro 5 comes in soon and euro 6 is twice as tough again.
They just cant get DI to work cheaply and as well HP wise.

[ceci]

339818 5% 339817 15% 339816 65% Throttle Position

At 9,000 rpm they look much the same at 0.8 bar and at 12,000 rpm they range from 0.7 to 0.85 bar. So might be able to work as an indicator of air flow.

For myself I still think the variance between maximum and minimum pressure each cycle gives more definition and is the approach I am going to try first.



What happens in the first graph 5%: the incoming gaseous mass barely has inertia because it is a slow and smooth displacement, in this case the piston creates vacuum faster, so the lowest pressure point is in the TDC where the piston stops creating vacuum.
In the graph 65%: it is the reverse what happens and the lowest point of pressure is advanced to the TDC

[TZ350]

.
Yes that makes sense.

[ceci]

When you let go of the accelerator: you cut the passage of the gaseous masses (creating turbulences that annul the inertia of this), but you do not reduce the inertia of the crankshaft and the linear speed of the piston.
How it influences the minimum pressure point: delaying it (which is anomalous since the rpm places it advanced)

[husaberg]

Just a thought dirty old 80's Turbo mods
You could actually feed two imputs to the map sensor, Crankcase and Pipe output
If you use correctly orientated check valves and a bleed and tees and Needle valves set up.
You could contol the outputs from the Map to reflect when the engine is on the pipe or not.
You could do this on the Fly on the Dyno or Track to vary the switching points to the two separate maps
The Map can take multiple imputs as Turbo Maps do just this.
Also using this method almost any Map sensor will do.
You could do this with computer inputs but i feel this would be the quick and dirty for a proof on concept.
It would be interesting to see what the out put signals would look like

[TZ350]

339844

Starting to make some sensible progress. Interesting, the huge differences in AF ratio.

Peak torque and therefor peak trapping efficiency is about 11,500 rpm and shows the richest mixture because there is the least short circuiting of air out the exhaust at that point.

At 9.8 its still a little rich there though, should be more like 12.5:1. Have to be careful about the lean areas as some of that can be un-trapped air/fuel escaping out the exhaust port before its closed.

I can see where the acceleration enrichment needs increasing. I am very happy with progress.

When I have got back to the same level of tune as I had with the Ecotrons system I will start modifying the software code for Speedys firmware so as to get the style of MAP input I want.
.

[ceci]

When you let go of the accelerator: you cut the passage of the gaseous masses (creating turbulences that annul the inertia of this), but you do not reduce the inertia of the crankshaft and the linear speed of the piston.
How it influences the minimum pressure point: delaying it (which is anomalous since the rpm places it advanced)




In addition to this, the pressure symmetry between the 150º and 210º points, which we see in the 5% and 15% graphs, reappears.

[ceci]

.
Yes that makes sense.


and this makes sense?

[TZ350]

.
That is informative. Thanks.

[ceci]

.
That is informative. Thanks.

Thanks to you and I hope it can be useful

[ceci]

I find it interesting
https://ac.els-cdn.com/S1876610217337578/1-s2.0-S1876610217337578-main.pdf?_tid=f2ec7ca3-dc28-4ab8-9fec-34bc2826a7b2&acdnat=1545303888_32aab4a5cf11b31434c98cd1376f3d10

[TZ350]

I find it interesting https://ac.els-cdn.com/S1876610217337578/1-s2.0-S1876610217337578-main.pdf?_tid=f2ec7ca3-dc28-4ab8-9fec-34bc2826a7b2&acdnat=1545303888_32aab4a5cf11b31434c98cd1376f3d10

Very interesting read.

Sadly they conclude in the first paragraph of page 7, that "Beyond 7750 rpm, air and fuel do not mix properly because of the short time available for the thermodynamic cycle."

339936339937

[TZ350]

339938 this is where I got to with the Ecotrons 2S EFI system.

People have asked what am I doing and I find it easier to explain that I had written a story using a foreign language (Ecotrons) and to change the ending of the story I need to use a different language ( Speeduino). But to change the ending I first have to re write the story in the new language so it reads the same as it did in the old one. When I have done that then I can change the ending to a more interesting one.

[TZ350]

.
I spent a very happy peaceful Christmas day in the shed. Slowly working through the typical issues you get with a new project.

I have traced the heart stopping engine flame outs to a faulty 12V power plug and have now soldered the 12V supply directly to the Arduino Mega board. And another issue causing flame outs and the motor stopping dead was the Mega board becoming detached from the Speedy board. On over run the vibration is sever enough to shake the board loose. Given how tight they are on the pins it was a bit of a surprise.

Having fixed that, I am now on the hunt for why the TPS signal appears so noisy on over run. The new analog smoothing option in Tuning Studio that comes with the latest Speedy firmware upgrade should be useful for sorting out if it is a mechanical or electrical issue.

339964
.

[speedpro]

Noisy TPS on overrun is probably the same cause as the Mega board becoming detached. I wonder if the TPS was as noisy with the Ecotrons system. I'd be looking at loose connections, or maybe the position sensor simply has a hard time with excess vibration due to the spring loade wiper being bounced about.

[TZ350]

Noisy TPS on overrun is probably the same cause as the Mega board becoming detached.

There might be something in that.

I had a look back at earlier traces before I secured the Mega board and there does not seem to be the same TPS noise. How that can be, I have no idea. It is hard to imagine anything on the board itself that could be affected by vibration.

Or maybe it is coincidental and like you say, the issue is something has started bouncing around in the TPS itself too now.

[TZ350]

.
The Steps for down loading and installing the Speeduino EFI software:-

From The Main Speeduino Home Screen.

Step 1. Download Firmware.
Step 2 Download Latest Stable Firmware.
Step 3. Unpack Zip File.
Step 4. Speeduino-INI setup file for Tuning Studio is in the Reference Directory.
Step 5. Speeduino-INI file to be loaded into Tuning Studio. This file is the Bridge between Tuning Studio and Speeduino.
Step 6. Speeduino Firmware code is in the Speeduino Directory.
Step 7. Compile and Upload Speeduino Firmware to the Arduino Mega board.
.

[TZ350]

.
Step 6. Speeduino Firmware code is in the Speeduino Directory.
Step 7. Compile and Upload Speeduino Firmware to the Arduino Mega board.
.

[TZ350]

.
I have finally got back to a point where the engine is running well enough for me to start looking at the "Two Stroke MAP" concept again. I am only working in the low throttle low RPM area but it is probably the most demanding if you are looking for differences in air flow between firing and misfiring.

Because that is the region where a two stroke can be four, six, eight or more stroking. Ie misfiring but not from poor spark but bad air fuel ratio in the cylinder caused by random mixing of the old burnt exhaust gas that has found its way back to the crankcase with the small amount of fresh air coming in at low throttle openings.

At low speed/throttle the crankcase has a high proportion of exhaust gas in it and it is not always certain that what fresh mixture that there is at low speed finds its way to the cylinder each cycle.

The average crankcase pressure at idle and low speed/rpm is close to atmospheric because of back flow from the exhaust. Idle air/fuel mixture is the small amount of fresh air that makes its way through this fog. A small amount of fresh air in equals the same volume but random ratio of dirty plus fresh air/fuel mixture out and up to the cylinder.

This randomness of how much incoming air/fuel actually makes it to the cylinder is another reason a four stroke style MAF sensor on the inlet side makes no sense for a two stroke. You would actually have to measure the flow and oxygen content going up the transfers to have a meaningful MAF style measurement.

At low demand it is rather random the ratio of dirty exhaust gas in the crankcase to combustible air/fuel mixture that finds its way to the cylinder on the next cycle. This randomness is the reason for the corn popping ring ding two stroke sound.


Another respose to deceleration Surge

http://www.powercdi.com/manual/combustion/pipe_bangs/en.html

He expressily fobids cut and paste.

