Speeduino 2T EFI Project

[ceci]

It has been shown that getting an injection system for the two-stroke engine that exceeds ecological standards is super difficult.
Since it has a high level of difficulty, it is shown by the large number of manufacturers that have tried and failed
One more https://gair.media.gunma-u.ac.jp/dspace/bitstream/10087/5138/1/Ph.D_E1-383.nishida.pdf

[TZ350]

.
After months of wandering lost in the wilderness I think I have finally found the reason for my starting issue. It is simple really.

I use a heavy duty drill running in reverse driving a big nut on the crankshaft to start the bike. This had worked well for ages. Then the old drills gearbox broke and I had to get another. Same reverse direction and speed. Also upgraded the firmware and had nothing but trouble since.

After finaly getting around to buying a scope and putting it on the multi tooth wheel's inductive trigger pickup and studying the trace and thinking about it for a bit I realized what was happening.

The crank looked to spin around impressively but the trace showed that as the piston started coming up against some real compression near TDC . The crank slowed down and there was not enough flux through the inductive trigger pickup to generate a signal big enough and with enough definition for the Speedy hardware to recognize. My 24-1 wheel looked like something with two gaps, one regular and another larger one where the flux was weak.

The answer is to get a powerfuller drill. Simple really........... it has only taken ages to figure that out.... https://speeduino.com/forum/images/smilies/icon_lol.gif

344335

The yellow trace is taken at Pin 19 on the Speeduino's Mega processor board and the purple trace is the inductive trigger pickup.
.

[Vannik]

Good! I have been coming here and to the Speeduino blog almost daily to look for progress!!!

[husaberg]

Good! I have been coming here and to the Speeduino blog almost daily to look for progress!!!

Hi Neels

You can set your thread subsricptions
(to subsribe to a thread post in it or use the button top right of each thread page , Use thread tools, subscribe to thread)
then alter your setting to notify you by email when there has been an update to the thread you are intersted in.
344350344351344352
Settings
General settings and default thread description method.

[Vannik]

Hi Husa,

Thank you, but I am aware of the facility. It is one of my daily pleasures to come to KiwiBiker first thing in the morning and see what was posted while I was sleeping. My "gripe" was not with comming here but with finding nothing :laugh:

And I will keep on doing it.

[husaberg]

Hi Husa,

Thank you, but I am aware of the facility. It is one of my daily pleasures to come to KiwiBiker first thing in the morning and see what was posted while I was sleeping. My "gripe" was not with comming here but with finding nothing :laugh:

And I will keep on doing it.

I did wonder, when you were clearly a bit better then averagely skilled wit that computer.:killingme

[ceci]

I apologize for not being at the level that corresponds to this forum.
I would have liked to bring photos and videos of a prototype.
I tried to make a prototype and for that I bought part of the components (I showed them on page 4 of this thread "the one I wrote in Spanish").
Due to lack of means to be able to do it (I expected the collaboration of a family member who does have mechanical means to do it, but he emigrated and with it the possibility of cooperation).
What I present is the scheme of the prototype I wanted to make.
The operation is that of a stratified load motor with the variation that it has two modes of operation, when it turns slowly, not all the pressurized air is used to carry out the mixing, the other part is introduced by the exhaust
When it turns quickly if you use all the air under pressure to mix

[ceci]

I apologize for not being at the level that corresponds to this forum.
I would have liked to bring photos and videos of a prototype.
I tried to make a prototype and for that I bought part of the components (I showed them on page 4 of this thread "the one I wrote in Spanish").
Due to lack of means to be able to do it (I expected the collaboration of a family member who does have mechanical means to do it, but he emigrated and with it the possibility of cooperation).
What I present is the scheme of the prototype I wanted to make.
The operation is that of a stratified load motor with the variation that it has two modes of operation, when it turns slowly, not all the pressurized air is used to carry out the mixing, the other part is introduced by the exhaust
When it turns quickly if you use all the air under pressure to mix



After a year and a half of starting the attempt to create the prototype, the progress is zero due to the lack of mechanical means to carry it out.
The components that I already showed, remain unmodified, which discouraged me from buying the rest of the components.
It seemed easy to make the prototype since it is made up of parts in 90% of other motorcycles and only 10% of parts created specifically for it (try to make these pieces of mechanization instead of casting, to have the maximum possibilities of making them)
But as it says in my land "God gives bread to those who don't have teeth" and that's the way it gives me ideas but doesn't give me the means to make them

[ceci]

I opted for the most complex option (to make a prototype) but at the same time the most appropriate one since real checks and data can be obtained.
The other simpler option was to make a model in 3D printing, with this option you could have made the first design, since this one has a majority of special designed parts (nonexistent on other motorcycles).
This second option would allow me to record videos of the operation, but I already have that through the simulation in the CAD.
These images are of the first design, in the non-use of the double piston in tandem for the compressor, what I use is a secondary compressor driven by spring and controlled by electromagnet

[TZ350]

.
Hi Ceci. It looks very interesting.

[ceci]

.
Hi Ceci. It looks very interesting.


Thank you.
Now I value more the projects that many of you are doing, because I know of the difficulty that it leads (I have failed in the attempt and that has made me understand it).
What I regret most is seeing the little interest in the 2S world to find anti-pollution solutions, I see (not in this forum) many attempts to overcome the work of Jan Thiel and Frits Overmars in obtaining more HP

[ceci]

I don't know about you, but I'm impatient to see the results of the international conference on 2-stroke direct injection engines, these days February 13 and 14 at IFP School!

[ceci]

The image is not from Orbital, it is from Yamaha that also had its intent in the Pneumatic direct-injection
In the scheme of the prototype I use the valve "Pneumatic direct-injection (PDI) system" of honda, in fact I did not plan to use that type of valve, I planned to use a hollow shaft with an outlet slot.
As for the first project, this used 1700cc Bosch EV14 injectors, transformed for double flow control operation, with a considerable loss of fuel flow, but in exchange for an improvement in the spray droplet size.





https://patentimages.storage.googleapis.com/ac/d0/af/b587abaeb851d7/imgf0005.png

[ceci]

I reached the figure of 100 posts in this forum, a figure that I set as a limit. Therefore this is my last publication.
I say goodbye by showing and clarifying what I did not show in my first publication (the internal operation of the video I published).
The video of the movement of the first project uses a hinged piston since the mechanical simulation program does not allow a piston without a hinge.
The images show the situation of the piston without hinge on the basis of a simulation.
The video of the attempt to create a prototype has no complexities
I am very grateful for allowing myself to enter your group of experts without being an expert.
I will continue visiting this forum since there are no more places that look for solutions to the problems of the 2S

https://www.youtube.com/watch?v=XP8tmtONnGY&feature=youtu.be

https://www.youtube.com/watch?v=BQgsGoX7YSE

https://www.youtube.com/watch?v=odkMKerW1KQ

[sonic_v]

Excellent Work!

As I understand things from reading the whole thread your present methodology is to take crankcase pressure readings and record the minimum and maximum pressures then subtract the minimum from the maximum and multiply the result by a constant - presently 3.0. This gives you a value that is representative of Delivery Ratio.

It is evident when using maximum pressure that it sometimes occurs 30 degrees past Transfer Port Opening. The problem with this is that part of the pressure rise is due to delayed flow from crankcase into cylinder due to cylinder pressure being higher than crankcase pressure - at TPO - at certain speed/throttle position combinations and hence this peak pressure is not fully representative of the Delivery Ratio.

Taking your pressure readings at TPO might well be the way to go.

[TZ350]

.
Yes. I think you are right. I will try and get some pictures of real time oscilloscope traces taken at the sensors output.

[TZ350]

.
344881 10,000 RPM 344882 5,500 RPM 344883 1,800 RPM (barely idling, much smoother at 3,000 rpm).

Some older oscilloscope traces. It is a bit hard running the bike up on the dyno by myself and grabbing the traces. Yellow = Ign and Blue = Crankcase pressure transducer. (transducer outputs every 1ms).

In some other old traces I can see a marked difference in heights of the crankcase pressure reading.

A bunch of traces at different RPM and Throttle positions and engine load would be very enlightening. I will see if I can get a friend to help me.

[sonic_v]

Are you stuck with reading crankcase pressure every 1ms or is this old data?

[TZ350]

Are you stuck with reading crankcase pressure every 1ms or is this old data?

I am stuck with a transducer that is reading/transmitting in complete 1ms cycles, it gives a stepped output. All affordable small pressure transducers I have seen are like this. I would be happy to try something else that has a 0-5V analog output and greater than a 1 kHz response rate (ie; << 1ms a cycle).

