Reesy's Big Turbo A3

[Reesy]

FAKE S3
FAKE S3
FAKE S3

If they did A3 clocks with white needles I'd have bought them.
It has a k04, like a fair few modded a3's.
Same brakes as a skoda vrs on the front, and r32 rears.
The only S3 bits are the spoiler, clocks and the wheels, and the wheels are off soon.

Plus you won't have to look at it for long as it'll be flying past you on the way to AITP :cheers:

 

[<tuffty/>]

Hmmmm, bit of a mixed bag tonight.

Didn't get the bumper on.
Got the S3 clocks on and they look awesome, but getting the flashing immob symbol so need to get that off, and change the mileage to the correct amount, but as I said they look awesome.
Fitted a new MAF and immediately the numbers were up.
With the proper s3 map it made 288bhp. The boost needs winding down and the map is very rich, but we think that's due to it trying to dump fuel as if it was egt related..

The more this goes on, the more I think it would be easier to just whack the narrowband loom on and stick the above map on a narrowband ecu.

Also, pretty sure I have a ps leak. Need to check in the morning.

You are running an S3 map? from what I remember the TQS ECU is a 512kb ECU so at a guess it would be an APY? they run rich out the box.. .in the 10's from memory...

Also... make sure you use a matching MAF sensor as the readings will be out... fitted an APY sensor to an AMK the other week and the readings were artificially high... if using an APY map then you should use an APY sensor else the airflow readings will be meaningless mate

<tuffty/>

 

[BENJAMIN]

You are running an S3 map? from what I remember the TQS ECU is a 512kb ECU so at a guess it would be an APY? they run rich out the box.. .in the 10's from memory...

Also... make sure you use a matching MAF sensor as the readings will be out... fitted an APY sensor to an AMK the other week and the readings were artificially high... if using an APY map then you should use an APY sensor else the airflow readings will be meaningless mate

<tuffty/>

His is a wideband 1mb ecu. He's running a bam map but as his engine is an early TT engine it has had the egt and vvt coded out, but for some reason it's being defaulted into requesting 0.649 lambda at all times. No fault codes though. I know dan wouldn't of got the lamfa map to request that. Very very odd. Pulls 229.97g/s but is rather rich lol.

 

[<tuffty/>]

His is a wideband 1mb ecu. He's running a bam map but as his engine is an early TT engine it has had the egt and vvt coded out, but for some reason it's being defaulted into requesting 0.649 lambda at all times. No fault codes though. I know dan wouldn't of got the lamfa map to request that. Very very odd. Pulls 229.97g/s but is rather rich lol.

Without an EGT sensor fitted this is what happens... coding the fault codes won't stop the logic from operating...

The coding for VVT is in the program code so VVT can't be switched off without setting NWS_SY and recompiling the code... I would imagine this is also the same for the EGT probe....

You would be better off starting from an AUM map and scaling everything up...

<tuffty/>

 

[BENJAMIN]

Without an EGT sensor fitted this is what happens... coding the fault codes won't stop the logic from operating...

The coding for VVT is in the program code so VVT can't be switched off without setting NWS_SY and recompiling the code... I would imagine this is also the same for the EGT probe....

You would be better off starting from an AUM map and scaling everything up...

<tuffty/>

Thats the plan as of yesterday. The original ko3s file for his ecu is in the process of being modified atm.

 

[<tuffty/>]

VVT will still 'technically' be an issue as it can't be actually turned off... not tried it myself but you 'could' get over it by coding out the fault codes and making the position 2 maps the same values as the position 1 tables so even if the program logic tries to go to the p2 tables it will continue to do the p1 values seeing as the cam won't have actually moved...

<tuffty/>

 

[BENJAMIN]

That's what dan has done I believe, the connector still needs resistance to stop the code showing on that map. But realistically I may swap the vvt unit into the TT engine anyway as its one less potential problem.