Husaberg posted a very interesting link.
.

[TZ350]

.
Well I got to play tonight with the 2S-Map software and coded up a simple Max - Min and smoothed pseudo MAP value. I scraped an exponential smoothing logarithm off the net .... yn = w × xn + (1 – w) × yn – 1 ... no array and just one easy line of code.

The engine was running sort of reasonably on Alpha-N and a fixed MAP input the other day. But with the variable MAP value the motor has become troublesome on the same tune. Coincidental or somehow affected by the new MAP values, I am not sure.

Anyway the MAP does change from about 0.3bar to 1.40bar, but not very smoothly. As the motor is not running that great I can't tell how useful the pseudo MAP results are going to be.

Still it is great to have something to work with and I will know more in a few days hopefully. With any luck I will be able to get the motor running better and be able to log some data. That should tell me if I am on the right track.

It is such hard work making "reality" conform to "theory" ........... :lol: ..... https://youtu.be/Ml7dmxb8ghw

<iframe width="560" height="315" src="https://www.youtube.com/embed/Ml7dmxb8ghw" frameborder="0" allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen=""></iframe>
.

[Nath88]

Looks great, what does the MAP input do in the code? In Microsquirt there is a MAPxRPM table, the value in the table is a multiplier for the value in the TPSxRPM table. Or it can switch from one TPSxRPM table to another based on a threshold MAP value.

I presume you're taking the MIN and MAX per cycle in the speeduino code.
At what crank position does the cycle start (MIN and MAX values reset)?
How often does the ADC read the MAP input?
How often is MAX-MIN calculated?
Can you log the MIN and MAX separately?
Can the speeduino do fast logging at the ADC speed?

Why are you smoothing? I'd be expecting MAX-MIN to jump from extreme to extreme depending on whether there's fire or not.
Is it clear cut enough that you can switch tables based on some threshold value, say 1 bar? Can you have an output power an LED when the table is switched, so you can see the LED flashing in real time with the pipe dings?

Sorry about the questions, love your work though.

[speedpro]

You don't work for KTM by any chance?

[TZ350]

Looks great, what does the MAP input do in the code? In Microsquirt there is a MAPxRPM table, the value in the table is a multiplier for the value in the TPSxRPM table. Or it can switch from one TPSxRPM table to another based on a threshold MAP value.

Hi Nath88.

Speeduino has two standard tables, VE and Alpha-n both can be multiplied by MAP and as the firmware code is open source you could have any scenario you like if you are prepared to re-program the firmware. Tuning Studio (Mega Squirt) is the application for Tuning the Speeduino firmware and Tuning Studio also handles all the data logging. Currently the software continuously looks for a maximum and minimum each cycle and works out the difference and translates that into a MAP value. The next move is to instantaneously measure the minimum at 10deg ATDC and Maximum at 160 deg ATDC only, which are the correct two places to look according to EngMod2T simulations.

All the 2S MAP Max Min stuff is add on code that I have done on my copy of Speeduino. The next move is to polish it up so its usable.

[husaberg]

You don't work for KTM by any chance?

I think thats Austria not Australia. Similar spelling but far worse smelling
As far as i know Flets still their go to guy for Southern Hemisphere R&D

[EfiCanuck]

TZ - thanks for doing all this work. I am a few steps behind with my 2 stroke EFI conversion, and this will help a lot.

Here are a few data points. Dont think they will help much, but are interesting.

- The OSSA trails bike - a 280 Keihin 2 stroke EFI - uses a map sensor in the air box. Very smooth and controllable bottom end. Perhaps since the TB is small (28mm) they can get enough of a reading to make it useful. Or perhaps they are reading it at specific crank angles like you are.

- The Vertigo trial bike. Uses a Athena GET ECU. Very smooth and controllable bottom end. They have an integrated IAT and MAP sensor on the airbox as well. No idea how they sample the sensor.

- The KTM 250/300 TPI MAP sensor is attached to the crankcase.

- I have a GET ECU on my KTM 250 TPI dirt bike and had a little chat with an Athena tech about how the crankcase pressure sensor is used. He indicated that the sensor on the crankcase is only used between 20%-30% TPS and 3.000-4.250 RPM. It was also implied that there was a "special" algorithm. Perhaps similar to yours - selecting values at specific crank angles. Of course this is for a enduro dirt bike, a street bike or race bike would be different.

[TZ350]

.


it´s my two stroke bike with speeduino https://youtu.be/zk55HR6bgO0



Great fuel injection project, I like it..... :drinknsin .... looks like more and more people are trying their hand at 2S EFI.

[husaberg]

.



Great fuel injection project, I like it..... :drinknsin .... looks like more and more people are trying their hand at 2S EFI.

He likely does even remember posting that or even joining KB;)
looks more like what we would call a Yamaha RS125.
I have seen the Rd125LC singles but never an aircooled one.

EDIT sorry its an RD135

Made mostly for the Brasilian market
https://upload.wikimedia.org/wikipedia/commons/b/b3/Yamaha_RD_135Z.jpg
looks like the cases are actually designed for a water cooled top end. looks like 5 speed box

Asia had the RZX135.
https://www.bikebd.com/wp-content/uploads/2016/12/yamaha-rzx-135-specification.jpg
Yamaha RX-Z 135 was a two-stroke naked bike manufactured by Yamaha Motor Corporation. Debuted in April 1985, the RX-Z it was very popular in Malaysia and Singapore and has been sold for more than two decades before the production was ended in 2011. [2] In 2004, the model was revised cosmetically and a catalytic converter was installed.
The Yamaha RX-Z was originally equipped with a five-speed transmission with solid front disc brake. The model shared the same platform and chassis as its predecessor, the water-cooled Yamaha RD125. Shortly after the launch, the RX-Z became popular among young motorcyclists especially in Malaysia. A few years later, the engine was upgraded with the installation of a six-speed transmission, together with newer instrument panel and handlebar switches, as well as a cross-drilled front disc brake rotor.

Since then, the design of the RX-Z remained unchanged until 2004 when the model was updated, with the rear lights borrowed from the Yamaha Y125Z. The new RX-Z was equipped with a catalytic converter, sacrificing 1 PS of maximum power (original maximum power: 21 PS). However, the maximum torque remained unchanged but the low-end torque was improved compared with the early models. Some owners of the earlier RX-Z motorcycles may have problems during take-off because the engine tends to stall when an inexperienced rider tries to take off in the first gear. However, the problem was resolved in the new model.

In Malaysia, the RX-Z was one of the motorcycle models often associated with Mat Rempit street racers. [3][4] As a result, many Malay movies including Remp-It use the RX-Z in their movies.
The production of the RX-Z ended in 2011 after being sold for more than two decades
https://en.wikipedia.org/wiki/Yamaha_RX-Z

So what are are is a RD125LC with an air cooled top end and a 5 speedbox that was later changed to 6 again.


Mat Rempit
is a Malaysian term for "an individual who participates in immoral activities and public disturbance with a motorcycle as their main transport", usually involving underbone motorcycles,[1] colloquially known as Kapcai, or scooters. Mat Rempits are not involved in street racing but rather go against each other for cheap thrills and to rebel against authorities while some of them perform stunts for fun, such as the wheelie, superman (lying flat on the seat), wikang, and scorpion (standing on the seat with one leg during a wheelie).
The word "Rempit" comes from "ramp-(rev)-it" (ramp the throttle). According to Kamus Dewan, the definition of "Rempit" is "the act of whipping with a cane".[2] An alternative source of the word is derived from the noise made by a 2-stroke motorcycle. "Mat" is a Malay slang term used to address or to refer to a male person who is usually of Malay descent but is generally used derogatorily.
Organization and traits

Mat Rempits usually travel in groups and create disturbance in bustling city centres on weekend nights. Mat Rempits are usually associated with gangsterism, gang robbery, street fighting, assault vandalism, theft and bullying. Most motorcycles used by the Mat Rempits do not meet standard specifications, or have been modified extensively for greater speed, or just to make the exhaust noisier.
Additionally, most Mat Rempits do not have valid motorcycle licenses, do not pay road taxes, and ride stolen motorcycles.A growing number of housing estates have also been turned into racing tracks.[8] It is estimated that there are about 200,000 Mat Rempits in Malaysia

https://en.wikipedia.org/wiki/Mat_Rempit

Marlon Brando without the Triumph but 200,000 of them.