Yes old data, I recorded these traces a while ago. I am going to try for some new more meaningful data with comparative traces that shows the difference in heights relative to engine load.

[ceci]

TZ350 excuse me asking, it may be EFIhttps://scontent.fvlc4-1.fna.fbcdn.net/v/t1.0-0/p640x640/94614497_2977246909007804_5764782517181743104_o.jp g?_nc_cat=109&_nc_sid=07e735&_nc_oc=AQkTdlQ7NEcR3NVjnIyp5FamLYhADJMk2NcEfYBwETH BAJEYjeYf_Ec1vwIhi0_LAko&_nc_ht=scontent.fvlc4-1.fna&_nc_tp=6&oh=7879322e70cfdf998e59b2b439447902&oe=5ECB98D9

[ceci]

As always my impatience defeats me, it is already clear if it is,

[TZ350]

.
(Single?) Fuel Injector. It would be interesting to know more.

[ceci]

.
(Single?) Fuel Injector. It would be interesting to know more.


I can't find information about this engine, but I have a theory, but I can't find a relationship between my theory and the engine.
I suspect it is the design of Marconi Pierluigi who is the designer of the Bimota Vdue, this is the patent, the location of the injector is the same.

https://patentimages.storage.googleapis.com/d1/eb/95/4e482a67a03d4e/WO1993022545A1.pdf

In this engine it is between that patent and the GasGas patent also from the same engineer. Which makes it also a combined injection

[ceci]

The engineer will not be the one I suspect, but the type of injection is combined

https://scontent.fvlc4-1.fna.fbcdn.net/v/t31.0-8/p720x720/1796012_1798541806951472_1305383168_o.jpg?_nc_cat= 102&_nc_sid=2d5d41&_nc_oc=AQms3HMTe9CPPMEC94apcI07VWge5qqJRAPPsACyGsO 1lUCVx83F506C8ZYFYHlihkc&_nc_ht=scontent.fvlc4-1.fna&_nc_tp=6&oh=60b81b18e5ede4c6cd8c89dfd3511a98&oe=5ED126D9

[TZ350]

.
It will be interesting to know which port is used for slow running and which for full noise up on the pipe.

[ceci]

.
It will be interesting to know which port is used for slow running and which for full noise up on the pipe.

I am not sure that it is from that engineer.
In the Spanish patent the main injection is in the cylinder.
The one in the carburettor body is like a powerjet

[TZ350]

.
345792

Does the cylinder injector squirt only under the piston like mine does. Or both above and below?

The injector in the throttle body for main fuel at higher power when the engine is on the pipe makes sense.

345793

I squirt my small low speed injector under the piston for best mixture homogenization at low throttle opening. The big injectors for power are in the B ports.

[husaberg]

the Bimota system injects direct to the bore
345794345795345796
plenty of pics here
https://www.kiwibiker.co.nz/forums/album.php?albumid=5064

[Flettner]

Interesting isnt it. All this complication just to try to avoid fuel / air spilling out the exhaust port.
Uniflow, throw the exhaust down the other end of the cylinder, and it all becomes so much easier to control.
I would bet it would be easier to have a clean exhaust with a carburetor Uniflow than with an injected loop scavenge twostroke.
Add some injection control to Uniflow, there is your answer. 😆😆

[Drew]

Forced induction, direct injection.

The problem is quite easily solved.

[ceci]

Forced induction, direct injection.

The problem is quite easily solved.

And why don't they market it

[Drew]

And why don't they market it

Never heard a train go by?

[sidecar bob]

Never heard a train go by?
Detroit 53 cu in per cylinder two stroke diesel engine went into production in 1957.
https://www.dieselhub.com/specs/detroit-53-series-two-stroke-diesel.html
Never understood why the same concept never evolved as a spark ignition petrol engine.

[husaberg]

Detroit 53 cu in per cylinder two stroke diesel engine went into production in 1957.
https://www.dieselhub.com/specs/detroit-53-series-two-stroke-diesel.html
Never understood why the same concept never evolved as a spark ignition petrol engine.
https://www.dieselhub.com/specs/detroit-71-series-two-stroke-diesel.html
http://usautoindustryworldwartwo.com/General%20Motors/detroit-diesel-the%20power%20to%20win.htm
Detroits were in production in ww2 twins in shermans, quads in landing craft but there is a reason you and drew seemed to have overlooked in why they have fallen out of favour despite their ruggedness, simplicity, power denisity and their cool noise.
they could not meet modern emissions

it looks like there was a methanol version was trialed though

https://www.sae.org/publications/technical-papers/content/831744/

[Flettner]

Simplicity? Valves, x4 per cylinder camshafts and a super charger.

Poppet valves are rubbish anyway, piston opening a port is far superior.

https://youtu.be/BPr694nlUKE

No poppet valves, no heads, no supercharger, no camshafts.
just simplicity, and uniflow.

[husaberg]

Simplicity? Valves, x4 per cylinder camshafts and a super charger.

Poppet valves are rubbish anyway, piston opening a port is far superior.

https://youtu.be/BPr694nlUKE

No poppet valves, no heads, no supercharger, no camshafts.
just simplicity, and uniflow.

or in a boat

https://www.youtube.com/watch?v=uiXsPkP9jvw

[Vannik]

https://www.autoevolution.com/news/the-ice-shouldnt-die-says-worlds-most-polluting-company-143085.html#agal_0

[sidecar bob]

Simplicity? Valves, x4 per cylinder camshafts and a super charger.

Poppet valves are rubbish anyway, piston opening a port is far superior.

https://youtu.be/BPr694nlUKE

No poppet valves, no heads, no supercharger, no camshafts.
just simplicity, and uniflow.

How are you managing the heat produced in the centre of the engine?

[Flettner]

I believe that is whats used to push the pistons down. 😆
If you are talking about managing the heat in the exhaust piston, yes thats an issue. In the boat engine above, the exhaust piston was cooled by fuel under the piston like a normal twostroke. I have since learnt other methods of exhaust piston thermal control, still using the crankcase as an air pump, like a normal twostroke.

[lohring]

Add a tuned pipe and you have the simplest uniflow engine with real power. Maybe with a version of your injection system it could meet emission rules. In any case, the catalytic muffler's back pressure isn't as harmful as in a four stroke.

Lohring Miller

PS I spent a Christmas day jet boating on the Haas River. I love New Zealand.
345813

[ceci]

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.

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.

You published this for a long time and at the time it was of great interest to me, but from my point of view they were only theories that I had not seen demonstrated.
This year 2020, two prestigious researchers from this forum have carried out experiments with mechanisms in which the pressure of the crankcase makes them work.
1nd Mr. Señor Katinas, published in March, reeds in the transfer, I am very struck by the graph of the 4th configuration.
2nd Mr. Ken, published in April, reeds in the butt, it seems to me that the behavior of the engine in the video is similar to the graph of the 4th Katinas configuration

[ceci]

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.

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.


My English is not bad, it is only from the Google translator, because I do not know English, there are some of my errors.
As I have already said the two experiments clarify certain things about what is expressed by TZ250, in both cases the crankcase is full of clean mix, since the contamination cannot enter due to the physical barriers of the reeds.
They also demonstrate that the maximum crankcase pressure is at low rpm, which is when the reeds resistance overcome.

[TZ350]

... in both cases the crankcase is full of clean mix, since the contamination cannot enter due to the physical barriers of the reeds.

Clean mixture in the crankcase would be a great advance on conventional two stroke technology. I wonder if the clean crankcase mixture does away with the 4, 6 and 8 stroking (misfiring) (corn popping) that you can get with a two stroke at low speed.

[ceci]

I wonder if the clean crankcase mixture does away with the 4, 6 and 8 stroking (misfiring) (corn popping) that you can get with a two stroke at low speed.



I wonder if that effect occurs in all engines or in certain settings of some engines

[ceci]

Whenever I have discussed this injection system, Ken Seeber has been referred to me for more information.
With this type of injection system some brands "aprilia, honda and ktm" have been investigated to apply it to engines over 50cc.
This new research work can help us understand why cylinder 250cc motorcycles have never been marketed with this type of system.

https://www.sciencedirect.com/science/article/pii/S0016236120300399


https://ars.els-cdn.com/content/image/1-s2.0-S0016236120300399-gr2.jpg

[MDJGyurgeiz]

Goodmorning everyone. I am converting my enduro bike (KTM exc 125 2006). i am doing all the necessary work and i already have all the necessary hardware (speeduino no2c, iat, CLT, wideband 4.9, tps, trottle body 38mm with iwp 043 330cc injector).
flywheel modified with 12-1 teeth
I wanted to understand with you who can help me for an initial tune to use
I need how many map sensors and where should I install them?
give me all the advice you can and if I'm doing something wrong

[TZ350]

Goodmorning everyone. I am converting my enduro bike (KTM exc 125 2006).