 

[beachbuggy]

Main problem has been the fact it's a bit of collection of parts at the moment more so the ECU, Loom, and sensors. The original ecu's on the is a 1mb me7.5 06a906032hn that was swapped for an Audi TT 180 ecu 8l0906018q ecu . All going into an APX engine. Ideally the same ECu and loom would be good especially if it was for a TT. Everything's been fine expect for this Rich running where it's solidly requesting 0.65 lambda. I am almost 99% sure it's down to the egt being coded out and not running a rear 02 sensor which has also been coded out and running an s3 map on the ecu. I know the ecu will run the map and the egt works IF it's on the correct loom but I don't think it's possible to fully delete the egt and hence the rich fuelling.
The plan is to stick with the original 06a906032hn ecu and then map it from there, I've got a bam ecu but sure the same thing will happen with the egt. We'll see

 

[Westy]

I've already told Reesy this but his injectors will definitely be maxed out at that airflow. Mine only flows the equivalent of 285bhp and the injectors were running at 130% on a 3bar FPR

Once you've got the mapping sorted he may need 550cc injectors and you'll have to map again lol

 

[Prawn]

I reckon that airflow will be false due to what Tufty said.....

There's no way on earth it'll be slowing 229g/s at 0.65 lambda request, SURELY?

the only way I could see that ever being possible is if it's leaned itself out to a sensible AFR at the top end due to the injectors being maxxed out!

Either way, not a nice thought.

 

[<tuffty/>]

Main problem has been the fact it's a bit of collection of parts at the moment more so the ECU, Loom, and sensors. The original ecu's on the is a 1mb me7.5 06a906032hn that was swapped for an Audi TT 180 ecu 8l0906018q ecu . All going into an APX engine. Ideally the same ECu and loom would be good especially if it was for a TT. Everything's been fine expect for this Rich running where it's solidly requesting 0.65 lambda. I am almost 99% sure it's down to the egt being coded out and not running a rear 02 sensor which has also been coded out and running an s3 map on the ecu. I know the ecu will run the map and the egt works IF it's on the correct loom but I don't think it's possible to fully delete the egt and hence the rich fuelling.
The plan is to stick with the original 06a906032hn ecu and then map it from there, I've got a bam ecu but sure the same thing will happen with the egt. We'll see

Pretty sure EGT issue will remain if you use the S3 map... EGT probe hardware is coded into the software logic as far as I am aware... ECU is for the most part irrelevant as most of the ecu's are pretty much the same hardware wise (within the same family) so you can essentially use an AUM wideband ECU in an S3 if flashed with the S3 code and a BAM ecu in an AUM A3 flashed with AUM code etc..

The physical hardware each code is expecting to see will dictate how the program code runs to a point... if the program code of the chosen map is expecting the hardware the program logic will react as its designed to... in the case of AMK/BAM engine code its expecting to see a physical EGT probe sensor... if it doesn't see the sensor because its faulty or in this case not present at all then the program logic does whats its programmed to do and go into a protection sub routine to ensure AFR is at a level suitable to protect the engine... its a fail-safe mode..

ME7's typical strategy for AFR is to use the richest value derived from the various functions in code... without an EGT probe then the EGT protection function wins as this will make stuff go mega rich...

The function diags kicking about the web/Nef may not show this as they are primarily ME7/ME7.1 rather than ME7.5 but not looked tbh as my primary focus has been my particular setup...

I have heard of tuners retro fitting the EGT probe and using S3 maps to gain the benefits of this type of AFR control as the strategies in physical EGT enabled code is different to that of those without... LAMFA for example runs mostly 1's as does KFLBTS on AMK/BAM maps... wideband K03 maps are tuned... this is reflected in the fact that AMK/BAM cars run lower CO than wideband A3's...

<tuffty/>

 

[BENJAMIN]

I reckon that airflow will be false due to what Tufty said.....

There's no way on earth it'll be slowing 229g/s at 0.65 lambda request, SURELY?

the only way I could see that ever being possible is if it's leaned itself out to a sensible AFR at the top end due to the injectors being maxxed out!

Either way, not a nice thought.

It does actually lean out towards the top. Lambda actual tails to 0.8 at the redline, injectors and/or pump the issue there, but .8 shows there is just enough scope in the fuel system currently. It was a bam map and bam maf so should be accurate, certainly feels like it as it goes well!

 

[Reesy]

I am seriously tempted, and I think it would be a lot easier, to stick the apx loom in and stick a narrowband ecu, or i believe Dan has an apy setup so might try and sort something out with him and stick that in the car. That way their are no issues. Plus I know Ben's narrowband and Welly's run fine, despite not being mapped perfectly and Prawn's seems alright as well..