[TZ350]

I haven’t successfully injected a two stroke yet but I have done an EFI for Mazda rotary which has many of the same exhaust tuning sensitivities as a two stroke (big intake and exhaust port timing overlap).

I think you are on the right track going with speed density. Alpha N was a PITA to get tuned right, almost impossible. I used a MAP sensor fed by an accumulator in turn connected to manifold with small gauge hose or with a small orifice restriction.

This formed a filter to mechanically average out the rapidly fluctuating manifold pressure. Only down side to this approach is throttle response may be less than instant. That was not a problem in my application (airplane) where all throttle changes are relatively slow.

I have tried the accumulator idea and find that with a 2S all you get is an average pressure. And the interesting thing about that is that at idle the average is near ambient and at WOT the average is a little lower.

I have written the code for a Arduino Nano to take a high pressure measurement at 135 deg ATDC and a low pressure reading at 315 deg ATDC. The code is triggered by a Timer Interrupt and an interrupt attached to the ignition trigger calculates the period. The end result after a bit of math manipulation is a MAP value that looks like a traditional 4S MAP. I am in the process of de bugging the code and hope to get to try it all out this weekend.

Small amplifier board on the left to amplify the ignition pulse simulated by my computer. Arduino Nano in the middle on a break out board with screw terminals. 12 bit DAC on the right near the calculator.

.340779

[OopsClunkThud]

I'm sure you know this but keeping the interrupt code as small and quick as possible is important. Just enough to capture the data. Do everything else in the normal loop.

[Vannik]

I have been following what you are doing here and on the Speeduino forum and are beginning to think you need to sample at a much higher rate, like two times per degree at max rpm and have a very fast processor to do some signal conditioning before using it. We often find the signal to noise ratio with pressure signals very bad where the noise has a 3 or more times bigger amplitude than the pressure amplitude. But not my area of expertice!

[TZ350]

I'm sure you know this but keeping the interrupt code as small and quick as possible is important. Just enough to capture the data. Do everything else in the normal loop.

I am just finding my way here and appreciated the reminder about the code.

[TZ350]

.... beginning to think you need to sample at a much higher rate, like two times per degree at max rpm and have a very fast processor to do some signal conditioning before using it. We often find the signal to noise ratio with pressure signals very bad where the noise has a 3 or more times bigger amplitude than the pressure amplitude.

340783

As I am not looking to track a complete cycle now but only looking to have two pressure measurement values taken at two distinct fixed points I was thinking of only two samples but if I can manage it I will make a series of readings of the MAP sensor clustered around each point and smooth them. Thanks for the pointer about noise.

I am using a CTC Timer Interrupt, being a newbe to coding, the challenge will be to reset the counter quickly enough.

It was a simple idea. Take a high and low crankcase pressure reading and use the difference between them as an indication of airflow. Easy to say but it sure has been harder to implement.
.

[Vannik]

I understand what you are trying to do but would have started with a full high performance super fast system and then worked at simplifying (less expensive) it rather than trying from the less complex side, adding more and more complexity and maybe missing something. But all in all a very worthwhile and noteworthy effort.

[TZ350]

I understand what you are trying to do but would have started with a full high performance super fast system ...

Problem is, I have come from a place of complete computing/programming ignorance and would not know a high performance super fast system if it hit me on the head........ :laugh: ... I have to work with the resources I can and in the end it may not be enough.

It has been suggested to me that a mechanical system might work for me better. Something that opens a port to a small chamber at the right time for the MAP sensor to read the crankcase pressure. Easy to do with a hole in the skirt of the piston and a gallery through the cylinder to a small chamber. The chamber could hold sufficient volume/pressure to allow the MAP sensor to settle. It is a tempting idea.

I am getting the code sorted, but it is like polishing a Turd, just takes time.

340803

[husaberg]

Problem is, I have come from a place of complete computing/programming ignorance and would not know a high performance super fast system if it hit me on the head........ :laugh: ... I have to work with the resources I can and in the end it may not be enough.

It has been suggested to me that a mechanical system might work for me better. Something that opens a port to a small chamber at the right time for the MAP sensor to read the crankcase pressure. Easy to do with a hole in the skirt of the piston and a gallery through the cylinder to a small chamber. The chamber could hold sufficient volume/pressure to allow the MAP sensor to settle. It is a tempting idea.

I am getting the code sorted, but it is like polishing a Turd, just takes time.

340803

I came across this Rob
https://www.megunolink.com/articles/coding/3-methods-filter-noisy-arduino-measurements/
but you know i like the remote chamber KISS reading

[TZ350]

I came across this Rob
https://www.megunolink.com/articles/coding/3-methods-filter-noisy-arduino-measurements/
but you know i like the remote chamber KISS reading

Interesting link Husa, thanks.

I have been using a traditional exponential smoothing function. Exponential uses very little memory and processing time unlike other methods that require arrays. The remote chamber thing looks like a good fallback position if I really have to go there. But the Arduino Nano will be "KISS" once it is sorted as copies could easily be added to a Link, Microsquirt, Megasquirt, or any other EFI system that takes a MAP input.

[TZ350]

.
Making progress with the Arduino Nano MAP co processor.

340825

The code runs one event interrupt for sync of a single lobe trigger 20deg BTDC and timer interrupt 1 with a 64 prescaler. Timer interrupt 1 triggers twice a crank revolution, once at 135deg ATDC and again at 315deg ATDC. So the compare match register changes twice a crank revolution and is slightly asymmetrical.

340824

It all works in theory on the test bench but I am a little concerned that 560 microseconds at 13,000 rpm will not be enough time for timer 1's code to make two low pressure analog reads and do a bit of math before the sync interrupt gets triggered 45deg later.

Anyway, it's time to tidy the code up, annotate it properly and remove the de bugging traces and see what happens when its run in real life on the engine.

340823

The trace in the box on the right shows the numbers are in the right place at a simulated 13,000 RPM and the numbers are the right size.
.

[TZ350]

.
The Arduino Nano co processor on a screw terminal breakout board and a 12 bit DAC for realistic analog output that is usable as a MAP signal for the EFI's CPU.

Hopefully get to try this out on the bike in the next night or two. If it looks promising I will post the code, if not its back to the drawing board.

.340849 340850

[TZ350]

.
Arduino Nano MAP co processor setup ready to test. Hopefully tonight.


340872 340873

[Michael Moore]

In case it might be of help, 55641075000 is the part number for the KTM TPI MAP/crankcase pressure sensor, US$39.31, (listed with the oil pump instead of the throttle body/ECU where I expected to find it). It seems like it should be some standard bought-in component. It may be worth your getting one to see if KTM specified something out of the usual range of sensitivity compared to MAP sensors sold for 4T use. Or if it matches up to what you are using you'll know you probably aren't too far off on your setup.

There's a KTM/Husky TPI thread at advrider and the sensor seems to often get replaced when people's bikes are having problems running correctly.

cheers,
Michael

[husaberg]

In case it might be of help, 55641075000 is the part number for the KTM TPI MAP/crankcase pressure sensor, US$39.31, (listed with the oil pump instead of the throttle body/ECU where I expected to find it). It seems like it should be some standard bought-in component. It may be worth your getting one to see if KTM specified something out of the usual range of sensitivity compared to MAP sensors sold for 4T use. Or if it matches up to what you are using you'll know you probably aren't too far off on your setup.