Hi MDJGyurgeiz. Great to see your project. The first thing to know is that there is a definite barrier about 9-10,000 RPM for fuel injecting two strokes. Above 10k RPM you will need staged injection. Below that you may get away with a single injector and at about 7-8k RPM a single injector is definitely possible.

This is because of limited time at higher RPM and the fact it takes an injector a finite amount of time to open and as its opening it dribbles fuel. An injector big enough to get the job done in the limited time available at high rpm dribbles far too much for successful tuning at lower RPM and Load. So two injectors are required, a small one and a much larger one. Two physical "B" port injectors can be treated as one logical injector when they are fired together.

KTM's EFI 300 with two physical injectors, one in each "B" port does a neat trick. At low RPM and Load they alternate the injections from side to side and when the RPM gets up or the Load increases they fire both together.

My 125 uses a 50cc/min small injector for low RPM and/or low Load and then switches over to the two big (160 cc/min each) B port injectors for a (logical) injection of 320cc/min.

MAP and MAF air flow sensors from a car will not work on a single cylinder two stroke. The reasons are all explained in earlier posts on this thread.

To get your bike up and running all you need is a TPS (throttle position sensor) and ambient air and cooling water temperature sensors. And maybe a MAP sensor connected to the outside air for variations in daily ambient air pressure.

The best MAPing method to start with, is the Alpha-N approach. This is just RPM vis TPS. Very easy to get started with and to tune for best power.

Alpha-N works well for anything pedestrian like a 300cc enduro bike or a two stoke car engine, basically anything two stroke that is not particularly high performance and maybe maxis out at less than 8,000 RPM.

A tuned 125 that relies on the pipe for its power needs a touch of VE volumetric efficiency in some feathering the throttle situations like approaching and exiting from a corners apex. The reasons for this is also explained in other posts on this thread.

But get your bike running with an Alpha-N map to start with, then add in a second VE map for fine tuning later.

A wide band O2 sensor is real handy for tuning for high power and areas on the map where the engine is running on the pipe and the engines trapping efficiency is at its best. But a O2 reading can be misleading when there is air short circuiting out the exhaust port or the engine is too rich and miss-firing. The O2 sensor will tell you air/fuel is lean when in fact its far too rich. Just another example of how four stroke thinking can be a trap with two strokes.

Good luck and please keep posting pictures and reports on your progress. Happy to answer questions here.

[MDJGyurgeiz]

Hi MDJGyurgeiz. Great to see your project. The first thing to know is that there is a definite barrier about 9-10,000 RPM for fuel injecting two strokes. Above 10k RPM you will need staged injection. Below that you may get away with a single injector and at about 7-8k RPM a single injector is definitely possible.

This is because of limited time at higher RPM and the fact it takes an injector a finite amount of time to open and as its opening it dribbles fuel. An injector big enough to get the job done in the limited time available at high rpm dribbles far too much for successful tuning at lower RPM and Load. So two injectors are required, a small one and a much larger one. Two physical "B" port injectors can be treated as one logical injector when they are fired together.

KTM's EFI 300 with two physical injectors, one in each "B" port does a neat trick. At low RPM and Load they alternate the injections from side to side and when the RPM gets up or the Load increases they fire both together.

My 125 uses a 50cc/min small injector for low RPM and/or low Load and then switches over to the two big (160 cc/min each) B port injectors for a (logical) injection of 320cc/min.

MAP and MAF air flow sensors from a car will not work on a single cylinder two stroke. The reasons are all explained in earlier posts on this thread.

To get your bike up and running all you need is a TPS (throttle position sensor) and ambient air and cooling water temperature sensors. And maybe a MAP sensor connected to the outside air for variations in daily ambient air pressure.

The best MAPing method to start with, is the Alpha-N approach. This is just RPM vis TPS. Very easy to get started with and to tune for best power.

Alpha-N works well for anything pedestrian like a 300cc enduro bike or a two stoke car engine, basically anything two stroke that is not particularly high performance and maybe maxis out at less than 8,000 RPM.

A tuned 125 that relies on the pipe for its power needs a touch of VE volumetric efficiency in some feathering the throttle situations like approaching and exiting from a corners apex. The reasons for this is also explained in other posts on this thread.

But get your bike running with an Alpha-N map to start with, then add in a second VE map for fine tuning later.

A wide band O2 sensor is real handy for tuning for high power and areas on the map where the engine is running on the pipe and the engines trapping efficiency is at its best. But a O2 reading can be misleading when there is air short circuiting out the exhaust port or the engine is too rich and miss-firing. The O2 sensor will tell you air/fuel is lean when in fact its far too rich. Just another example of how four stroke thinking can be a trap with two strokes.

Good luck and please keep posting pictures and reports on your progress. Happy to answer questions here.

thank you very much for the advice, I will proceed as you advised me. do you think the map sensor at the moment is better to connect it to the throttle body?
can you recommend a petrol pump with integrated regulator to connect the outlet to the petrol tank?
can I ask your help to configure the tunerstudio tune?

[TZ350]

... do you think the map sensor at the moment is better to connect it to the throttle body?

No. do not connect it to anything, just let it sense the outside ambient air pressure.

I am using the Speeduino V4 board with the onboard MAP sensor. I have not connected a hose to the sensor, it just reads ambient air pressure. That way it can adjust the tune for elevation and air pressure changes in the day.

Get your project started with the Alpha-N methodology first, (TPS vis RPM) then we can add a pseudo MAP sensor later for an additional VE map.


can you recommend a petrol pump with integrated regulator to connect the outlet to the petrol tank?

This is the fuel pump I use https://www.ecotrons.com/wp-content/uploads/2020/08/ECOTRONS-Fuel-Pump-technical-spec-V1.0.pdf you will be able to find something similar and cheaper on AliExpress. I use a Honda (but could be anything) 3,5 bar fuel rail pressure bypass type regulator salvaged from a car wreckers.


can I ask your help to configure the tunerstudio tune?

Yes, I may be able to help. What Speeduino board are you using? For technical Speeduino problems like noise on the signal line you may have to go to the Speeduino projects forum but anything two stroke EFI tuning with your project I may be able to help.

EFI and two strokes does work, people have had great success with it.

A big must. Get the paid for full version of Tunerstudeo and their Data logging and graphing software.
.

[TZ350]

.

Later, for a two stroke VE table we need to do something different to four strokes.

We need to create a pseudo MAP value that mimics four stroke behavior.

This search will bring up a lot of references about it on this thread.

First follow this link:- https://www.kiwibiker.co.nz/forums/showthread.php/185773-Speeduino-2T-EFI-Project? (https://www.kiwibiker.co.nz/forums/showthread.php/185773-Speeduino-2T-EFI-Project?p=1131168950#post1131168950)

Then copy and paste this in front of the https: Crankcase pressure site:

And you should see a Google page that looks like this:-

346865

If you have Google and you can use the "site:" option then you can find things much more easily.

[TZ350]

.

You can also search for interesting things by using the thread images search tool.

346868 go to "Thread Tools" at the top of the page Then "View Thread Images".

346867 select "Beginning"

346866 select "70 to Page"

346870 press "Show Results"

346869 press the "N/A" on any picture and it will take you to the original Post.

[MDJGyurgeiz]

No. do not connect it to anything, just let it sense the outside ambient air pressure.

I am using the Speeduino V4 board with the onboard MAP sensor. I have not connected a hose to the sensor, it just reads ambient air pressure. That way it can adjust the tune for elevation and air pressure changes in the day.

Get your project started with the Alpha-N methodology first, (TPS vis RPM) then we can add a pseudo MAP sensor later for an additional VE map.



This is the fuel pump I use https://www.ecotrons.com/wp-content/uploads/2020/08/ECOTRONS-Fuel-Pump-technical-spec-V1.0.pdf you will be able to find something similar and cheaper on AliExpress. I use a Honda (but could be anything) 3,5 bar fuel rail pressure bypass type regulator salvaged from a car wreckers.



Yes, I may be able to help. What Speeduino board are you using? For technical Speeduino problems like noise on the signal line you may have to go to the Speeduino projects forum but anything two stroke EFI tuning with your project I may be able to help.

EFI and two strokes does work, people have had great success with it.

