We can only help if we understand the plan and the actual situation.

Those who try to summarize (for the reader, not just for themselves!) thematically are on the good path, and those who don't care to detail (the circumstances, plans) in a way the reader can understand should reformat (yes, I know that crying without exactly specifying actual state and what is desired is valid interaction in mothers-babies relation, but forgive me I cannot afford to tend a 1000 babies).

Project subpages:

The rest of this page (even if just one engine project) is mostly chronological (perfectly OK, but not sufficient for tech help)

If help is needed on certain part => split to subpage (use thematical organization, not chronological, and refer to this main index page)

Audi RS2 Avant



Mahle pistons

scat rods

GT3071R turbo (0.63AR)

OEM inlet/exhaust manifold

custom intercooler/downpipe

water-meth injection

70mm catback with x2 200cpi BN pipes catalysts

Stock cyl head with stock ADU cams

Spec 3+ clutch kit for dual mass flywheel

running VEMS with megatune 1.1.27 for 9 months now!!!

Results with Perfomance box: around 410hp with 1.85bar overb. and 1.7-1.8bar steady to redline on 4th gear pulls.

Max torgue reached: 600Nm, but keep it around 580Nm now.

Next year:

Holset super HX40 with 0.70AR

reverse inlet manifold

SQ exhaust manifold

New downpipe


Idle control finally solved but always small modifications for fine tuning. I use diode for ISV valve and a combination of ignition based idle and IAC intervention to keep idle steady and allow minimal play when fan kicks in or A/C is working. So far im very pleased with the results

Version 1.1.27 is very stable but would like to switch to 1.1.62 with vemstune.I managed to connect the laptop to ECU on 1.1.27 with vemstune. I will wait for the 1.1.63 version which has the boost control bug fixed, as i use vemstune to control boost with an industrial solenoid valve. Very accurate boost control!!!


Removing the MAF from the airfilter caused some overboost cut outs for some reason and i had to adjust the boost control figures. Noticed that the offset feature will not accept value 15 on 1.1.27. It will only go from 16 to 14, although when i did save and burn to ECU the value 15, i did notice a change in the highest boost reached.


After more tests, i had to reduce the boost offset to 12 and redo the boost control vs rpm in order to avoid fuel cuts as it was spiking. But everything seems to be working fine now. I do get a bit of drop in boost on higher revs, but i dont mind. it is minimal,and maybe required as well.

The idle has been altered a bit,but here is a video of how it works now. I am really pleased,and the only thing which has been posted before about A/C control and idle is the one thing that can possibly make it even better.

Marcel has my settings and can work with them or improve them,but the way they are now, it is working quite well.

Car is making about 420hp and 620Nm of torgue.


car now has:

SQ exhaust manifold

New downpipe

GT3076R turbo, 0.82AR

reverse Wagner inlet manifold

220psi water injection pump

7A flywheel and southbend clutch kit

Car now produces around 500hp

Im still using 1.1.27version of megatune as Vemstune still gets upgraded every other week and there isnt a version yet that can utilise boost offset feature. Im planning to upgrade to a later version of vemstune as soon as i feel that it is stable and working nicely.

So far im pleased with the results from the car and i drive it everyday!

TODO: cleanup this and add vemslog

Pasting some info from mail from Vasilis:

IDLE: (1.1.81 firmware now).

i used the following method to tune idle speed:

Ignition based idle: DISABLED

integral limits: ZERO

cold start the engine and set idle ref DC

I use for the bosch 2pin IAC valve: 35hz and limits of 0 and 100% DC

I managed to make the ref DC table close to my cold and warm idle speeds using the IAC control alone.

Problem: Once IAT starts to go higher than 25-30C the idle speed goes UP!!!

I have disabled the refpos for MAT slope feature completelly and it still does it!!

with ignition idle settings enabled i end up having something like: 1050-1080rpm idle speed on CLT:83-85.

If IAT is anything between 0-18C --> idle rpm is 940-970 (which is my target rpm)

On the bosch valve and with the above settings, any DC below 28 on the refDC table, makes the idle speed HIGHER!.

So setting the integral limit more than 20 takes the refDC below 28, and the idle speed goes higher!

my refdc table has the number 29 for refDC from CLT of 70 to 90C.

In the log in Idle mode the minimum was 839 RPM, can you excactly show your problem?

  • Could you post a log file with high and low MAT, so we can investigate why idle speed goes up

Do you use ego-correction? I`ve seen simular problems with ego correction off, and I would have to modify the MAT/TPS table at 0% trottle. That was related to an MAT-sensor heatsoaking issue, causing lambda to go lean and rich after different tempratures. -reidar


thanks for the answer, i do use EGO correction and have some heat soak issues for which i used a phenolic spacer which minimised the problem. However according to some tuners, if the IAT goes higher, the idle usually drops, it doesnt go higher, which is why mine does the opposite for some reason.

I have now modified the TB a bit and i get a bit better reaction, so im happy with it.

Tuning cruising lamda and ignition:

may i ask what is the optimal that tuners have used for the areas:



compression ratio: 9:1 and fuel: 97 octane

what lamda and what timing?