There's a KTM/Husky TPI thread at advrider and the sensor seems to often get replaced when people's bikes are having problems running correctly.

cheers,
Michael

https://www.kiwibiker.co.nz/forums/showthread.php/185773-Speeduino-2T-EFI-Project?p=1131118311#post1131118311
Check out the O2 sensor price and location a few posts down
https://www.kiwibiker.co.nz/forums/showthread.php/185773-Speeduino-2T-EFI-Project?p=1131118379#post1131118379

[TZ350]

.

Well my first try of the pseudo MAP co processing simulator did not work so well on the bike. So I have spent two weeks developing a high/low pressure simulator to mimic the fluctuating two stroke crankcase pressures and then tested the pseudo MAP co processing unit and it was a real failure as far as timing and staying in sinc with the ignition trigger. So no surprise it did not work so well.

341103 Arduino Nano crankcase pressure simulator on the right and the pseudo MAP co processing unit on the left.

With the help of the Arduino Nano crankcase pressure simulator I made/programmed and a scope I finally got the pseudo MAP co processor sorted. The co processor is triggered of the ignition pulse and stays in sink now with the crankcase pressure changes from 500 RPM to 14,000 RPM and can clearly see both peaks and troughs. The scope display is at 12,000 RPM.

Hopefully I now have a pseudo MAP co processing unit that can sense the changes in airflow through the two strokes crankcase and interpret the changing differences between the high and low points as changes in airflow and send that as a 4T like MAP value to the EFI's CPU.

Hopefully get to try it on the dyno tomorrow night after work.

[patouille]

Hey,
I didn't give any news for a while but I was still investigating my troubles.
I replaced my hall sensor with a pwm generator to see why I was unable to make it run over 10K RPM. It turns out the injector starts messing up at ~11K, I tried a different one (CBR 125) with the same troubles at the same RPM.
Did you have the same issue ?
For me the next step is to modify the throttle body in order to add a second injector. I think I will trigger one injector after the other.

[Drew]

At some point I think ya'll are gonna have to accept that the best method is direct injection to the combustion chamber, and let it oil the way they do normally.

[TZ350]

I was unable to make it run over 10K RPM. It turns out the injector starts messing up at ~11K, I tried a different one (CBR 125) with the same troubles at the same RPM.
Did you have the same issue ?

Sounds about right. An injector small enough to give good tun-ability at lower rpm will be to small to deliver enough fuel at 11k rpm. It is time to get the job done that is the problem.

And no.... I have always used two logical injectors a small and a big in staged injection. They swap over about 7-8k rpm. Both big and small deliver the same amount of fuel (more or less) per crank cycle. It is just the big one has much less time to do it so has to really shovel it in. But then of course because it has to be open for a minimum amount of time it is to big for low rpm.

The Ecotrons injectors I have been using so far have been good to 14,000 rpm firing on each crank cycle (233 Hz).

[TZ350]

At some point I think ya'll are gonna have to accept that the best method is direct injection to the combustion chamber, and let it oil the way they do normally.

Why? and that would also limit rpm to something less than 9k because of the time it takes to evaporate the injected fuel and stir it properly into a combustion-able air/fuel mixture.

[Drew]

Why? and that would also limit rpm to something less than 9k because of the time it takes to evaporate the injected fuel and stir it properly into a combustion-able air/fuel mixture.

Pointed slightly outward from the center, the squish will take care of stirring it with the fuel.

[TZ350]

Pointed slightly outward from the center, the squish will take care of stirring it with the fuel.

No, that whole approach has proven to be dud for high rpm performance engines, just not enough time to get things done. The whole thing is RPM (read "time") limited.

[TZ350]

<iframe width="560" height="315" src="https://www.youtube.com/embed/ZKNxrs7E-4s" frameborder="0" allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen=""></iframe>

Video of the Pseudo MAP concept:

Some smoothing required and the bike is just idling, I need another pair of hands to video a dyno run, but for now things look promising.

[TZ350]

.
Success ....... https://youtu.be/xAp77XQx9R8


I have finally got the pseudo 2S EFI MAP co processor working well with the output smoothed. Now I have something useful to work with to explore the possibilities of using MAP and a VE like table with fuel injection for a two stroke 2S. The co processor is measuring the high and low crankcase pressures and converting the difference between them to a pseudo MAP value.
.

<iframe width="560" height="315" src="https://www.youtube.com/embed/xAp77XQx9R8" frameborder="0" allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>

[Vannik]

This is just so great!

I was wondering in your calculation of the MAP values (I think it was (Phigh-Plow)*3) whether you should not maybe add an atmospheric based Patm to your equation? Maybe you are planning it but for now it is not an issue.

[TZ350]

I was wondering in your calculation of the MAP values (I think it was (Phigh-Plow)*3) whether you should not maybe add an atmospheric based Patm to your equation? Maybe you are planning it but for now it is not an issue.

Yes, thank you for your encouragement and I think you are right, the equation needs developing. But for now, I am happy just getting the (Phigh-Plow)*3 actually giving a meaningful result. A lot to do yet.

The great news is that the Pseudo MAP value behaves like I hoped. It increases as the motor starts making power and drops in areas where the TPS increased but the motor did not fire. You can see that in the middle of the picture where the TPS (green line) increases but the RPM (white line) decreases because the motor did not pick up and the MAP (yellow line) and rpm dropped. To the right of the graph you can see where the motor picks up and the MAP follows suit. Fueling is crap so the traces are not very neat.

341171

Altogether it looks like the concept is going to work.

[Flettner]

Thankyou, I'll take one.

[TZ350]

.

Ok. Another bump in the road. The Pesudo MAP thing works pretty well BUT .... there seems to be another issue that comes up when the throttle is closed ie TPS = 0%.

With the pseudo crankcase pressure MAP (Phigh - Plow)*3 I get a range of 0.2 bar, TPS = just open and motor firing to 0.9 bar, TPS fully open and engine firing. The motor responds to changes in MAP value just like you would expect.

But with the throttle fully closed ie TPS = 0% the pseudo MAP value is 0.6 bar when it should be 0.2 at most. I think this is due to the atmospheric pressure at the stinger/muffler tip overcoming weak suction by the pipe and pushing exhaust gas back into the crankcase. And without the suction action of the pipe the residual average pressure in the crankcase is higher than it would be if the throttle was a little open and the motor firing well enough to get some suction action going in the pipe.

The 0.6 bar is the result of (Phigh - Plow)*3 working on contaminated gases filling the crankcase whereas as soon as the motor fires properly the pipe sucks on the crankcase and the pseudo MAP value drops back to 0.2 bar.

The EFI CPU can't work with 0.6 bar for two entirely different situations.

The fix might be to bring the TPS into the mix and when it is = 0% then force the pseudo map to be 0.2 bar. Easy enough and I hope to try it soon. There is another interesting project on the dyno at the moment so it might be a week or so before I can get back to the 2S EFI project.

[TerraRoot]

I've nothing to add, but my thanks for documenting all of your work, warts and all.

actually i do have something to add, i'm not really surprised that TPS data would be needed!

[Vannik]

I am no expert on this but would expect both throttle position and later also rate of throttle change to be part of the control loop.

[husaberg]

I am no expert on this but would expect both throttle position and later also rate of throttle change to be part of the control loop.

I think thats the key, rapidly shutting off a throttle and rapidly opening a throttle have very different fuel requirements.
For instance diaphragm Pumper carbs as used on Jet skis and Karts keep pumping due to crankcase pressure, even when the throttle is closed. This results in over run over fueling. (the way to avoid this is not shut off completely which is not practical on a bike)
a conventional carb needs the airflow a open carb sees when the throttle is open. Consequently this doesn't happen.
Even with the map sensor i am not sure if the difference will be enough without, a throttle closing or opening rate of change comparison.
I dont have the brains to figure out how to do this. But you would have to set up a table showing greater or lessor over a time frame, maybe over a few corners to see what is the norm.
Maybe it just needs the two sensors airflow and map. Or a additional maps only for overun or acceleration much like a carb how it has a different circuit for acceleration (ie accelerator pump)and lean cruise circuit.
All i know GM spends 100's of millions doing stuff like lean cruise and overrun down hills etc.
I am only spit-balling this, as I hope it isnt, as Rob is doing some very neat stuff and sharing it openly, i certainly dont want to sound like its criticizing him or the approach.