A big must. Get the paid for full version of Tunerstudeo and their Data logging and graphing software.
.

great thank you very much. I'm proceeding as you advised me. I just need to finish the wiring and then I can start starting the engine. i am using as ecu speeduino NO2c which is the smaller and more compact version of speeduino and is able to manage two injectors. I made a 12-1 flywheel, audi rs COB coil, Bosch water and air temperature sensor. tunerstudio and megalog I have already bought them so even at the software level they are fine. I started doing a tunerstudio project but I would like you to help me set it up. I attach it here346874

[TZ350]

great thank you very much. I'm proceeding as you advised me. I started doing a tunerstudio project but I would like you to help me set it up. I attach it here346874

I have not tried to open someone else's TunerStudio project before and have utterly failed with yours. Anyway get your bike wired and lets see what happens.

[MDJGyurgeiz]

I have not tried to open someone else's TunerStudio project before and have utterly failed with yours. Anyway get your bike wired and lets see what happens.

try this file346928

[TZ350]

try this file346928

Yes, I can open that one.

[MDJGyurgeiz]

Yes, I can open that one.

perfect. if you can give me a hand to make a first configuration of the project I would be grateful. so you start with a solid foundation to start with.

[TZ350]

perfect. if you can give me a hand to make a first configuration of the project I would be grateful. so you start with a solid foundation to start with.

I made a few changes, hopefully they will help you get going. You will still have to fill out the VE table. You will still have to enter the TPS and RPM break points and fill out the cell entries. Talk more when you are ready start trying to fire the bike up.

346952

[TZ350]

..........

[MDJGyurgeiz]

..........

perfect. I entered all the settings you recommended on tuner studio. you can help me also for these that I am attaching in the photo. ?

[MDJGyurgeiz]

I made a few changes, hopefully they will help you get going. You will still have to fill out the VE table. You will still have to enter the TPS and RPM break points and fill out the cell entries. Talk more when you are ready start trying to fire the bike up.

346952
can you give me a hand in these settings that I have attached above? do you know what advance I can use if I had to clone the original CDI ECU?

[TZ350]

can you give me a hand in these settings that I have attached above? do you know what advance I can use if I had to clone the original CDI ECU?

Because you have so many 255 entry's I suspect you did not load the Speeduino Base Tune first when you started your project.

Ignition, a safe place to start is 18 or 20 deg BTDC and as it gets into the higher RPM steadily retard it 4 to 8 degrees. Yes with a two stroke you retard the ignition at high RPM.

Myself I am not using the Speeduino for ignition control, I only use it for fuel injection. I have a different digital DC powered CDI ignition for lighting the fire.

My EFI settings can only be a rough guide. If you had exactly the same bike as me you could follow my settings exactly but as it is, you will have to develop your own settings that suit your bike.
.

[TZ350]

.

Loading the Tuning Studio Firmware Definition ini file and the Speeduino base tune.
Speeduino Manual:- https://speeduino.com/Speeduino_manual.pdf
Page 11:- Creating a new Project in TunerStudio.
Page 12:- TunerStudeo Firmware Definition File.
Page 15:- Speeduino Base Tune.


Basically.
Step 1:- go to the reference sub directory in the Speeduino folder.
Step 2:- load the Speeduino ini file. (Firmware Definition File for the Tuning Studio software).
Step 3:- Go to the Base Tunes sub folder
Step 4:- and load the Speeduino base tune.
Then you can modify Speeduino base tune to suit your project.
.
347008 347009 347010 347011
.
By loading the Speeduino base tune all of the screens are filled out for a four stroke. Easily changed too two stroke settings and you wont overlook some obscure initial setting that might cause you a lot of grief until you discover it.
.

[obelsik]

The KTM 65 SX has the power valve driven open by exhaust gas pulses into a diaphragm. Could this be used as a pseudo map sensor?

Sent from my SM-G960F using Tapatalk

[TZ350]

The KTM 65 SX has the power valve driven open by exhaust gas pulses into a diaphragm. Could this be used as a pseudo map sensor?

Intuitively seems like a good idea, worth trying if you have the time.

Using the Google "SITE:" search option and this threads URL you can read about other things that have been tried in the search for a useful MAP reading. Some more successful than others.

347012

[MDJGyurgeiz]

Because you have so many 255 entry's I suspect you did not load the Speeduino Base Tune first when you started your project.

Ignition, a safe place to start is 18 or 20 deg BTDC and as it gets into the higher RPM steadily retard it 4 to 8 degrees. Yes with a two stroke you retard the ignition at high RPM.

Myself I am not using the Speeduino for ignition control, I only use it for fuel injection. I have a different digital DC powered CDI ignition for lighting the fire.

My EFI settings can only be a rough guide. If you had exactly the same bike as me you could follow my settings exactly but as it is, you will have to develop your own settings that suit your bike.
.

I made the whole configuration by putting together your advice and a first configuration that I did. can you give a final check to the project that I am attaching to you and modify or tell me what you don't like?347032

[TZ350]

.
Trigger angle is the most important setting to get right.

347042

With the piston at TDC (top dead center). Use a protractor to count the number of degrees from the falling edge of number one tooth that follows the "Gap" to just past the center of the hall sensor / trigger coil.

347043

I would definitely use TPS not MAP for the ignition.

347044

Correct calibrations are a must too.

347045 347051

The cell entry's in a TPS/RPM (Alpha-N) Map represent the % percentage of your 2.5ms setting in the Required Fuel Calculator in your Engine Constants. So a cell with a 100 means 100% of 2.5ms injector time is required by that cell.

The maximum cell entry allowed is 255. That is 2.5 times the calculated "Required Fuel Calculation".

[MDJGyurgeiz]

I made the whole configuration by putting together your advice and a first configuration that I did. can you give a final check to the project that I am attaching to you and modify or tell me what you don't like?347032

Hello. I don't know if you managed to see my latest tunerstudio project that I have attached to you above, however I wanted to ask you what you think of these advance curves that I am attaching to you. I found them on the internet and they correspond to a variable advance CDI kit for my KTM 125. which do you think is the best of these to use by copying the in the tunerstudio project? last question, what afr value do you recommend to use in tunerstudio?

[TZ350]

.

https://www.kiwibiker.co.nz/forums/showthread.php/86554-ESE-s-works-engine-tuner/page2000
347052
Start here with 18 - 20 deg.

Massive advance that works on full throttle will promote detonation and engine damage on semi closed throttle.

Beware..... four stroke thinking, while you are tuning a two stroke, will get you into real trouble.

Use the Google "site:" option to search this thread to understand more about two stroke tuning.

https://www.kiwibiker.co.nz/forums/showthread.php/86554-ESE-s-works-engine-tuner (https://www.kiwibiker.co.nz/forums/showthread.php/86554-ESE-s-works-engine-tuner/page2000)

.

[MDJGyurgeiz]

.

https://www.kiwibiker.co.nz/forums/showthread.php/86554-ESE-s-works-engine-tuner/page2000
347052
Start here with 18 - 20 deg.

Massive advance that works on full throttle will promote detonation and engine damage on semi closed throttle.

Beware..... four stroke thinking, while you are tuning a two stroke, will get you into real trouble.

Use the Google "site:" option to search this thread to understand more about two stroke tuning.

https://www.kiwibiker.co.nz/forums/showthread.php/86554-ESE-s-works-engine-tuner (https://www.kiwibiker.co.nz/forums/showthread.php/86554-ESE-s-works-engine-tuner/page2000)

.

finally after a long time and thanks to your help, I was able to start the bike

https://youtu.be/KRvxEpjPV10
https://youtu.be/CSclGFfqEno
https://youtu.be/YMk_75FsPNY

[TZ350]

.

To change the scale on the face of a gauge.

347061347060347059347058347062

Step 1: click on the gauge.
Step 2: click "Designer Mode"
Step 3: click "Dashboard Designer"
Step 4: click "Properties Dialog" and make your changes to the scale and limits.
Step 5: Un click "Designer Mode.

[TZ350]

I was able to start the bike

https://youtu.be/KRvxEpjPV10
https://youtu.be/CSclGFfqEno
https://youtu.be/YMk_75FsPNY



Great work. I am very impressed......:drinknsin

[MDJGyurgeiz]

Great work. I am very impressed......:drinknsin

I succeeded because I have passion and am skilled in the electronic field, but you have given me a good hand to solve my doubts and you have dedicated your time to help me. I am immensely grateful to you

[MDJGyurgeiz]

Great work. I am very impressed......:drinknsin

question. do you think it is possible to insert an egt probe and a knock sensor to the engine to manage, for example, a leaning of the carburetion with the egt probe and the ignition advance table with the knock sensor?

[TZ350]

question. do you think it is possible to insert an egt probe and a knock sensor to the engine to manage, for example, a leaning of the carburetion with the egt probe and the ignition advance table with the knock sensor?

