ESE's works engine tuner

[Peter1962]

That is interesting, KTM carburetor engine makes significantly more power than the EFI version. My experience with two stroke EFI is that you can easily match or better a carb engine for maximum power and over rev. With access to the fueling maps EFI is so easy.

On the ktmtalk forum there are a lot of stories on problems with the SX125 with failing crankshafts since model 2022. Some think that KTM did not properly solved the problem, but in stead limited the Maximum RPM on the TBI 125 SX model 2023 to avoid catastrophic failures... Poor choice. https://ktmtalk.com/threads/2023-125-sx-crank.569456/

The ktm 125 model 2022 had more HP than the model 2023.

But the 250 SX TBI 2023 has more HP than the previous model, so I do not think that it depends on the injection but rather to the choices that the ktm engineers implemented concerning the mapping.

[husaberg]

You have to love the irony that nearly 15 years ago ktm had injection on their 2 stroke gp bikes that worked on overrun to avoid crankshaft failures......

The most important thing, however, is that the injection enables things that are not feasible with carburettors alone under certain load conditions, such as overrun or in the partial load range. Knock sensors in the cylinders can detect and compensate for minor inaccuracies in the spraying. In addition, it is possible to drive much leaner in the partial load range and also to lean individual load and speed ranges in a targeted manner, which improves drivability and torque curve

[katinas]

.

I am not sure why 2S EFI should find it harder to meet Euro standards than a carb engine.

Flettner who has way more real world practical riding experience with 2S EFI than me. Reports that with transfer port injection (TPI) there is a noticeable fuel saving which would suggest there is less fuel wasted through charge short circuiting. That has to mean a cleaner engine exhaust.

KTM copied Flettners TPI idea then went to reed port injection. Injecting before the reed valves looks like a poor mans carburetor to me. With all the problems of a crankcase full of fuel mixture, some of which is lost through short circuiting.

Again from my 2S EFI experience. A 125 is much harder to map properly than a 300. I suspect, maybe KTM has found mapping a 125 for the vastly more dynamic on pipe/off pipe conditions a challenge. There is a way but it is much more tricky than working with a 300.

Factory set just two fuel mapping programs and looks like one lean, other even more leaner.
The only change that we done with EFI engine, new head insert with corrected squish to 0.8 mm ( 1.2 mm from factory with 12 cc volume) and 12 angle 3mm piston dome.
This add 1.2 hp.
Crankcase pressure sensor connected directly to C port through 16 mm pipe.

[TZ350]

Crankcase pressure sensor connected directly to C port through 16 mm pipe.

The way the KTM's EFI software makes use of the pressure readings interests me. I would love to know more about that.

[koenich]

Someone on VitalMX posted the below info...