[Vannik]

It it wasn't for the impressive and neat things Rob is doing we would not be able to have this discussion.

The automotive engineering world use terminology that if you are unfamiliar with it you can miss important publications, maybe it is well known but was strange to me:

Throttle opening: Tipp-in
Throttle closing: Tipp-out

[husaberg]

It it wasn't for the impressive and neat things Rob is doing we would not be able to have this discussion.

The automotive engineering world use terminology that if you are unfamiliar with it you can miss important publications, maybe it is well known but was strange to me:

Throttle opening: Tipp-in
Throttle closing: Tipp-out
Well i had never seen it mentioned.
Intersting stuff tipp-in

the period it checks throttle delta is probably about 10hz, and I've found it is hard to get it to activate more than about 2 times per second. If it fires off once, it seems like there is a delay period before it is allowed to fire off again.

I think there is probably a watchdog thread that watches the output value of the map. The map is checked as a normal thread on a periodic schedule, constantly updating the value. If the value is not zero when the watchdog is run, it triggers a special routine to buzz the injectors the appropriate amount, then starts a countdown timer so it cannot be triggered again for 1/2 second or so. This is just from experience logging and listening to the buzzing. I haven't bothered trying to look at it in detail in disassembly because I don't feel there is a need.
http://www.romraider.com/forum/viewtopic.php?t=1735

[TZ350]

I am no expert on this but would expect both throttle position and later also rate of throttle change to be part of the control loop.


I think that is the key, rapidly shutting off a throttle and rapidly opening a throttle have very different fuel requirements.

Yes I think so too. Pretty much any EFI CPU is for a four stroke and they typically have two provisions for rate of change of throttle position. One for rate of change of opening (accelerator pump) and another much cruder one for rate of change of closing for over run lean out, usually to reduce wet fuel and bore wash.

I have a TPS and the rate of change of throttle position maps are already built into the four stroke EFI CPU I am using and I can modify their characteristics easily. These maps would be normal for any typical four stroke EFI CPU and all off this works Ok with my two stroke.

A typical simple four stroke EFI CPU expects to see:-

RPM
Crank Position.
Cam Position (but not always essential).
Exhaust O2. (good to have, but not all that essential either).
IAT (inlet air temperature).
CLT (coolant temperature).
TPS (position and rate of change).
MAF (manifold absolute air flow)
or
MAP (manifold absolute air pressure). MAP is what my four stroke EFI CPU is looking at to judge how much air flow is passing through the motor.

In a 4 stroke, MAP is at a minimum when the throttle is closed and at its highest when the throttle is fully open. MAP is a good indication of air flow and has a huge influence on the amount of fuel delivered each cycle.

I am trying to run my two stroke with a simple four stroke EFI CPU that expects to see MAP.

And my problem is to measure the fluctuating two stroke crankcase pressures that occur each cycle and present that as a pseudo MAP value to the four stroke EFI CPU in a way that the CPU understands. My fluctuating two stroke crankcase pressure has to look like a steady four stroke MAP value. That is where my pseudo MAP co processor made from an Arduino Nano comes in.

With pseudo MAP = (Phigh - Plow)*3 I am almost there. It is the high crankcase pressure at closed throttle that is a problem and I think you are right, I will have to also use TPS here too. But just to tell the pseudo MAP co processor that the throttle is closed and instead of measuring the crankcase pressure the pseudo MAP value should just be at its lowest.

341354

Speeduino four stroke EFI CPU (top) with Arduino Nano co processor (bottom) for the pseudo MAP value and vari pot to simulate TPS for the EFI's CPU. The LED's are there to replicate the Injectors and so I can see what is happening.

[TZ350]

Throttle opening: Tipp-in
Throttle closing: Tipp-out

I like it. Very easy to visualize.




Intersting stuff tipp-in http://www.romraider.com/forum/viewtopic.php?t=1735



Interesting and very familiar.

[crbbt]

Loving the project!

are you having any problems with the jumper leads losing signal because of the vibrations?

[TZ350]

Loving the project! are you having any problems with the jumper leads losing signal because of the vibrations?

No not really but it is all very temporary and when the basics are sorted everything will get a tidy up. Working on the CVT sidecar at the moment, hope to get back to the 2S EFI project maybe next week.

[ceci]

More reading for EFI addicts two stroke enginehttp://suw.biblos.pk.edu.pl/resourceDetailsRPK&rId=4079

[TZ350]

341784

I have been fighting a bit of a battle with the PesudoMAP sensors software thinking there was something wrong with it but after extensive testing with the simulator and oscilloscope I am very confident with it. The issue is that the PesudoMAP value drops away over 8,000 RPM or so as seen in the picture above (Yellow line).

Currently my code can make a sensible reading every 8 deg at 13,000 RPM and with a bit of clever coding it could make a reading every 2 deg. 8 deg is good enough for the three high pressure readings I make at 15 deg intervals. 115, 130, 145 deg ATDC. If the software code is good then the issue must be with the crankcase pressure sensor.

341785

So I put the oscilloscope on the crankcase pressure sensor tonight, purple line, yellow line is the ignition pulse. I wanted to see what sort of signal it produced and what it looked like compared to EnginMod's prediction. Pretty close I recon. The next move is to try and capture a trace at 12,000 rpm.

I have been having problems with the MAP value dropping away above 8,000 RPM and suspect either the pressure sensor is not keeping up or the speed of response from it is so slow that the indicated pressure peak moves away from my last measuring point at 145 deg ATDC.

341786

[Grumph]

What Husa said about GM reminded me of a point Rob might appreciate. They cheat on throttle closing. Most of the auto EFI systems have a damper on the throttle linkage which regulates the rate of closure over the last few degrees of throttle plate movement.
When that area between working throttle and reducing to idle is a constant in terms of rate of change, it's much easier to map.

[TZ350]

TeeZee , the comment about ignoring the absolute value , just measure the delta between high and low values is a valid one.

Yes, it is my intention to only measure the delta between high and low, then multiply it by a simple correction factor to increase its magnitude to something similar to what a regular four stroke MAP value would be.

That way I can use a regular common after market four stroke type EFI CPU unit and traditional Alpha-N and VE volumetric efficiency mapping. It has taken a bit of work to figure out how to do this, not quite there yet but I feel I am closing in on it.

341952

On my two stroke GP/NSR110. 3 times the Delta crankcase pressure roughly equals the MAP reading you would expect to see from a four stroke's inlet manifold.

341951

This is an actual scope reading at about 11,500 rpm of a MAP sensor reading the crankcase pressure of my GP/NSR110. The yellow spikes are the ignition trigger pulses.

The purple steps are the sensors output each milli second. The sensor outputs its last (averaged??) reading every milli second for a milli second, hence the steps.

It is easy to see that the low pressure is around TDC and the high pressure at BDC as seen by this pressure sensor. I think processing delay shifts the timing from reality a bit. But that is not a problem as it is consistent.

My mission is to capture the readings with my Arduino Nano, process them and then pass a pseudo MAP reading to the EFI's CPU.

The Arduino Nano can easily take the two readings required every cycle, in fact it takes three high and three low and picks the best. Then averages them over ten readings, finds the delta and multiplies it by 3 and finally outputs it as an analog pseudo MAP reading. I have successfully simulated the fluctuating crankcase pressure and tested the pseudo MAP software on the scope to 15,000 rpm.

341953

Theoretically it is technically simple to do but it is just taking me a bit of time to achieve it in working practicable reality on my bike.