A knock sensor would be very useful if it retarded the ignition and/or enriched the mixture to quell detonation. I use one to retard the ignition when necessary.

A temperature sensor for sensing detonation or lean mixture is problematic in a two stroke. Detonation increases head temperature and reduces exhaust gas temperature. So when the exhaust temperature goes down. The question is, is it rich and needs further leaning out or is it really lean and detonating? The sensor would not know.

If you are brave you could use an Arduino (Nano would be my pick) to watch the head temp, exhaust temp and look for detonation. The Arduino could interpolate the results and adjust the mixture. Maybe by adding an offset to the Baro reading.

[TZ350]

.
Another step forward in getting a clean swap between small and big injectors with staged injection.


Initially with my changes to the code in "STAGING_MODE_AUTO" I could get it to swap. But was troubled with fluttering between the big and small injectors. Now I have some hysterias built in. Once the small injector has swapped over to the big one, in my code the fuel demand has to drop 10% before it swaps back.


I also had to add some adjustment to "inj_openTime_us" to balance the initial fuel dribble from the big injector with the small one so as they both delivered the same amount of fuel for a given pulse width.
.
347316

Small injector, Green line. Large injectors Yellow line.

The AFR lower Yellow line stays consistent for the same VE as the injectors swap. I had to adjust the "inj_openTime_us" to get a smooth transition.

347317

[TZ350]

.
My final code changes in the Speeduino main block of code that have the staged injection swapping cleanly from the primary injector to the secondary injectors with some built in hysteresis to inhibit fluttering between injectors.


//.............................................. Rob2
//uint32_t inj_openThistime_uS = 0;
//uint32_t PWlastTime = 0; //
//uint32_t PWthisTime = 0; // these need to be coppied into the definition area near the top of the page.


else if(configPage10.stagingMode == STAGING_MODE_AUTO)
{
currentStatus.PW1 = tempPW1;
if(tempPW1 > pwLimit ){PWlastTime = (tempPW1 * 0.9); PWthisTime = tempPW1;}// tempPW3 has to drop 10% to avoid unwanted short term jumping between two states (ie; small and big injectors) - hysteresis
//If automatic mode, the primary injectors are used all the way up to their limit (Configured by the pulsewidth limit setting)
//If they exceed their limit, the whole duty is passed onto the secondaries

if(PWthisTime > PWlastTime)
{
uint32_t extraPW = tempPW1; //The open time must be added below because tempPW1 does not include an open time.
currentStatus.PW1 = 0; // Completely turn off the primary injector.
currentStatus.PW3 = ((extraPW * staged_req_fuel_mult_sec) / staged_req_fuel_mult_pri); //Swap to the secondary injector and convert the fuel amount from primary injector scaling to secondary
inj_openThistime_uS = (inj_opentime_uS * 0.5);
currentStatus.PW3 += inj_openThistime_uS; // adjustment of inj_opentime_uS because there are two injectors fired together as one logical injector. This value may need adjusting.
PWthisTime = tempPW1;
}
else { currentStatus.PW3 = 0; PWthisTime = 0;} //If tempPW1 and tempPW3 < pwLImit it means that the entire fuel load can be handled by the primaries. Simply set the secondaries to 0
}
//................................................ Rob2

[jbiplane]

May be a bit offtopic?

I have other simple idea - want to share. May be someone will implement or help me to implement. I believe it is possible create simplest ECU which will calculate injection duration and ignition angle based only on rpm. No other sensors. Small tune can be performed by mechanical fuel pressure regulator. In future it can be separate board which regulate fuel pressure by simple circuit board (themperature and barometric corrections). I would be happy to find someone who can help me in electronics to do this joint project from the scratch and will do all mechanical stuff and testing.

Possible make fuel table by throttle angle (instead of rpm) if it simpler...

In this case processor should perform 1000th times fewer calculations will be cheap and reliable.

[JanBros]

simpelest and probably cheapest programmable ignition based on only rpm, only requires an Arduino Nano, an IBGT and a simple voltage regulator :

http://loutrel.org/aeduino.php (sorry, only in french :innocent:)

someone who knows arduino-stuff can probably easily do the same for injection.

[TZ350]

simpelest and probably cheapest programmable ignition based on only rpm, only requires an Arduino Nano, an IBGT and a simple voltage regulator : someone who knows arduino-stuff can probably easily do the same for injection.

I would not want to do it but, yes probably quite simple(ish) using the Alpha-N fuel mapping topology. The open source, freely available Speeduino firmware for the Arduino Mega could even be stripped down to basics for the Nano. They both have the same processor speed. The Nano just has less IO. A disadvantage would be that the Speeduino runs inductive ignition coil(s) instead of CDI but there would be a way around that too.

[TZ350]

.
After trying my code changes I discovered a flaw .........

I have been trying to get the staged injection to cut cleanly between the primary (Inj - 1) and the secondary injectors (Inj - 2).

This it can do and after building in some hysterias it does this without fluttering between primary and secondary injectors.

And I could get it to run nicely at the low end on the primary injector and ok at the top end on the secondary injectors.

But the problem is I could not get it to swap cleanly from the Inj-1 region to the Inj-2 area.

By changing the allowed "Duty Cycle" I could force Inj-1 or Inj-2 to try and run in the region in the middle.

But no amount of fuel mapping trickery. Fiddling with the logical injector sizing or alteration of the injector dead times helped.
.
347490
.
As it turns out the issue seems to be in the original Speeduino code. The code may be right for when the staged injection just adds the secondary's to the primaries as needed but does not work when its a clean cut over.

I had to change the way the ratio of secondary to primary is calculated. A quick run on the Dyno looked promising but we will know for sure next week when I have time to test it properly.
.
At any point of the Alpha-N map Inj-1 80cc and Inj-2 at 320cc will have different pulse widths for the same VE number but both should deliver the same fuel volume. Its obvious but it took me a while to take the step of looking at this.

My primary Inj-1 is 80cc and secondary Inj-2 is 320cc . A ratio of 4:1.

In the data log the Red line is VE. The Green line is Inj-1 and yellow for Inj-2. Previously the PW ratio in the graph was 3:1 but with the code changes the ratio now looks to be 4:1, the same as the injectors themselves.

What I am unsure about is the dead time. I am not sure if the dead time is included in the PW and PW3 as displayed or if the PW is logged before the dead time is added or after. I would like to know because it would make for a better idea of the actual ratio of fuel delivered.
.
347492
.
Basically either side of the point where the injectors swap over and the map number is the same the difference in pulse widths should be in a similar ratio to the injector ratio. In my case 4:1 because Inj-1 = 80cc and Inj-2 = 320cc.

[jbiplane]

simpelest and probably cheapest programmable ignition based on only rpm..
I saw such simple - digital ignition and injection based on rpm
https://www.homebuiltairplanes.com/forums/threads/forum-made-efi.30920/page-15#post-559277

[jbiplane]

Speeduino news
2 injection+ 2ignition channels 47x25mm board
https://speeduino.com/forum/download/file.php?id=7447

A bit bigger board
low-cost and mini-sized ECU for speeduino firmware (45 x 55 mm)
4 high impedance injectors
4 active ignition coils or 1 passive ignition coil with separate ground plane
1 crankshaft sensor with VR-conditioner
1 hall input for camshaft sensor
6 analog inputs: CLT, IAT, TPS, BAT, LMM, MAP
1 optional baro sensor MPXH6400A on bottom side
1 digital input for disco swaggering (aka. launch control)
Onboard Wideband Lambda Controller with Bosch CJ125 (LSU 4.9)
1 fuel pump out
1 idle valve out
1 tacho out
1 boost out
USB programming interface with CP2104

https://speeduino.com/forum/download/file.php?id=6826

[jbiplane]

Finally my company ECU = digital ignition + injection integrating to throttle body will look like on this picture.
We will make it rather big 70x70mm placing most of components on one side to fit ECU+throttle body in $120 FOB
2 ignition and 2 injection channel allow install on most of 1...4 cylnder engines. MAP, IAT TPS onboard, so very
few and only required wires will go out which will simplify instalation. We trying make PWM modulation of
fuel pump to reduce electricity consumption.
https://www.homebuiltairplanes.com/forums/attachments/%D0%9F%D0%BB%D0%B0%D1%82%D0%B0-%D0%B2-%D0%B7%D0%B0%D1%81%D0%BB%D0%BE%D0%BD%D0%BA%D1%83-jpg.103283/
Wiring:
CLT - cooling liquid temperature, can be used for air cooling as well
Al + CHR thermocouple type k to be fitter under spark of one cylinder which is in worst conditions
O2 - narrow band lambda
2 ignition channels
2 injections channels
2 wires for Crankshaft sensor
2 wires for Camshaft sensor
R-in programmable to fill any resistive sensor, can be switched to V sensing, says for exhaust temperature control using AD595
IAC - idle control
FP and FPM fuel pump and fuel pump modulation to minimize electricity consumption on small engines

Possible use free version of EFI Tuner Studio with our hardware
Finally we can make Speeduino version of our ECU to allow users make any programming easily

[jbiplane]

Added magnetic sensor instead of external TPS

https://www.youtube.com/watch?v=_RlguxPV8TQ

[Dexxa]

Hi,
I appreciated a lot all the work done by TZ350, very very great work.
Since that I'm trying to EFI an Aprilia RS 125, i'm here to collaborate with this thread and with all the people that are reading this thread.