I think it's just a case that the software doesn't want to co-operate with the mix and match it has at the moment, but if values of the map it has can be copied into a narrowband setup, I'd be more then happy as it really flies. Whether the numbers are right or not anything within 5-10% less of this would still be very fast.

I have a 4 bar fpr coming to see if that'll help things, and the fuel filter still needs to go on which might help the flow of juice, as I doubt this one has ever been changed..

Also, I think the rack is causing the power steering leak.

 

[badger5]

It does actually lean out towards the top. Lambda actual tails to 0.8 at the redline, injectors and/or pump the issue there, but .8 shows there is just enough scope in the fuel system currently. It was a bam map and bam maf so should be accurate, certainly feels like it as it goes well!

what does the request do...
logged 001,031 and 002 to see what occuring?

 

[badger5]

I am seriously tempted, and I think it would be a lot easier, to stick the apx loom in and stick a narrowband ecu, or i believe Dan has an apy setup so might try and sort something out with him and stick that in the car. That way their are no issues. Plus I know Ben's narrowband and Welly's run fine, despite not being mapped perfectly and Prawn's seems alright as well..

I think it's just a case that the software doesn't want to co-operate with the mix and match it has at the moment, but if values of the map it has can be copied into a narrowband setup, I'd be more then happy as it really flies. Whether the numbers are right or not anything within 5-10% less of this would still be very fast.

I have a 4 bar fpr coming to see if that'll help things, and the fuel filter still needs to go on which might help the flow of juice, as I doubt this one has ever been changed..

Also, I think the rack is causing the power steering leak.

stick with wideband imho

 

[BENJAMIN]

what does the request do...
logged 001,031 and 002 to see what occuring?

Its always requesting 0.649, actual lambda tails from 0.75 until 4500rpm and by 6000 rpm it progressively tails to .83 at the redline. Injector m/s are very high but being wideband it's going to try and compensate if it was the pump that was dying I guess. Don't think the a3q came with an s3 pump so my money's on that?

 

[<tuffty/>]

Its always requesting 0.649, actual lambda tails from 0.75 until 4500rpm and by 6000 rpm it progressively tails to .83 at the redline. Injector m/s are very high but being wideband it's going to try and compensate if it was the pump that was dying I guess. Don't think the a3q came with an s3 pump so my money's on that?

I suspect if you logged live lambda control (001) this will be at 0... this is a result of no EGT probe feedback so the ECU fuels at 0.6 and no live correction is performed... you need to fix the EGT probe (or lack of it) issue and this will then be irrelevant...

No point in chasing around trying to adjust it out as it won't happen on the current setup..

<tuffty/>

 

[Reesy]

stick with wideband imho

What are the downfalls Bill? I know with wideband the car will adjust to certain things, and you lose that flexibility in a narrowband, but I mean on a more technical level.

 

[BENJAMIN]

I suspect if you logged live lambda control (001) this will be at 0... this is a result of no EGT probe feedback so the ECU fuels at 0.6 and no live correction is performed... you need to fix the EGT probe (or lack of it) issue and this will then be irrelevant...

No point in chasing around trying to adjust it out as it won't happen on the current setup..

<tuffty/>

I didn't mean the fuel pump was the cause of all the probs.
A solution using the original ary map as a base is being worked on I believe.

 

[<tuffty/>]

Ok, with you...

<tuffty/>

 

[badger5]

Its always requesting 0.649, actual lambda tails from 0.75 until 4500rpm and by 6000 rpm it progressively tails to .83 at the redline. Injector m/s are very high but being wideband it's going to try and compensate if it was the pump that was dying I guess. Don't think the a3q came with an s3 pump so my money's on that?

block 112 would be showing lots of % egt oh my god its hot adjustment... and 001 switches off in deferance to egt kicking in
as its leaning out with revs tho, I would be checking fuel pressure... and log your injector duty cycles... max;d out perhaps

 

[Broken Byzan]

So you want to use a dumber ECU so you don't have to tune it correctly?

Weird way around it,lol

 

[Reesy]

Not really, I was just using the others as an example.

It started life as a narrowband engine, the turbo came as part of the narrowband lump, and has no egt sensor. It came with the narrowband loom. All I need to do is plug that in and run a narrowband ecu and it should all mate up.