KTM


A small tube in the back of the cylinder is connected with an intake pressure sensor, which supplies pressure data to the control unit. .... https://www.ktm.com/en/enduro/150-exc-tpi/engine-exhaust


I think KTM have got it sorted and are doing something similar on their latest EFI 2S bikes.


I think so, you might find KTM and TM are spying here. KTM were certainly aware of my work documented on TSM.

[Carel H]

341951

This is an actual scope reading at about 11,500 rpm of a MAP sensor reading the crankcase pressure of my GP/NSR110. The yellow spikes are the ignition trigger pulses.




The scope shows a cycle time of nearly 6ms, just over 10000 rpm. It looks like the scope's calculation slipped over the chopped blocks of the sensor.

With some fantasy you can see a pseudo cycle of 5 ms (and the disadvantages of the slow sample cycle).

You cant choose when it takes a sample, so you could make a voting system over a longer period. Or walk away from automotive parts in general.

[TZ350]

The scope shows a cycle time of nearly 6ms, just over 10000 rpm.

The trigger was on the yellow channel and set to 4,75V rising edge of the initial trigger pulse and I was looking at the f=193 Hz displayed top right. But, yes you are right, time scale is 2ms a division so 6ms a cycle or 167 Hz. ?????? Why would the scope show two different frequencies??? more shit I don't understand ...... :scratch:


You cant choose when it takes a sample, so you could make a voting system over a longer period.

At the moment my pseudo MAP system takes the best high and low reading of each cycle and rolling averages them over ten cycles. I would welcome ideas around how I could do this better.

[TZ350]

.

.... walk away from automotive parts in general.

I was hoping not to have to and at the price of a KTM spare sensor I don't think KTM are using anything too fancy themselves.

Automotive sensors can be picked up cheap on the Net.

The Speeduino EFI CPU's firmware code is free and open source so you can modify it to suit your needs.... https://speeduino.com/wiki/index.php/Compiling_and_Installing_Firmware#Installation_-_Easy_Method

TunerStudio for adjusting the Speeduino's mapping, the basic version is free.... http://www.tunerstudio.com/index.php/tuner-studio

The Speeduino Shield DIY naked board for the Arduino Mega is cheap and they can also be brought fully made up.... https://speeduino.com/forum/app.php/page/buy

The Arduino Mega itself is relativly inexpensive.... https://www.amazon.com/ARDUINO-MEGA-2560-REV3-A000067/dp/B0046AMGW0

The Arduino Nano is cheap .... https://www.amazon.com/OSOYOO-Arduino-ATMEGA328P-Microcontroller-Without/dp/B00UACD13Q

So Inexpensive, Cheap or Free. As a DIY project I hope others might try, I am keen to keep this affordable.

[Carel H]

The trigger was on the yellow channel and set to 4,75V rising edge of the initial trigger pulse and I was looking at the f=193 Hz displayed top right. But, yes you are right, time scale is 2ms a division so 6ms a cycle or 167 Hz. ?????? Why would the scope show two different frequencies??? more shit I don't understand ...... :scratch:



At the moment my pseudo MAP system takes the best high and low reading of each cycle and rolling averages them over ten cycles. I would welcome ideas around how I could do this better.

I saw on the bottom left screen "CH1" and I assumed the blue line was CH1.

Doesn't matter, for it applies to both: a scope will very happily give you a nice frequency reading on clean wave signals, but with complex signals the going gets tough.

Looking at the signal, I would synchronize with the ingnition signal, take a few samples during 1.5 - 2 ms, take the lowest and put it in the moving average filter.

[Carel H]

.


I was hoping not to have to and at the price of a KTM spare sensor I don't think KTM are using anything too fancy themselves.

Automotive sensors can be picked up cheap on the Net.

The Speeduino EFI CPU's firmware code is free and open source so you can modify it to suit your needs.... https://speeduino.com/wiki/index.php/Compiling_and_Installing_Firmware#Installation_-_Easy_Method

TunerStudio for adjusting the Speeduino's mapping, the basic version is free.... http://www.tunerstudio.com/index.php/tuner-studio

The Speeduino Shield DIY naked board for the Arduino Mega is cheap and they can also be brought fully made up.... https://speeduino.com/forum/app.php/page/buy

The Arduino Mega itself is relativly inexpensive.... https://www.amazon.com/ARDUINO-MEGA-2560-REV3-A000067/dp/B0046AMGW0

The Arduino Nano is cheap .... https://www.amazon.com/OSOYOO-Arduino-ATMEGA328P-Microcontroller-Without/dp/B00UACD13Q

So Inexpensive, Cheap or Free. As a DIY project I hope others might try, I am keen to keep this affordable.


The sensor you use now is meant for a multi cilinder 4 stroke engine and will give for that purpose a nice signal. It is also calibrated and will therefore give no tuning demands on the assembly line or replacement.

No need for fast peak to peak measurements. The signal you want is on the chip, not on the thermal compensated and calibrated output.

That's automotive.

As for cheap and fast, not calibrated, not compensated (check the box "Freescale"):

https://www.reichelt.nl/Sensors-SMD-wired-/2/index.html?ACTION=2&GROUPID=8097&START=0&OFFSET=16&SID=94XEIjaawQATIAAGA4k1Id0a736d0b45e774ca58425b11 e7f8497&LANGUAGE=EN

I had a quick look at the Speeduino and stopped at the marketing talk of Tunerstudio. Same as the KTM site: hardly at specification level, but all about the experience.

Also intended for multicilinder four strokes, which makes parametric tuning harder and will eventually make it necessary to change the software.

For your purpose an Arduino plus two Mosfet injector drivers would be a nice (cheap) start.

I looked at EFI systems before. I find it a narrow bridge between programmers and mechanics. The programmers don't understand an engine, the mechanics don't understand electronics and proud of it on both sides.

Resulting in these stacked rough database contraptions, far removed from engine management.

Preferably in a heavily finned, colour anodised, billet machined housing.

End of rant.

[TZ350]

For your purpose an Arduino plus two Mosfet injector drivers would be a nice (cheap) start.


That could easily be the future.


As for cheap and fast, not calibrated, not compensated (check the box "Freescale"):

https://www.reichelt.nl/Sensors-SMD-wired-/2/index.html?ACTION=2&GROUPID=8097&START=0&OFFSET=16&SID=94XEIjaawQATIAAGA4k1Id0a736d0b45e774ca58425b11 e7f8497&LANGUAGE=EN


I am grateful for your input. This sensor looks like the unit that came on my Speeduino board.

Why would it have so many wires? Do you think it would be worth getting a spare and trying it?

341966

[TZ350]

.

Looking at the signal, I would synchronize with the ignition signal, take a few samples during 1.5 - 2 ms, take the lowest and put it in the moving average filter.

Pretty much what I am doing:-

The ignition signal initiates an interrupt.
From that the time for a complete cycle is determined.
Then the time in micro seconds are calculated to the six sample points. Three low around TDC and three high around BDC.
At the appropriate time the analog reads are done.
The best of the lowest and highest are chosen.
Then they go into their respective FILO 10 cell liner arrays and then the 10 current entry's are averaged.
The rolling average values are used in the Delta pressure calculation. Ie; High - Low = Delta
The Delta is rolling averaged too then multiplied by 3 to give a pseudo MAP value.
The pseudo MAP value is continuously output and updated each cycle.

341971Air Cooled and Fuel Injected Suzuki GP125

A series of EFI dyno runs taken at different throttle positions on my old air cooled 125. I had spent a lot of time mapping the Alpha-N fuel injection table. Alpha-N only looks at TPS and RPM and for each cell on the table you map an engine "Load" value that is used by the CPU to determine how much fuel to inject at that particular point, every time, nothing changes. This worked well for anything above 45% throttle position but it is not a flexible fueling system. More suited to drag racing than road racing.