There is a series of “tables” that are related to each other, mathematically, that effect each other, and adapt to each other that related to ignition timing, fuel delivery and pv.
It is all based off the TPS and any changes to the engine (by use, intentional mods such as external TPS adj, wear and tear etc), and the tables attempt to compensate for these changes automatically, they hold their value, and sometimes for the worse.
Simply put, if you log 25 hrs on your top end, at some point during those hours, the wear and tear has effected the cps and changed the pressure under the piston, this has effected the way your ride, it made more heat, all the while the tables morf to attempt to compensate what it can to get it back to what it considers it’s zero point.
The tables shift and hold onto that value, it has “adaptive learned”.
You put in a top end and leave all as is, the ecu may still consider it’s held values it learned..this is why it is so important to clear the adaptive shift and get the engine back to its electronic zero so it’s baseline is restored.
This process to me is more important to the pv reset or anything else at this point.
The changes it morfs to can never go to an extreme and eventually it hits a wall and can’t move any more.
This is where if all those negative engine attributes stay held, the performance gets worse and worse.
The ex) of TPS adj has now changed my mind about how it is done.
If the TPS is adj, you are shifting the table of timing, fuel and pv, they are connected like a chain.
In very small movements, it will only effect the tables slightly, however as the engine started to wear down, the morfing gets more pronounced and “held”, or remembered.
This is also true with the sensing capabilities of the engine, it has held values determined by the sensors.
If you go down the the line at 9am for mx practice and it’s 45 deg, go back to the truck and shut it off and leave it until 1pm and start your bike on the line, your motor held the 45 degrees and will run that way until it catches up with itself after it learns.
What I would recommend in this situation if it’s 1pm, start your bike at the truck and let it run and blip for 20 seconds and then shut it off…then go to the line.
Your sensors will absorb the new data that it’s warmer and adapt.
This is one of the tables that has an X-Y axis as in engine rpm and throttle opening.
All those cells have cross over points over certain rpm ranges that effect each other.
Move one area, it shifts the table to effect fuel, timing, pv, etc.
I’m just getting the hang of the cause and effect of this but I can tell you one thing.
As for the actual combustion process of a two stroke motor with a pv, this is super simple….same same.
Regarding the fuel delivery, pv, timing, tps, crank pressure, etc throw everything you consider the “same” as a carbed bike out the window.
They are now two different things and should be treated that way, I have much more to share but am trying to keep it super easy to say and think about how to say it.
Stuff the cases with epoxy and dont compensate for that increase presssure under the piston with the ecu? It will also increase the velocity, lean out your fuel delivery and you will explode you motor..it’s that sensitive, unreal.
More soon- sorry if this is wordy, I’m learning and sharing at the same time lol.
Clear the adapted learned changes from your bike all the time and especially when large scale engine changes have occurred like top ends and if large environmental changes happened like cold am motos and hotter pm motos.

Source: https://www.vitalmx.com/forums/Tech-...1408609?page=5

I don't know any better but this would be an extremly complex system w/o much benefit. My best guess is it senses load by a fast pressure transducer and has some sort load depending alpha-N mapping.

[TZ350]

.

Thanks.....

[_____]

Source: https://www.vitalmx.com/forums/Tech-...1408609?page=5

... My best guess is it senses load by a fast pressure transducer and has some sort load depending alpha-N mapping.

If I would need to bet my money, I would say the sensor in the boostport is a transducer that collects crankcase pressure in two or more points of engine rotation. When the sensor also has a temperature probe installed, you are able to calculate the filling of the crankcase with the ideal gas equation.
https://en.wikipedia.org/wiki/Ideal_gas_law
Because you have certain knowledge of t, p, V. The values for M and R might vary with AFR in the crankcase, but I think that is minor and might be adjustable with measuring twice.

So say they got that under control: When they know how much air mass (with fuel) is in the crankcase, it is possible to add the required fuel to it.

When you combine that with an Alpha-N map for acceleration and use the calculation of the air mass in the crankcase, you may either inject according to the ideal gas law, or reduce your acceleration quantity with a fudge factor coming from the gas law equation.


Regarding the MX guy:
I doubt they have self learning in that depth because it would not make any sense. However, I have not ridden the bike.

Regards
Chris

[dutchpower]

ecu 125 sx 2023

[Flettner]

The original YZ250TPI would run between 14 and 21% less fuel burnt than a standard, same spec, carburetor YZ.
Compared same days, same tracks, aulternating riders, over many events. Same top end power.
14% on the more open fast events, 21% was seen at slow low speed events.
One would have to conclude that that's a funtion of less fuel short circiting out the exhaust. Also partly because the engine would sort of be a stratified charge combustion at lower speeds with fuel only entering the B ports. A decient squish and toridal head seemed to help with this last ditch fuel / air mixing. Power is definitely more mellow down low in the rev range compared to the carburetor bike. Interestingly I could turn in a faster time on the TPI bike but the carb bike felt faster. Way more linner power delivery with less ' surprises'.