I am struggling to get good fuel economy on my car. I know i should have gone to a dyno and did the proper thing, but id like to know what other people use for the time being.

i can tell you why idle speed goes up when MAT goes up.

integral degrease limit tries to keep idle speed low when the MAT goes up(as idle goes up with it if IAC is off completelly, i dont know why), sometimes its only the IAT sensor heatsoaking probably.

if integral degrease limit takes the IAC DC below 30, then the audi IAC goes into its safe mode which from factory means that it starts to open again instead of keeping shut. For audi IAC, 30% DC is the completelly closed position. So as integral is taking the DC to lower levels (for example 25% or 20%, the idle speed goes high.

Try it in an audi with this mechanic IAC valve and you will see that if you use DC of 20 or less, the idle goes higher. WHich is why i set the lower DC limit to 29 so as the idle to try and stay as stable as possible.


The engine is rebuilt with new pistons (oem mahle 81.51), ported head with larger valves, higher compression than 9:1(probably closer to 9,3:1), southbend clutch kit full feramic.

I replaced the injectors (siemens deka) with EV14 650cc.

I have the simplified dead time values from the shop:

8volt 2.55ms = 2550us

10volt 1.70ms = 1700us

12volt 1.20ms = 1200us

14volt 0.92ms = 920us

16volt 0.75ms = 750us

Apparently when i use these values i get another validation error from VEMS, although they are the correct spec from the shop they tested them.

Vems help says that i may get very low values on the VE table around idle. Other than that, is there any other reason i would worry about this new validation error, or should i ignore it?

Above values seem OK, validation values are too strict, we'll change them, Ignore the warning for now


Car now has been fully mapped and water injection is back online after a faulty controller. Looks like it may be spraying too much mixture (70/30) but ill need to keep logging to see the results and if the lamda becomes too rich robbing off some power.

I still get secondary trigger errors above 6800rpm. I use a 7A flywheel with a larger welded pin, as VEMS was not able to 'see' the OEM pin on the flywheel. It doesnt misfire or anything, but its annoying, as the rpm above 6800 jump around and cannot be logged properly by VEMS.

Unless this gets fixed in future firmwares, i would advise a 60-2 on audi I5 engines to make them read properly

Car is now producing around 394whp and 380ftlbs on the virtual dyno runs, done by using VEMS logs. When i compared these dyno runs to a dyno-dynamics dyno i was recently on, they were pretty accurate if you input the weight and frontal area of the car, as well as the gear ratio. On dyno-dynamics i had 368whp with less boost and higher IATs. A comparison between the same excact run on the dyno-dynamics and virtual dyno was:

315whp dyno dynamics (with 30C IAT) vs

330whp on virtual dyno (with 10C IAT)

Here is a 4th gear pull:


Wheel speed control is still not working and i assume i would need a hardware change for that, but boost control is acceptable on all gears now, so i dont mind.



After a few improvements on the boost settings the curve now is much better and boost builds up quicker. I could try a faster N74 mac valve(4port) but i think its not going to be all that different.

As EGTs get higher after 6500rpm, i have reduced the boost thereafter, and now the car makes more power than before,as full power seems to be there up to 6400rpm.

here is an example of the boost improvement:3d gear pull


PID settings are very car specific and therefore need tuning from zero. No settings from someone elses car will work 100%.

Also, cold starts are faster as well with new IAC cranking DC updated!



A few adjustments on the settings the last few weeks include:

1) A fix of a weird idle fueling issue which allowed on hot engine, high MAT the car to become VERY rich if MAT/TPS stayed at 100, and once fuel was drawn away with the table, going down to 90 from 100, the car 'felt' lean. Checked lamda, settings etc and replaced the car's battery in the end. Result is getting the car to behave again by showing that heatsoak is creating a lean mixture when MAT is high and fuel needs to be added rather than subtracted.

2) Replaced the N80 valve and trying to figure out a testing procedure through VEMS as it seems that it doesnt respond although new, and once 12v is given to it, it reacts. Settings are in place for it to work on deceleration only, but need to make sure it actually works. This valve is CLOSED once 12v is applied and OPEN when no power goes through it.

3) Fixed a small issue with rich mixture being caused by EGO riching up the mixture after deceleration and once rpms reach idle speed, as it tries to enrich a closed fuel supply and lamda of 1.29 causing a rich mixture whith took about 10sec to go back to normal lamda, and obviously made the city driving quite high in consumption!! Beware of your EGO settings for this, as it will make the car very uneconomical.

Awaiting of a pressure valve to see if my A/C will be fixed and if so, do the final testing of idle speed with the A/C being on. I have already tried this on a friends RS2 which i tuned a month ago and it works brilliantly. At last idle speed without issues!


please use this link to get config for updating to new ECU

EVAP and AC idle control


I would like to incorporate 2 features on my latest (??) ecu :

What should I do?

As I have told you before, I do not possess the knowledge to make such changes as you describe. You know what the ecu has from hardware

and since the AC feature was done only recently and requires more than just a config setting

If a hardware change is required in order for the ECU to be able to use the settings provided in 1.2.38 firmware

If you do not wish to help, then please direct me to someone who can.

By the way this system of pages is so old and complex that i do not understand how to make subpages of a page.