At the momemnt I am trying to run the narrowband lump with a wideband ecu, which seems a more odd way around it, as does fitting an egt sensor. Surely it would make more sense to get it working to what it was originally, rather then trying to make it work with something it shouldn't or won't very easily? This seemed the best option to begin with, as I was under the assumption that the egt could be coded out.

Rather then changing the ary ecu map so much, it would be easier to run it narrowband. Unless their is something majorly wrong with narrowband that I am unaware of.

 

[<tuffty/>]

Either works but wideband is a nicer place to be... issue you have to get over is the maps for wideband ECU's have basic hardware requirements so a wideband lambda (obviously) and VVT being the main ones (DBW TB and MAP sensor too if coming from an AGU)....

The EGT probe issue is just related to the map you are using... its an expected sensor for the program code so the logic around it depends on its presence.. as far as I am aware most if not all ME7.5 ecu's are EGT probe capable its just the program code that determines its use... you could wire an EGT probe into an AUM ecu and it won't do anything as the software isn't aware of it...

If you stick to the wideband ECU and use a K03 map (a 150hp one rather than a 180hp map as they had issues) then it will be fine... modelled EGTs etc can be sorted as can MAF scaling/injectors and fuelling...

Just remember that despite the age of the car etc its still the same basic hardware engine wise that can be found in an 05 plate TT...

<tuffty/>

 

[BENJAMIN]

Just buy a f**king V-TEC Reesy you g**.

 

[badger5]

Not really, I was just using the others as an example.

It started life as a narrowband engine, the turbo came as part of the narrowband lump, and has no egt sensor. It came with the narrowband loom. All I need to do is plug that in and run a narrowband ecu and it should all mate up.

At the momemnt I am trying to run the narrowband lump with a wideband ecu, which seems a more odd way around it, as does fitting an egt sensor. Surely it would make more sense to get it working to what it was originally, rather then trying to make it work with something it shouldn't or won't very easily? This seemed the best option to begin with, as I was under the assumption that the egt could be coded out.

Rather then changing the ary ecu map so much, it would be easier to run it narrowband. Unless their is something majorly wrong with narrowband that I am unaware of.

with that logic - you should have left it all standard then

lol

 

[beachbuggy]

I suspect if you logged live lambda control (001) this will be at 0... this is a result of no EGT probe feedback so the ECU fuels at 0.6 and no live correction is performed... you need to fix the EGT probe (or lack of it) issue and this will then be irrelevant...

No point in chasing around trying to adjust it out as it won't happen on the current setup..

<tuffty/>

Agree 100% ... It seems and hasn't really been confirmed anywhere that you can't delete thè egt, you can leave it plugged in but can't actually code it out completely.

 

[Broken Byzan]

I'm still going to change to WB for the better features available for the 1mb ECUs. I need to get hold of the extra bits, and still believe the hassle / cost is worth the benefits over the NB ecu.

M383 / 382 like in prawns are perfectly tunable but you need to remember that it isn't as clever in regards to it's learnt values, so if something goes wrong there is no compensation available to keep the car running without it leaning out and nuking a valve / piston for instance

 

[<tuffty/>]

Agree 100% ... It seems and hasn't really been confirmed anywhere that you can't delete thè egt, you can leave it plugged in but can't actually code it out completely.

I think the reason for this is that the EGT probe (and VVT) is crucial to the strategies and models used by this particular code... removing them physically means the code can no longer function as intended...

The feedback from the EGT probe and the function of the VVT in context of the 210/225 maps is built into the execution of the code of these particular maps and without them the program logic is essentially broken...

There are fail safes of course hence the rich fuelling... VVT is controlled (from what I understand) by a system constant SY_NWS:
SY_NWS: camshaft adjustment
= 0: No NWS
= 1: 2-point NWS
= 2: steady NWS
> 2: not defined

I asked this very question on Nef when I was thinking my VVT tensioner was in need of replacing...
VVT delete

There is also this...
Dilema - BAM content inside

...but I fear its not as straight forward as suggested due to most of Nef references being S4/ME7.1 related rather than ME7.5...

<tuffty/>

 

[Prawn]

M383 / 382 like in prawns are perfectly tunable but you need to remember that it isn't as clever in regards to it's learnt values, so if something goes wrong there is no compensation available to keep the car running without it leaning out and nuking a valve / piston for instance

Very true. 512 ECU's are dumb. Wellys ME7 is better than my 3.8, but that's still fairly dumb too from what I can gather.