[SwePatrick]

I thought i´d share this with you guys:



I had some problems with the chain didn´t feel the urge to stay in place.
Need to mount the engine solid(now it is rubber hinged) and reinforce the rear swingarm.

So we couldn´t push it as hard as we wanted to.

[Wos]

The original YZ250TPI would run between 14 and 21% less fuel burnt than a standard, same spec, carburetor YZ.
Compared same days, same tracks, aulternating riders, over many events. Same top end power.
14% on the more open fast events, 21% was seen at slow low speed events.
One would have to conclude that that's a funtion of less fuel short circiting out the exhaust. Also partly because the engine would sort of be a stratified charge combustion at lower speeds with fuel only entering the B ports. A decient squish and toridal head seemed to help with this last ditch fuel / air mixing. Power is definitely more mellow down low in the rev range compared to the carburetor bike. Interestingly I could turn in a faster time on the TPI bike but the carb bike felt faster. Way more linner power delivery with less ' surprises'.

A narrow squish helps every 2 stroke to reduce the consum of fuel!

For example:
We reduced squish on a yz 125 to 0,7 mm and had to lean the mixture

Effekt... better throttle reponse...more power everywhere...consumption went down

My opinion...

If you have large squish often it is difficult or nearly impossible to find a satisfying setup for the carb...cause of low MSV...
which leads to low burn velocity...and that makes it necessary to richen the mixture up...

All the same on a yz 250..

Your experiences welcome to widen my horizon

Grüße Wolfgang

[wobbly]

Increasing MSV has two effects on combustion.
Firstly it increases the turbulent eddies that impinge on the expanding flame front , thus increasing the burn speed across the chamber.
This is analogous to advancing the ignition timing , without the down side of increased pumping losses on the way to TDC after the spark kernel has formed.
Secondly the reduced vertical clearance decreases the volume of end gases trapped in the squish area , thus making more A/F mixture available to burn earlyer
within the cycle - increasing the combustion efficiency.
All in all a win win , and a rare free lunch.

[Pursang]

All in all a win win , and a rare free lunch.

Not too often that Wobbly offers us a "free lunch"

[husaberg]

Not too often that Wobbly offers us a "free lunch"

He has offeredto do more than just buy your dinner....

Long time ago ,if I remember correctly , I said free blowys for life to anyone making over 30Hp at 10500 in a bucket - cant remember if that was 100cc ( with 24mm carb ) or 125 aircooled.
No ones even come close, so I am safe, for now.

12 bar i think?

[oldjohnno]

I've long been interested in chamber design. In the four-stroke world modern engines with shallow pent-roof chambers utilise tumble to promote fast flame propagation, with little or no squish/quench. Spark timing is typically around 20deg for these, about half of that used in many older closed-chamber engines.

I always thought the transfer flow in a 2T would promote as much motion as these current 4Ts, possibly more. As an experiment I machined a head with a roof that matched the slight dome of the piston crown, so that the chamber was a simply a slightly dished disc of constant thickness. I made it the same volume as the two heads I'd previously run, the first one having a traditional 50% squish band and a roughly hemispherical central bowl. The second had about a 35% area band and a low flat roof and performed slightly better.

On the dyno the new, open chambered head was surprising. It wanted exactly the same spark timing as the other heads and torque and peak hp was about 3% or so higher. Also surprising was the fact that it wanted significantly less fuel - about 4% or 5% less. However it wouldn't tolerate richness like the other heads would. Where they would run with jetting in excess of the optimum with little loss of power the open chambers output would drop much more rapidly with excess fuel. I don't know if these results are a weird anomaly or not but I was pretty happy with it overall.

In the 4T world the trend has long been away from squish and towards shallow, open chambers. Mainly driven by emissions requirements I think and the need to reduce UHCs. With these it's not squish that's the problem, it's the quench that goes along with it. And because the two are inseparable it's hard to tell what causes what. I kinda suspect the benefits of tight clearances are as often as not because of the reduced quench volume rather than mixture motion.