I got the Alpha-N mapping fairly good and even rode the bike like that a few times and towards the end, sort of successfully but throttle control coming out of corners was not so sweet.

The current plan is to use Alpha-N above 45% TP and VE, volumetric efficiency below 45%. VE looks at engine volumetric efficiency vis manifold absolute pressure (MAP). That way fueling can be variable in response to the airflow changes induced by the pipe.

This way I can have the best of both worlds. Alpha-N which has proved reliable and very tune-able when the motor is on song and responsive variable fueling at lower throttle openings.

Combining the two mapping typologies, Alpha-N and VE should be relativly easy with the Arduino Nano pseudo MAP system. Below 45% I make the output variable and above I make it fixed at atmospheric.

[Carel H]

That could easily be the future.



I am grateful for your input. This sensor looks like the unit that came on my Speeduino board.

Why would it have so many wires? Do you think it would be worth getting a spare and trying it?

341966

It's the "DeLuxe" version with build-in amplifier to make life easy, three pins are No Connect:


https://www.nxp.com/docs/en/data-sheet/MPX4250A.pdf

Globally speaking, you need a differential one, because you want to measure vacuum and pressure, and specify a pressure range.
But for experiments you can start with the one you have and go from there. If your board is assembled, all you need to do is connect a tube and the scope.

To add: the datasheet specifies a 10-90% swing in 1ms. Meaning a 30-70% swing in 0.5ms. So a higher pressure sensor is faster with a lower resolution. So one step at a time.

Did some further research, because the inbuild amplifier inhibits positive/negative readings, just to answer my own question. I did pos/neg before and now it was gone.

On this page you will see sensors with -x/x specifications:

https://www.nxp.com/products/sensors/pressure-sensors/vacuum-sensors-50-to-115-kpa:VACUUM_-50-115KPA#/

Select one at the top:

https://www.nxp.com/products/sensors/pressure-sensors/vacuum-sensors-50-to-115-kpa/-7-to-7kpa-differential-and-gauge-pressure-sensor:MPXV7007DP

Go to Buy and do a sample request over several pressure ranges and as much as they will hand out per pressure spec. Works normally well and it's as cheap as you can get, ie free. The law of permanence of problems here is that zero pressure is at 2.5 volt. So while you're busy, request also a couple of vacuum sensors.


I think they will do it, considering it is for New Zealand.

[TZ350]

.
Successfully fuel injecting a single cylinder high performance two stroke, i.e., something making more than 10 bar BMEP and greater the 10,000 RPM has been very difficult until now. Because it was impossible to apply four stroke air flow measurement techniques like MAF, manifold mass air flow and MAP, manifold absolute pressure to a two stroke.

A two strokes inlet manifold and crankcase have entirely different pressure profiles to a four stroke. Whereas a four strokes manifold pressure decreases when the throttle is closed and is at its highest when at WOT, wide open throttle with a range of 0.2 bar to 1 bar on a normally aspirated engine. A two stroke is the other way around. Maximum manifold and average crankcase pressure at minimum throttle and reduced pressure at WOT, wide open throttle and the range is 1 bar at no throttle to about 0.7 bar at WOT, wide open throttle.

The reason for the two strokes pressure profile is because at WOT and in the power band, the pipe is sucking hard on the crankcase and so reducing the pressure there. At closed throttle there is little or no pipe suction and atmospheric pressure floods back down the stinger and up the expansion chamber flooding the crankcase to atmospheric pressure with exhaust smog.

A two stroke can idle with a crankcase full of smog because a small amount of fresh mixture randomly finds its way through this soup to the cylinder. At idle a two strokes cylinder is filled with random proportions of smog and fresh mixture each cycle. So you have to run a bit rich here to give a fuel molecule a better chance of finding a O2 to combine with.

342089

It is very easy to setup a fuel injected two stroke for maximum power on WOT wide open throttle as air flow there is consistent and the Alpha-N fueling topology works well there. Alpha-N is TPS throttle position vis RPM and relies on consistent airflow.

But accurate fueling becomes increasingly difficult at lower throttle settings because the suction action of the pipe becomes increasingly variable and changes in air flow less consistent or predictable at lower throttle settings. At lower throttle settings the VE, volumetric efficiency topology is more appropriate. VE, volumetric efficiency is MAP, manifold absolute pressure vis RPM and is suited to variable airflow.

In a two stroke variations in air flow through through the motor can be seen by watching the Delta crankcase pressure. The difference between the lowest crankcase pressure near TDC and the highest near BDC. This delta is a reliable indication of changes in air flow. The greater the delta the greater the air flow.

The theoretically correct crank angle for the highest pressure is around 160 ATD and lowest about 20 BTDC but the quality and response of your pressure sensor may change things.

342090

The first step is to see how your sensor responds. Maybe run the engine up with a carburetor and view the sensors output and timing on a scope.

342088


This is how I would setup the EFI fueling using the delta crankcase pressure concept.

The Red area below 45% TPS and in the RPM area where the pipe is working I would setup the fueling for VE volumetric efficiency. MAP vis RPM.

Every where else in the Green area I would setup fueling as Alpha-N. TPS vis RPM. Airflow is consistently predictable here.

And in the Purple area below the RPM where the pipe works and TPS = 0% the MAP value needs to be set at 0.2 - 0.4 bar. This is necessary because when the pipe is not sucking strongly the crankcase pressure rises to someplace close to 0.8 - 0.9 bar of dirty air. This happens because the pipe is not drawing airflow through the motor and atmospheric pressure and stale exhaust gasses flow back into the crankcase raising the pressure there and in the inlet tract.

The motor can idle with a crankcase full of dirty air because some fresh air/fuel makes its way through the smog and randomly finds its way into the cylinder. At idle the cylinder has some fresh air and a lot of smog.




342091

A copy of my PesudoMAP software in PDF format for an Arduino Nano.

Basically this determines the Delta between high and low crankcase pressures and outputs it as a pseudo MAP value that can be used by any EFI system that can use MAP as an indication of airflow.

With their latest 150cc TPI offering KTM have shown that high performance two stroke EFI is possible. They use a crankcase pressure sensor but aren't about to reveal how seeing changes in air flow is done. But with the Delta crankcase pressure reading concept I have revealed the secret of how to see the all important changes in airflow through a two stroke motor.

[TZ350]

342099

342100

This is the software I developed for finding the difference between High and Low crankcase pressure using an Arduino Nano.

1) It has one interrupt triggered by the ignition pickup.
2) From the time for the last cycle it works out the time in micro seconds to each analog read point. Six in total.
3) It takes three analog high pressure readings and three analog low pressure readings.
4) It selects the best of each three because the exact crank angle moves with rpm so you have to straddle the optimum area.
5) The best of each gets added to a moving average.
6) The Delta difference is averaged again.
7) The Delta is multiplied to get its magnitude to look something like a real MAP value.
8) The result is continuously output via a DAC card as a 0-5V MAP value until the end of the next cycle.

342102 The Speeduino EFI CPU is in the box. The Arduino Nano and MCP4725 DAC, MAP co-processor are shown beside it.

Currently the software looks at the TPS and if its over 45% then the output is standard atmospheric as read by the map sensor at startup for Alpha-N fuel mapping.
If it is less than 45% then the Delta pressure is used for the MAP value and VE fuel mapping.
And if the TPS is 0% then the MAP value is forced to 0.2 bar otherwise the Delta pressure would be to high due to exhaust gasses flooding the crankcase.

342101


Clearly the high vs low pressure comparison works, but I can't get my head around only using 3 interrupts for taking the readings? For my engine, which uses crankcase reed valve induction and using the same cylinders as yours. The transfer port timing is symmetrical regardless of the revs, opening at 118 degrees ATDC (62 BBDC) and closing at 242 ATDC (62 ABDC or 118 BTDC). So shouldn't there be more interrupts around the opening/closing times?

