Those cars have different software versions, but the mapping is 100% the same on those versions so results should be identical.

I dont understand why you say its an AFR issue. Its clearly a load control issue

i dont know the ins and outs of mapping, but it is quite obvious that the AFR is not correct on daves car. why its incorrect, i have no idea, but its not helping the situation which i why i mentioned the possibility of it MAF/MAP or lambda sensor related.

I was always under the impression that AFR was best kept between 11.6-12. but that said i know nowt about EGT, so i can only go off what other tuners have mentioned previously. What about the requested rail pressure? iirc it was requesting 108bar peak. thats stage 1 std HPFP territory much lower than many stage 2+ cars i have seen.

Could you expand on the load issue. bit confused with that. why would the load differ from timmys car. They are ran identical on the dyno, so the loads should be near as dammit the same?

Also you say it is a car issue? what do youthink it is, and if its an issue with the car, how can software get around this, would it not be wiser to fix the hardware issue.

thanks for the input. i love all this geek stuff

i dont know the ins and outs of mapping, but it is quite obvious that the AFR is not correct on daves car. why its incorrect, i have no idea, but its not helping the situation which i why i mentioned the possibility of it MAF/MAP or lambda sensor related.

Yes but you're assuming the AFR being "incorrect" is the reason for the car making lower hp than Tim's, when infact it's the other way around. The AFR is fluctuating slightly *because* the engine load is changing. Your cause and effect are the wrong way around that's all

Basically the car is backing off slightly for some reason, which is causing the load to reduce which inherently increases AFR, it's just like taking your foot off the pedal (although by a much smaller amount).

As Mikko has said, although the software files are two different ones, the mapping inside is identical and therefore they should perform in just about the same way barring any car issues, which is where we are now.

Yes but you're assuming the AFR being "incorrect" is the reason for the car making lower hp than Tim's, when infact it's the other way around. The AFR is fluctuating slightly *because* the engine load is changing. Your cause and effect are the wrong way around that's all

Basically the car is backing off slightly for some reason, which is causing the load to reduce which inherently increases AFR, it's just like taking your foot off the pedal (although by a much smaller amount).

As Mikko has said, although the software files are two different ones, the mapping inside is identical and therefore they should perform in just about the same way barring any car issues, which is where we are now.

what do you think the issue is?

Pre-MED17 gen ECU's are load/torque controlled.

When you press the throttle, ECU converts that input into relative load request. Then it calculates how, when, why and where it needs to modify engine parameters to achieve that load.

This load is the % you see usually maxing up to 191,7% on logs (its actually 191,25 since its 8bit value with factor of 0,75).

When ECU decides that its necessary to fully open the throttle and request x millibars of boost to achieve that load, it does just that. ECU sends a signal to throttle valve, "open, sesame", and runs the requested boost value through a very sensitive PID controller.

The PID controller output is then filtered through a linearisation curve that defines the physical characteristics of the turbocharger, and through that filtering we get the duty cycle %age used to control the N75 valve.

Now, ECU sees that it wanted 180% of load with 2420mbar of boost, but suddenly load is 188% and boost 2520mbar. It has a list of stuff to take care of. In this case, the overboost isnt too big, so it just changes the n75 duty cycle a bit to lower the boost.

But, (and this is the car issue I suspect...), this time the turbocharger, for whatever reason, isnt reacting the way the PID controller is set up to expect. So the boost drops too much, to lets say 2210mbar. So ECU sees this, and it increases it a bit again, causing another overboost, which causes the oscillation we see midrange on the print.

(If the overboost would be serious, it would start closing down the throttle, then doing ignition, then a complete fuel cut if it isnt enough, but this is not the case here).

When we are close to maximum possible output of different sensors and using high duty cycles on different control valves, the slightest difference or a very minor problem in the car might cause these kinds of issues. Easiest way to go round them would be to start removing different protection features built in these ECU's, in this case, increase the allowed boost overshoot. We aim for the more bespoke way of doing things.

From that load % ECU calculates the required injector opening time and requested rail pressure.

When the load % fluctuates, as it now clearly does, it messes up the AFR calculation, because the opening time is defined as basic opening time in milliseconds per 1% of load.

When actual load is offset from requested load, it makes fuel control inaccurate.

Thats a VERY simplified explanation of what goes on there if it is what I think it is. (the Bosch documentation for MED9.1 ECU is 6000 pages, and just the load control algo is over 1000 pages of flow charts...)

what do you think the issue is?

Not sure yet, we've been discussing it while at MPH.

The basics are:

Tim's car runs a 100% identical map to my own VRS

Dave's car runs a 100% identical map to your old VRS Sy

Now the interesting thing is, and I'm not pointing any fingers here at all, that Dave has an ITG and so did yours Sy. The problem that Dave is experiencing appears to be a similar although less pronounced problem as you experienced. The car is holding back for some reason. Now from listening to the dyno video of Dave's car it's not timing pull as it's not knocking (or the exhaust doesn't sound like it is). I don't think it's fuelling (or at least I don't think it's leaning out) either although one possible theory is that the fuel pump is cavitating.