I have a Microsquirt module to control the efi on my Aprilia, I know very well this ECU and her software, it's very similar to Speeduino since it's using the same software (tuning studio).

I've done some project with this ecu on a 4 stroke, but now i want to apply it on a 2S.

I have a question: I don't want to use 2 injectors with staged injection because im afraid about the switching of them at 10000 rpm.
So this is what i want to do.

PWM FUEL PUMP CONTROL, this in order to adjust the fuel pressure vs RPM or LOAD.

Pwm fuel control could change the pressure in order to have more o less flow throw the injector, so you can have a low pressure ad idling (2 bar I think would be ok) and than 4-5 bar at max power. With this method you can use only one injector with a solid idling and a good flow at high rpm.

The other benefits are:
Removing the mechanical fuel pressure regulator
You have no more fuel return to the tank, so your fuel is more cooler.

All is sounding great but the bad news is that Microsquirt (even Speeduino) can not control the fuel pump with a PWM control.

I'm not very good at programming a new part of the firmware but I could help if someone is interested in doing this.

[JanBros]

Hi,

The other benefits are:
Removing the mechanical fuel pressure regulator
You have no more fuel return to the tank, so your fuel is more cooler.


this makes no sence ???
how are you gonna create preasure when the pump has nothing to push against ?
or if you gonna block off the valve, where is the excess fuel supposed to go to ?

[Dexxa]

The pwm control of the fuel pump allow you to don't have the mechanical regulator because the fuel pump is controlled by a duty cycle from arduino or something else.
There are a lot of impulses between 0 and 12 volt to the pump. The more their frequency is high the more the pump will go faster.
So if the pressure gauge indicates 4 bar and in your imposed table you want 5 bar at that rpm, the duty cycle of the pump will be increased to have 5 bar on the fuel line.
This in a closed loop control.

Did you understand? ;)

[JanBros]

yes, I understand pwm.

so you are gonna close the return ? injectors use way less fuel than the pump supplies, meaning the pump will turn for a brief moment untill the wanted preasure is obtained. than it needs to stop to prevent overpreasure. preasure drops fast because of fuel injection, pump needs to be turned on again etc...

[pete376403]

https://www.vaporworx.com/resources/fds/8-how-pwm-returnless-fuel-systems-work/

[husaberg]

https://www.vaporworx.com/resources/fds/8-how-pwm-returnless-fuel-systems-work/
Neat explanation.

[ceci]

I am the least expert in the field of chromatography (in this forum there are real experts in this field "they participated in the creation of the best 2S injection system" DITECH).
But I think the pressure with the drop size achieved in the Spray is very important and in turn the drop size with the evaporation time

[ceci]

What I'm trying to say is:
5 bar = spray droplet size ≤ injector orifice diameter Ø
2 bar = spray droplet size > injector orifice diameter Ø.
Although when idling there is more time for evaporation, the larger droplet size (the smaller the droplet size, the greater number of droplets and more dispersed covering a greater area) would influence so that a uniform mixture cannot be created, creating a stratified scavenging. which may have the poorest fuel zone closer to the spark plug

[TZ350]

Ok. So its drag racing, but a two stroke drag racer.

Great two stroke sound at WOT. And also if you pay attention to the technical talk and trip around the work shop you will pickup that its fuel injected.

<iframe width="560" height="315" src="https://www.youtube.com/embed/eRItqiCFuQE" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen=""></iframe>

The fastest 2T drag bike in the world has fuel injection.

[HankStarr]

Hi..... The SR50 and Scarabeo 50 early forms were both Fuel infused and cut. Placing a motor in a Zuma would require the ECU a dashboard, radiators and a great deal of work. The actual hurries are incredible and in case it was the main one to play with most huge tuners change them over to carbs. I'm keeping mine infused on the grounds that its amusing to play with. It's shown me a ton about CVT tuning. Yet, when you have a sputter or where it's excessively rich or incline there are no planes to trade out. Stay with the carb you can fabricate a significantly more impressive machine.

[ceci]

Hi,
I appreciated a lot all the work done by TZ350, very very great work.
Since that I'm trying to EFI an Aprilia RS 125, i'm here to collaborate with this thread and with all the people that are reading this thread.

I have a Microsquirt module to control the efi on my Aprilia, I know very well this ECU and her software, it's very similar to Speeduino since it's using the same software (tuning studio).

I've done some project with this ecu on a 4 stroke, but now i want to apply it on a 2S.

I have a question: I don't want to use 2 injectors with staged injection because im afraid about the switching of them at 10000 rpm.
So this is what i want to do.

PWM FUEL PUMP CONTROL, this in order to adjust the fuel pressure vs RPM or LOAD.

Pwm fuel control could change the pressure in order to have more o less flow throw the injector, so you can have a low pressure ad idling (2 bar I think would be ok) and than 4-5 bar at max power. With this method you can use only one injector with a solid idling and a good flow at high rpm.

The other benefits are:
Removing the mechanical fuel pressure regulator
You have no more fuel return to the tank, so your fuel is more cooler.

All is sounding great but the bad news is that Microsquirt (even Speeduino) can not control the fuel pump with a PWM control.

I'm not very good at programming a new part of the firmware but I could help if someone is interested in doing this.


Hi Dexxa.
Are you trying to adapt the injection of the RSW 500 to an RS 125?
I suspect that you have worked this type of injection before

[ceci]

The pwm control of the fuel pump allow you to don't have the mechanical regulator because the fuel pump is controlled by a duty cycle from arduino or something else.
There are a lot of impulses between 0 and 12 volt to the pump. The more their frequency is high the more the pump will go faster.
So if the pressure gauge indicates 4 bar and in your imposed table you want 5 bar at that rpm, the duty cycle of the pump will be increased to have 5 bar on the fuel line.
This in a closed loop control.

Did you understand? ;)

I had doubts if I could be wrong when affirming that you have worked on the RSW500, but after analyzing other mechanical injection systems (in the way you indicated to work the pump I thought it was mechanical and not electronic).
Analyze the injection system of MRD, which uses a mechanical pump (pinion) but the injection of this system is continuous injection like that of K-jetronic (it uses an electronic pump "with constant pressure" and a pressure regulator to supply the desired flow) and not sequential injection as you describe.
Another piece of information that makes me think that I am not mistaken is that you indicate the use of a pressure gauge to control the injection pressure and that I have only seen in the RSW500, it is part E99061 in the throttle body.
https://www.kiwibiker.co.nz/forums/attachment.php?attachmentid=349626&d=1631529700

[RD400C]

Hello.. I just sat and read all 24 pages and it is a fascinating read!

I am in Ottawa, Canada and am about to set out on my own two-stroke efi project.

I have a 1976 RD400C completely stock.... just the way I like them. They arent horse power machines in stock trim but they are fantastic for cruising around town. Id never mess with intake or exhausts or porting because I love the torque right in the lower and mid rpms. Yamaha says 40hp stock but I suspect it's really 35 or so.

At this point I have purchased a Microsquirt because it comes in a small sealed unit with water tight amp seal for the wiring harness. I plan on not only controlling fuel but also ign curve. I'll need more power for fuel pump so I have upgraded to a VAPE ignition system. It produces 150 watts and now the ignition is CDI with a VR sensor. There is no ECU in this setup... I have it set at a static 20 degrees (2.3mm BTDC on this bike) The microsquirt will be able to take direct control using the VR as a signal in and also will also directly control the ignitor circuit in the CDI/coil. Microsquirt can log two EGT channels so I have also purchased those. I will also have a CHT sensor under one of the plugs to be used for start up enrichment and also logging purposes. I have a heated Wideband O2 sensor that I will use as well but as we know... the readings may not be an accurate picture of what's happening in the combustion chamber. An intake air temp sensor as well.