The low crankcase pressure is just before TDC where the piston is still moving up and the incoming air has not had time to completely fill the crankcase. The high pressure is just before BDC when the piston catches up with air that has not yet been sucked up the transfers by the pipe. Gas inertia is everything here, port timing has only a small part in it.

The quality and speed of the pressure (MAP) sensor changes the timing of things so best to scope the sensors output at different RPM and loads then adjust the software's timing to suit. The three readings are so you can straddle the optimum crank angles for the best pressure readings.

[jbiplane]

May be interesting - uni project based on high speed ARM processor:
8 cylinder ECU with parts for 64USD
https://skemman.is/bitstream/1946/29570/1/LokaskyrslaRikardur.pdf

[obelsik]

Saw this posted on thumpertalk.

Could this sensor be used to detect miss-fire and therefore know the engine is not on the pipe and hence dial back the fuel?

https://www.cycleworld.com/2015/07/23/technical-update-detonation-detection-by-analyzing-ionization/

[dmcca]

I’ve had a lot of success with alpha-n on my YZ125... it runs much better than a carb at all throttle positions, especially low throttle and low/mid revs, and revs out to over 13,000. Makes 40 at the wheel so it’s in a moderate state of tune, not too high or too low. ItÂ’s one of the best bikes I’ve ever ridden off-road.

To be honest it’s very confusing because everything you’re saying makes perfect sense about using VE, and before I started building it a didn’t think alpha-n would work properly on a 2 Stroke however this bike has responded extremely well to TPS/RPM only since the very beginning. It’s been running for almost 3 years now and I’ve come full circle back to a throttle body with 2 injectors in it. Previously tried injectors in the cylinder and cases but TBI works fine with much less hassle.

IÂ’m still unsure as to why it’s worked so well though. Maybe there’s something about the engine/pipe/riding style (off-road) thatÂ’s more forgiving for EFI?? Maybe temps are more stable and therefore have less impact?? I’ve got no idea!


As an aside I’ve been reflashing the KTM TPI ECUs... figured out a way to access and modify the hex data using a BDM interface. Took about 200 hours on the dyno to figure out where the maps were and how to change them. Getting huge gains in performance though, especially at part throttle. Based on the maps I’ve identified I’m pretty sure the crank pressure sensor is only referenced up to 4,000rpm and less than about 20% throttle. Could be wrong though as I’m still figuring it all out and it’s not at all straightforward. The stock KTM ignition maps are ridiculous and are the cause of most of the spluttering that the TPIs suffer from at part throttle.

[TZ350]

.
That is very interesting, very encouraging, thanks.

If you would like to post more about your project and your adventures with two stroke EFI, that would be great.

[jbiplane]

This is automatically tuning system FAI EFI for 4-stroke.
https://www.ebay.com/itm/FAI-motorcycle-scooter-buggy-moped-UAV-EFI-engine-Electronic-Fuel-Injection-kit-/273178456487
Seems should be simple and elegant way use it on 2-strokes?

[husaberg]

How quickly does the exhaust gas temp change with different running conditions Rob.
I guess judging by water injection falure, its to slow to be useful as a calibration?

[TZ350]

How quickly does the exhaust gas temp change with different running conditions?

I have not done much work with EGT for myself but what I have seen with other bikes on the dyno the EGT seems as responsive as my O2 sensor. In other words, pretty good.

[TZ350]

.
Someone was asking about RPM and TPS spacing.

The steps between RPM and TPS points does not have to be even. The trick is to position them so that the fuel "Load" steps between cells is smooth as possible, no big jumps.

Typically the RPM steps follow the Torque curve and some of the RPM steps are bunched together in the rpm area where the torque curve is steepest as the motor is getting on the pipe. The idea is to keep the increasing fuel steps as even as possible as the RPM changes.

342711 upside down again %^$^$%)% ...:angry2:

The graph is a typical carburetor flow curve. I tested my ball valve inlet and got the green line. 50% of air flow is typically at 25% throttle open. So at least half the fuel steps need to be below the 25% throttle line to get the increasing fuel demand steps due to the steadily increasing air flow in as even as steps as possible.

The point is RPM and TPS steps can be uneven, its the steps between Fuel load cells that need to be as even as possible.

[TZ350]

TZ, when looking at your VE table, it looks like you've only allocated 6 columns from 3000 to 7200 rpm for the crankcase pressure differential measurement, whilst giving 10 columns to Alpha N from 7800 to 13500 rpm? I thought that the area (in rpm) below where the pipe starts working would require finer control and thus more columns. On a separate note are you running the engine with premix or oil pump?

Premix. 20:1.

No, the area below (and above) the transition area where the pipe starts to work does not need as many columns as the transition area itself.

I have abandoned the VE map approach and for now are concentrating on Alpha-N multiplied by the MAP/Baro ratio.

Alpha-N is great for everywhere the airflow is consistent. So Alpha-N is good for everywhere that is not on the pipe and every where on the pipe and above 50% TP throttle position. There is only a smaller area where Alpha-N does not work well and that is in the area where you could be on the pipe and less than 50% TP. That is because air flow here is variable due to inconsistent pipe action.

Currently my crankcase PseudoMAP outputs a MAP value equal to the Baro value in the areas where airflow is going to be consistent. So the Alpha-N map here is multiplied by the ratio of 1:1 i.e., Alpha-N * 1.

And in the area where the engine should be on the pipe but the TPS is less than 50% the PseudoMAP outputs three times the Delta crankcase pressure. So in the variable air flow area the Alpha-N map is multiplied by the ratio of PseudoMAP to Baro pressure. i.e., Alpha-N * (PseudoMAP/Baro).

Anyway this is the current approach I am exploring, hopefully find out soon how well this concept works.

342724

[husaberg]

We can only imagine a bunch of Austrians are feverishly looking into it now Rob.

[ceci]

We can only imagine a bunch of Austrians are feverishly looking into it now Rob.



You what group of Austrians you mean.
If they are these, I see no future

https://www.visordown.com/news/industry/ktm-buys-majority-stake-gasgas-build-new-electric-motorcycles

[husaberg]

You what group of Austrians you mean.
If they are these, I see no future

https://www.visordown.com/news/industry/ktm-buys-majority-stake-gasgas-build-new-electric-motorcycles

What the attendance like at the Electric formula 1 compared to the ICE F1.

[patouille]

Hi guys,

I was very quiet since my last post but I was still developing my EFI.
As my engine was struggling to rev over 10,000RPM, I came to the conclusion I needed staged injection. But no...
Using the oscilloscope in my hall sensor signal with the carb, I discovered that kind of sensor was guilty.
So I changed the way to pickup the rev information. I used an optical switch and now everything works perfectly !

https://i.imgur.com/fzfe5Uk.jpg

I use a CBR125 throttle body but the original injector went over 80% duty cycle despite its 140cc/min flow rate. I use now a 200cc/min injector which a bit high but works good.

I usually complain that most of 2S EFI projects in the internet are static tests at low RPM. Here is a small glance of the bike delivering all the power on the track. Yes it is race proven ! :Punk:


https://youtu.be/v_-1vPlJtmw

[ceci]

Some time ago I suggested that KTM was the brand that would have most investigated the EFI two stroke engine.
Here we have one more test

https://diglib.tugraz.at/download.php?id=5aa249252fafb&location=browse
https://diglib.tugraz.at/download.php?id=5891c6f4c7db6&location=browse
https://diglib.tugraz.at/download.php?id=5aa24914439ca&location=browse

[ceci]

Some time ago I suggested that KTM was the brand that would have most investigated the EFI two stroke engine.
Here we have one more test

https://diglib.tugraz.at/download.php?id=5aa249252fafb&location=browse
https://diglib.tugraz.at/download.php?id=5891c6f4c7db6&location=browse
https://diglib.tugraz.at/download.php?id=5aa24914439ca&location=browse



I was wrong not yet from KTM