So we're pretty much left with boost control or fuelling fluctuations due to a sensor reading. As I said above, I don't think it's knocking so it shouldn't be that, and if it was then the cut would be more severe if knock protection kicked in. It isn't too lean otherwise EGT protection would just kick in and shut the car down (so it's definitely not running too lean as we're one of the few companies that leave the EGT protection *exactly* where it should be ). It could be a MAF problem, it could be an N75 problem, it could even be a knock sensor problem but I really doubt that from those results and the way in which it fluctuates.

cheers mikko.

Makes sense, im familiar with a how a PID control system works and how it should smooth out the oscillation and over shoot to provide that solid flat AFR. But the problem im seeing is whilst you can remove the safety feature thats triggering the oscillation in the first place, its not solving the potential hardware issue in the first place. what hardware could be replaced to solve this.

Edited 15 November, 201015 yr by janner_Sy

On a slightly different note Sy, I do like your little turbo graphic

ben

As i said in an earlier post, REVO had approached ITG to develop an intake that could flow enough Air to replicate an issue that guys in the USA had been having with really high flow intakes. I have also read of guys with APR software having issues with the twintake/ITG, which when the swapped inthe awesome-GTI amnesty has disapeared.

could it be a result of that?, although iiirc daves MAF was quite low for stage 2+. around 204g/s compared to vRSCarls making 234g/s

cheers mikko.

Makes sense, im familiar with a how a PID control system works and how it should smooth out the oscillation to provide that solid flat AFR. But the problem im seeing is whilst you can remove the safety feature thats triggering the oscillation in the first place, its not solving the potential hardware issue in the first place. what hardware could be replaced to solve this.

No no no, we do not REMOVE it! We just move the threshold a bit. Removing safety features always makes me go RRRRrrrrr...

Perhaps its the intake, perhaps its that this car was built on monday and not tuesday, it could be just about anything. Sometimes its better to tweak the map slightly than to start randomly changing parts. It would be very educational to compare the intake kits though.

It only takes a very slightly worn MAF, N75 valve, MAP sensor, or the slightest leak somewhere, to cause this issue.

Lambda PID is untouched, so we need to get the load smoothened to make fueling PID work as it should.

Hi chaps!

I'm stuck in Orly airport in Paris with a 3 hour+ delay so this thread is keeping me from insanity. Thanks! I can just about follow the techie stuff too

Sy, you are right - my MAF readings were always way low.

Mikko/Ben - do you remember the data I sent that Keith at JKM produce? I think we are talking about the same problem as then. There was still an element of guesswork then, of course, but boost control was mentioned at that time.

Dave

ben

As i said in an earlier post, REVO had approached ITG to develop an intake that could flow enough Air to replicate an issue that guys in the USA had been having with really high flow intakes. I have also read of guys with APR software having issues with the twintake/ITG, which when the swapped inthe awesome-GTI amnesty has disapeared.

could it be a result of that?

I don't know, but it has to be said that in some of the TFSI maps I've seen from Revo a lot of deviation is often allowed before the ECU does something to rectify a condition, and in some cases the rectification procedure has even been effectively removed.

Sometimes this can mask a hardware issue.

What I'm saying is that perhaps a "problem" that was there with APR's software isn't there any more because the ECU doesn't care about that problem any more.

No no no, we do not REMOVE it! We just move the threshold a bit. Removing safety features always makes me go RRRRrrrrr...

Sorry pal, i misread that there mate

"Easiest way to go round them would be to start removing different protection features built in these ECU's, in this case, increase the allowed boost overshoot. We aim for the more bespoke way of doing things."

Perhaps its the intake, perhaps its that this car was built on monday and not tuesday, it could be just about anything. Sometimes its better to tweak the map slightly than to start randomly changing parts. It would be very educational to compare the intake kits though.

It only takes a very slightly worn MAF, N75 valve, MAP sensor, or the slightest leak somewhere, to cause this issue.

Lambda PID is untouched, so we need to get the load smoothened to make fueling PID work as it should.

whats involved in the N75 valve change. MAF, MAP and lambda are easy to swap out. is the N75 the waste gate isnt it?, is that an intergal p[ar t of the turbo, ie if the N75 goes, a newturbo is required, or is it replacable as valve only

ben/mikko

looking at timmys torque curve, whilst being very good and holding on very well, it almost looks like it was drawn with a ruler at 300ibft for almost 2000rpm. Which in all fairness is not a bad trait in any way shape or form, but if you were to apply the normal hump style torque curve to that plateau it alooks like it could make around 320Ibft (obvioulsy not an exact science )

Is that to purposefully done to limit the torque to a safer amount?

Edited 15 November, 201015 yr by janner_Sy

Ps - if there was any knocking then I'm sure Keith would have picked that up.

Just curious...Would a slight leak in the manifold/downpipe join do this? I always have little faith in myself even although I'm sure I tightened the heck out of it when fitting!

ben/mikko

looking at timmys torque curve, whilst being very good and holding on very well, it almost looks like it was drawn with a ruler at 300ibft for 2000rpm. if you were to apply the normal hump style torque curve to that plateau it alooks like it could make around 320Ibft (obvioulsy not an exact science )

Is that to purposefully done to limit the torque to a safer amount?