First thing in my plan is build a hydraulic gear pump dyno with a needle valve to control rpm. I have everything needed and will soon start welding a frame to hold it all together. The plan is make it so it all fits between the swing arm rails and the stock chain will drive the gear pump. The pump will be able to rotate a little... there will be a 12inch arm attached to the pump and it will press a load sensor that will also feed into the microsquirt and be logged. I will trick it to think it is something else but it outputs a 0-5v signal so thats what matters. I have two stepper motors... one will manage the needle valve and the other, the throttle cable. I have written but not tested code for the Arduino to go to a specific throttle % and hold it there while the one controlling the needle valve will use engine rpm to open/close the valve in small increments to keep the engine at desired RPM.

The plan.... set arduino to hold certain cell value on fuel table and matching cell in spark advance table then I will monitor EGT and O2 and torque produced as I add/remove fuel to achieve good power but safe power then I will add/remove spark advance as it plots the torque curve so I can see where it makes the most power then remove a couple degrees to leave it on the safe side of things. I am thinking about installing a knock sensor to prevent me from destroying the motor during ignition tuning but its an air cooled motor covered in fins and there is no where to really put it... plus I dont want to weld anything to the engine. However... it is my understanding that high octane pump fuel will deter knock and allow me to advance spark enough where I will see torque production drop off a little without actually knocking. Then I can chose the part of the curve before the peak and set the timing there. Lots to learn and loads will go wrong but im hoping to see just how strong a stock setup can be. Also hoping to tune it to be a smooth ride with no surging or bucking etc. You can never tune this out of an RD400... i have mine pretty damn good but i have a really rich low rpm do achieve it. smokes like a chimney at idle.

I am planning on Alpha-N as I ran a vacuum signal test run the bike with the carbs and it looked like a jagged saw tooth. Completely unusable. I used the oil injection nipples on the carbs to get the vacuum so.... maybe when the R6 throttle bodies are on the bike that have actual vacuum ports it will perform better. Testing will tell. I also have ITB control method which is method that uses Alpha-N and switches over to VE assuming the signal is more consistent in higher rpms.

I love what you are doing here... just wanted to introduce myself and share what I have planned. I'll pop in from time to time and share anything I have learned :)
-John

[ceci]

Hello.. I just sat and read all 24 pages and it is a fascinating read!

I am in Ottawa, Canada and am about to set out on my own two-stroke efi project.

I have a 1976 RD400C completely stock.... just the way I like them. They arent horse power machines in stock trim but they are fantastic for cruising around town. Id never mess with intake or exhausts or porting because I love the torque right in the lower and mid rpms. Yamaha says 40hp stock but I suspect it's really 35 or so.

At this point I have purchased a Microsquirt because it comes in a small sealed unit with water tight amp seal for the wiring harness. I plan on not only controlling fuel but also ign curve. I'll need more power for fuel pump so I have upgraded to a VAPE ignition system. It produces 150 watts and now the ignition is CDI with a VR sensor. There is no ECU in this setup... I have it set at a static 20 degrees (2.3mm BTDC on this bike) The microsquirt will be able to take direct control using the VR as a signal in and also will also directly control the ignitor circuit in the CDI/coil. Microsquirt can log two EGT channels so I have also purchased those. I will also have a CHT sensor under one of the plugs to be used for start up enrichment and also logging purposes. I have a heated Wideband O2 sensor that I will use as well but as we know... the readings may not be an accurate picture of what's happening in the combustion chamber. An intake air temp sensor as well.

First thing in my plan is build a hydraulic gear pump dyno with a needle valve to control rpm. I have everything needed and will soon start welding a frame to hold it all together. The plan is make it so it all fits between the swing arm rails and the stock chain will drive the gear pump. The pump will be able to rotate a little... there will be a 12inch arm attached to the pump and it will press a load sensor that will also feed into the microsquirt and be logged. I will trick it to think it is something else but it outputs a 0-5v signal so thats what matters. I have two stepper motors... one will manage the needle valve and the other, the throttle cable. I have written but not tested code for the Arduino to go to a specific throttle % and hold it there while the one controlling the needle valve will use engine rpm to open/close the valve in small increments to keep the engine at desired RPM.

The plan.... set arduino to hold certain cell value on fuel table and matching cell in spark advance table then I will monitor EGT and O2 and torque produced as I add/remove fuel to achieve good power but safe power then I will add/remove spark advance as it plots the torque curve so I can see where it makes the most power then remove a couple degrees to leave it on the safe side of things. I am thinking about installing a knock sensor to prevent me from destroying the motor during ignition tuning but its an air cooled motor covered in fins and there is no where to really put it... plus I dont want to weld anything to the engine. However... it is my understanding that high octane pump fuel will deter knock and allow me to advance spark enough where I will see torque production drop off a little without actually knocking. Then I can chose the part of the curve before the peak and set the timing there. Lots to learn and loads will go wrong but im hoping to see just how strong a stock setup can be. Also hoping to tune it to be a smooth ride with no surging or bucking etc. You can never tune this out of an RD400... i have mine pretty damn good but i have a really rich low rpm do achieve it. smokes like a chimney at idle.

I am planning on Alpha-N as I ran a vacuum signal test run the bike with the carbs and it looked like a jagged saw tooth. Completely unusable. I used the oil injection nipples on the carbs to get the vacuum so.... maybe when the R6 throttle bodies are on the bike that have actual vacuum ports it will perform better. Testing will tell. I also have ITB control method which is method that uses Alpha-N and switches over to VE assuming the signal is more consistent in higher rpms.

I love what you are doing here... just wanted to introduce myself and share what I have planned. I'll pop in from time to time and share anything I have learned :)
-John
Hi John. Welcome, it is a good approach, I wish you success

[lohring]

Welcome. My only criticism is of your test method. An inertial dyno is not only easier to build; it is much easier to run. We typically made up to 50 individual test runs in an afternoon. Each run generates a full power curve. This lets you test a lot of changes in a short time. We ran one test series on a water brake dyno. It took all afternoon and generated only one partial power curve. See http://performancetrends.com/tdkmotorsports/index.html for more information. We used the Performance Trends dyno data logger, but there are others.

Lohring Miller

[RD400C]

Welcome. My only criticism is of your test method. An inertial dyno is not only easier to build; it is much easier to run. We typically made up to 50 individual test runs in an afternoon. Each run generates a full power curve. This lets you test a lot of changes in a short time. We ran one test series on a water brake dyno. It took all afternoon and generated only one partial power curve. See http://performancetrends.com/tdkmotorsports/index.html for more information. We used the Performance Trends dyno data logger, but there are others.

Lohring Miller

Thanks for your feedback... these forums are invaluable for people... I have access to so many people's experience and feedback.

I looked into building a DIY home inertial dyno but gathering the parts it needs and building it seemed out of my scope. The gear pump dyno is nice and compact and should fit in the swing arm and hook right up to the chain. I think the biggest benefit, for me, of this type of dyno is the ability to plot a torque curve for each cell while the computer holds a steady rpm. Im hoping this will allow me to optimize each and every cell of the fuel and spark tables. I'm not certain but I dont think an inertia dyno will allow me to have that sort of accuracy for a specific throttle/rpm setting. I am terrified of destroying my motor with too much advance or too little fuel... im hoping this method will give me the accuracy to avoid that.

I was thinking if I wanted an rpm sweep torque curve I could set the needle valve to mimic what the bike would feel on the hwy and then do a 2000 to 7500 run and plot the torque and get something meaningful. Im not really concerned with accurate numbers... just the ability to see dips in the curve.

Thats the plan... fingers crossed!

I'm also going the be recording everything I do and posting it on youtube so people can watch and offer advice and ideas to me. Once I get things started I'll come here and post the link for anyone that's interested.

Thanks for the welcome :)
-john

[ceci]

Hello.
The issue of whether the TBI or the TPI is better is not new, there are many threads in forums.
Could he be winning the TBI? , the FI in the RSW500 got more HP.
This can also happen in MX 125?

https://www.vitalmx.com/features/SPOTTED-2023-Husqvarna-TC-125-With-New-Fuel-Injection,7921

[ceci]

It's just a divagation of mine, since I discovered the work of this German boy on YouTube, I can't stop thinking about repeating history as it happened with others before, in which manufacturers take ideas without rewarding their creators.