Yep, that's a torque limiter. The car should always run on a torque limiter rather than being allowed to make as much power as mechanically possible.

Yep, that's a torque limiter. The car should always run on a torque limiter rather than being allowed to make as much power as mechanically possible.

surely another 20ibft would be ok though??

Hi chaps!

I'm stuck in Orly airport in Paris with a 3 hour+ delay so this thread is keeping me from insanity. Thanks! I can just about follow the techie stuff too

Sy, you are right - my MAF readings were always way low.

Mikko/Ben - do you remember the data I sent that Keith at JKM produce? I think we are talking about the same problem as then. There was still an element of guesswork then, of course, but boost control was mentioned at that time.

Dave

First of all, sorry to hear that mate. Orly is one awful place to be stuck in (well, I once did CDG T3 for 7 hours, 2100-0400 with the normal CDG-style lost luggage thrown on top of that...)

Your horsepower is quite high for that MAF reading. Have you tried a new MAF sensor in your car? (Old MAF sensor reads lower than actual air flow values, which cause all kinds of trouble on these)

surely another 20ibft would be ok though??

It's not really a case of that mate, it's about choosing a limiter and allowing the car to run to that limiter throughout. The limiter that you choose should govern the amount of power the car is creating when at WOT. A typical humped graph, while it feels faster, is usually the product of a flat-lined torque limiter (i.e. set higher than it'll ever reach) and allowing the car to make power in a comparitively uncontrolled manner.

surely another 20ibft would be ok though??

Well, no. After that, IAT starts to increase rapidly because the turbocharger runs out of efficiency and starts producing more heat than airflow. Yes, it would give that 20lbft that looks nice on a dyno print, but it wouldnt be faster on the road at all.

Our engine is 100% controlled by the ECU, not by the amount of air mass turbocharger is capable of producing. The extra 20lbft would cause alot of stress to clutch and powertrain, wouldn't make it any faster on the road, and in my opinion, makes it too jerky to drive.

It's not really a case of that mate, it's about choosing a limiter and allowing the car to run to that limiter throughout. The limiter that you choose should govern the amount of power the car is creating when at WOT. A typical humped graph, while it feels faster, is usually the product of a flat-lined torque limiter (i.e. set higher than it'll ever reach) and allowing the car to make power in a comparitively uncontrolled manner.

what do you guys consider the peak torque for the k03 engine/block.

about time we had a proper thread like this. beats the " whats your delivery date" threads :p

Well, no. After that, IAT starts to increase rapidly because the turbocharger runs out of efficiency and starts producing more heat than airflow. Yes, it would give that 20lbft that looks nice on a dyno print, but it wouldnt be faster on the road at all.

Our engine is 100% controlled by the ECU, not by the amount of air mass turbocharger is capable of producing. The extra 20lbft would cause alot of stress to clutch and powertrain, wouldn't make it any faster on the road, and in my opinion, makes it too jerky to drive.

agree with you there., much more and you'll stress the clutch although you would notice 20ibft especially when on the motorway.

Edited 15 November, 201015 yr by janner_Sy

agree with you there., much more and you'll stress the clutch although you would notice 20ibft especially when on the motorway.

Which is exactly where you don't want it. That's what will kill clutches and cause massive IAT hikes which will actually lose power. OK for one dyno run, not so great on the road.

what do you guys consider the peak torque for the k03 engine/block.

In my experience, 110-120Nm per cylinder seems to be a nice safe figure. It is a strong little engine (you dont see many engines with so nicely balanced internals from factory...)

In my experience, 110-120Nm per cylinder seems to be a nice safe figure. It is a strong little engine (you dont see many engines with so nicely balanced internals from factory...)

whats that in old school money. ....

1nm=0.74lbft

110nm=81lbft

120nm=88lbft

81lbft/cylinder x 4=324lbft

88Ibft/cylinder x4 =352lbft

so 324Ibft-352ibft!!!

whats that in old school money. ....

1nm=0.74lbft

110nm=81lbft

120nm=88lbft

81lbft/cylinder x 4=324lbft

88Ibft/cylinder x4 =352lbft

so 324Ibft-352ibft!!!

Hehe yes sorry, should've used imperial instead of SI

I think the ultimate load limit for the connecting rods is somewhere around 400-450lbft on these engines. 350ish is safe, or atleast thats how I feel after opening a "few" of these out of interest, but I've seen clutch problems with cars running only slightly over 300lbft.

Hehe yes sorry, should've used imperial instead of SI

I think the ultimate load limit for the connecting rods is somewhere around 400-450lbft on these engines. 350ish is safe, or atleast thats how I feel after opening a "few" of these out of interest, but I've seen clutch problems with cars running only slightly over 300lbft.

but for a guy who uses his car on track with full range of chassis mods and uses the gears properly the limit could surely be raised. I for one would rather take the extra torque and instead invest in an uprated clutch and gain more mid range. to be honest id not want more than 320ibft, that would be the limit of usability IMO, even on a 'setup' car

Edited 15 November, 201015 yr by janner_Sy