Minarelli in two consecutive years has been presenting the EFI in the two stroke engine.
The first year I only presented it in the 300cc and with injection trans the reed valve.
https://enduro21.com/images/2021/november-2021/minarelli-2t-fuel-injected-engine/motori_minarelli_300_2t_fi_eicma2021__0341.jpg
https://enduro21.com/images/2021/november-2021/minarelli-2t-fuel-injected-engine/motori_minarelli_300_2t_fi_eicma2021__0354.jpg

The second year he presented the 300cc with injection again but this time in the throttle body, he also presented the 50cc with injection in the style of the 300cc of the previous year

https://www.dueruote.it/content/dam/dueruote/it/news/moto-scooter/2022/11/11/minarelli-presenta-il-nuovo-300cc-2t-ad-iniezione/gallery/rsmall/MM300%20-%20300cc%202T%20FI%20Euro5%208.jpg
https://motoriminarelli.it/wp-content/uploads/2022/12/IQJB4858-e1669980328502-1024x777.jpg
https://motoriminarelli.it/wp-content/uploads/2022/12/GEBX4885-1024x1024.jpg

It is clear that Minarelli is interested in bringing EFI to the two stroke engine, if we follow the trajectory of the 300cc and apply it to the 50cc, we can expect to see at EICMA 2023 a 50cc with injection in the throttle body, also if we take into account Details such as those of the power valve servomotor, in the 50cc it would be in the throttle body


https://www.youtube.com/watch?v=7CYgLiBbQ4A


https://www.youtube.com/watch?v=JWTqWEztjzI&t=278s

[ceci]

Another option is if Minarelli or any other manufacturer buys it.
As it happened to Bob Horvath who bought his idea, the manufacturer of Tillotson carburetors.

https://tillotson.ie/wp-content/uploads/2015/07/IMG_2425sijpgv1348152908.jpeg

Another option is for the Chinese manufacturer CPI to implement this mechanism in its engines.

https://www.apriliaforum.com/forums/attachment.php?attachmentid=37181&d=1326387883

https://www.apriliaforum.com/forums/attachment.php?attachmentid=37182&d=1326387883

EICMA 2005

[ceci]

It looks more compact and simple than the one used by Minarelli on the 4S.

https://youtube.com/shorts/mz8Kyo2FE-U?feature=share




https://sbd.etailercdn.de/media/pimg/CG/thumbs/540340_4452178.jpg

[JB21]

Hi guys,

Is here still someone reading? 😁

I am currently getting everything together to speeduino my ktm sx 125 23 tbi.

I want to use the original crank case pressure sensor. So is there anything already "New" to make the Firmware better?

I would like to use everything onboard, so no external map Filter.

It uses a small range map sensor for ccp.

Both injectors are the same https://cars245.com/en/item/bosch-0280158416-injector/
One infront the butterfly and one behind.

I am currently tuning the OEM ecu's, due the lag of Power they come with (34hp).
But what i noticed and still cant get rid of it (due the not existing informations about this m4c ecu) is that if the piston wears, the ccp drops and the power drops drastical.

So i want to use the speeduino to be able to "ignore" the wear of the piston some how.

Maybe we can keep this thread up to life.
Everything that could help is really appreciated!

Have a good time guys!

[JB21]

Following this awesome build!

I am currently trying to get running a Speeduino based ecu on my ktm sx 125 23 TBI.

I want to get rid of that way too restricted oem ECU.

[TZ350]

.
Hi JB2

I stopped working on my own project when the racing class it was for got too hectic for me to continue competing in. I never felt I had got my bike running perfectly but hopefully you will do better with yours.

I would love to follow your project here.

Have you seen the Speeduino NO2C https://bossgarage.eu/en-ww/products/speeduino-no2c

It is a two channel system and physically smaller than the common four channel version that overhangs the MEGA 2560 mother board. The NO2C is the same size as the MEGA 2560 board itself. It would be my choice for a single cylinder bike.

Speeduino Manual:- https://speeduino.com/Speeduino_manual.pdf

[JB21]

Hi TZ,

I just read all you posts about the freaking journy of converting a two stroke to efi.
Just crazy how many hours of research you just spent!

Now to my project- I want to convert the original ktm sx 125 23 TBI ecu to a good ecu!

So I want to throw that oem part away and replace it by an own build Speeduino powered pcb that fits the oem connector without changing anything on the harness.

The bike comes stock- like you must have heared already - a ccps.

So I am into tuning the oem ecu's, but they keep restricting power always when piston wear accures. So after 4-5 hours runtime the bike keeps loosing power because the ecu restricts it.
And I think the only sensor doing that restriction is the ccps- minimal changes in the value and the ecu takes ign low.

The problem is, that I just don't find the table, where this happens- maybe it is an algorythm in the code itself...I dont know, because I am just working in the Binary file without any definitions.

SO to get rid of that, I want to use the speeduino to get good and overall consistent results.

The bike comes stock with 2 injectors, one infront, one behind the butterfly. Both are equal: https://cars245.com/en/item/bosch-0280158416-injector/

What I cant get over- the bike doesnt want to hit the redline : at around 12500 that thing just keeps barking, so it doesnt want to rev higher- like its just getting way too much fuel- or not correctly timed.

I can provide you some tables of Start of injection and so on, if you are generally interested in these.

Also the ccps and ambient pressure sensors are way different to yours- they only have a small window of reading. (https://www.autoersatzteile.de/7890684-vdo)




Thank you so much for all the work your putting in.

[T0M]

Hey, I managed to get my 2 stroke Vespa to run very smoothly with fuel injection. It runs nice at all throttle positions.
At some time - I switched from speeduino to MaxxECU Mini.

A dyno run and the used parts are shown in this video:


https://youtu.be/dOfxx0HjXEw

[TZ350]

Hey, I managed to get my 2 stroke Vespa to run very smoothly with fuel injection. It runs nice at all throttle positions.
At some time - I switched from speeduino to MaxxECU Mini.

A full power dyno run and blipping the engine afterwards as seen in the video is relatively easy. Alpha-N topology is good enough for that.

The real test is blitzing at full noise up to a slow corner, breaking hard, feathering your way around to the apex, then pulling away smoothly and getting back on the pipe again.

Basically, its on the pipe, throttle off for the corner collapses the wave action in the pipe, rpm drops and then throttling back on to try and re establish the wave action back in the pipe.

This has been my problem and I have not really got that smoothly pulling away from the apex working very well myself. Tuned 125 with competition pipe.

Before I rush out and buy a MaxxCPU to replace my Speeduino. If you can I would be very interested in a video clip of how well the MaxxCPU handles fast paced riding and corners for you.

A PDF of the MaxxCPU manual would be very welcome too.

I admire your work.

[T0M]

This has been my problem and I have not really got that smoothly pulling away from the apex working very well myself. Tuned 125 with competition pipe.

Before I rush out and buy a MaxxCPU to replace my Speeduino. If you can I would be very interested in a video clip of how well the MaxxCPU handles fast paced riding and corners for you.

A PDF of the MaxxCPU manual would be very welcome too.

I admire your work.

I drive only on the street, and there is not one situation in which the fuel injection does worse than the carb. Even, if I pull the trottrle to full opening from very low rpm. However, I can try to simulate your situation next time, when I will be on the dyno. It has a mode, wher it simulates real driving by using speed and drag coefficent.

I guess the MaxxECU will not resolve your problem. My engine worked also well with the speeduion, however I was sick of replacing damaged circuit bords. The last one I destroyed by my self, as I tried to add a log-to-SD-card functionality . This was the time I switched to MaxxECU.

It is not clear to me why, but the VE table gained from the dyno runs did not worked well on the street. I need to log the fuel-injection and AFR during real riding in order to be able to create a good working VE table.

The MaxxECU seems to cost more than the speeduion, but the software is allrady included (similar to tuning studio) and you get a very nice wiring harness with a good connector. The manual is online:https://www.maxxecu.com/webhelp/index.html

The wirings (and other stuff) is here: https://www.maxxecu.com/downloads
For the MaxxECU Mini, which I use it is: https://www.maxxecu.com/files/Documentation/Wirings/MaxxECU%20MINI%20(REV10+)%20-%20Wiring-en.pdf

The MTune software is what is tuner studio for speeduino. The free download is here (incl. manual): https://www.maxxecu.com/mtune

I attached my tune file: 354344

The VE Table has some quite high values for low rpm at full throttle.

https://www.kiwibiker.co.nz/forums/attachment.php?attachmentid=354345&d=1707404456


The lambda table is quite lean for low rpm and low throttle openings. The carb was alrady showing this behaviour.

https://www.kiwibiker.co.nz/forums/attachment.php?attachmentid=354346&d=1707404456

Hopefully, I did manage to include this pictures correctly within the text.
354344
354345
354346

EDITH attached my last TunerStudio project: 354349

[TZ350]

.
Thank you, I will read your post carefully. Its very interesting.