What are peoples thoughts on tuning WB vs Non WB?

Niki told me when he mapped mine, that he set the fuel request at the top end to run 11.9 AFR, but if it was tuning a WB ECU he would request 12.4. he said he kept it more rich on mine as a safeguard should anything start to go out of tune....

 

[Westy]

Well well well, this thread has come a long way since the N249 delete on page 1 lol. Very interesting stuff though.

 

[badger5]

Agree 100% ... It seems and hasn't really been confirmed anywhere that you can't delete thè egt, you can leave it plugged in but can't actually code it out completely.

despite what prj might have said on nef?

"just a few bytes to code it out he says" - but does'nt say which ones.. go figure !?!?!

 

[badger5]

I think the reason for this is that the EGT probe (and VVT) is crucial to the strategies and models used by this particular code... removing them physically means the code can no longer function as intended...

The feedback from the EGT probe and the function of the VVT in context of the 210/225 maps is built into the execution of the code of these particular maps and without them the program logic is essentially broken...

There are fail safes of course hence the rich fuelling... VVT is controlled (from what I understand) by a system constant SY_NWS:
SY_NWS: camshaft adjustment
= 0: No NWS
= 1: 2-point NWS
= 2: steady NWS
> 2: not defined

I asked this very question on Nef when I was thinking my VVT tensioner was in need of replacing...
VVT delete

There is also this...
Dilema - BAM content inside

...but I fear its not as straight forward as suggested due to most of Nef references being S4/ME7.1 related rather than ME7.5...

<tuffty/>

I wonder what input the ecu is seeing for egt... we knows its an "Audi egt sensor" which is most likely a type K thermocouple, but goes into an audi "box of electronics" which will be converting egt into... more than likely a 0-5v output... So.... you see where I am going with this? wonder if you can log actual sensor egt voltage on the neflogger to see what voltage correlates to what temp, and then just frig the 0-5v input with a resistor to give it "some volts" to keep it happy.. Well, happy somethings connected, so coding out implausable's will likely be required also.

worth a punt for someone with some time on their hands.. ~(not me)

 

[Westy]

worth a punt for someone with some time on their hands.. ~(not me)

Says the man with enough time to mix methanol, screen wash, and salt together then set fire to it

 

[Broken Byzan]

I'll give it a go, but wont be for a few months yet when i upgrade to WB, the less parts i have to buy, the better,lol

 

[beachbuggy]

I wonder what input the ecu is seeing for egt... we knows its an "Audi egt sensor" which is most likely a type K thermocouple, but goes into an audi "box of electronics" which will be converting egt into... more than likely a 0-5v output... So.... you see where I am going with this? wonder if you can log actual sensor egt voltage on the neflogger to see what voltage correlates to what temp, and then just frig the 0-5v input with a resistor to give it "some volts" to keep it happy.. Well, happy somethings connected, so coding out implausable's will likely be required also.

worth a punt for someone with some time on their hands.. ~(not me)

I think i may ( i say may....) have found the missing byte to disable the EGT

From FR

Procedure:
===============
===
Switch off the function:
========================
Prohibition of the exhaust gas temperature control: code word CATR [bit0] = 0.
Affected functions:
========================
¨ Lamko% over dlamatr_w, dlamatr2_w


This isn't noted on Nef anywhere, but I have looked at all the EGT maps I have and this on all of them is set to 1 where as on the non Egt ecu's such as the HN , HJ etc it's set to 0.

reading through FR as well it looks like this along with

CATR FW configuration byte exhaust temperature control
FW DLATRMN minimum exhaust temperature limit controller
DLATRNLN NMOT KL delta lambda in emergency exhaust temperature control
KFRLATR NMOT TANS KF Minimum load f ¨ ur exhaust gas temperature control
Latro FW limit lambda f ¨ ur exhaust gas temperature control
SY_STERVK SYS (REF) system constant condition: stereo from Kat
TABGSS FW exhaust temperature setpoint f ¨ ur exhaust gas temperature control

Interestingly DLATRMN is the min temp for egt control, so this could potentially be set for a higher value and KFRLATR is for min load for temp control,

 

[badger5]

Says the man with enough time to mix methanol, screen wash, and salt together then set fire to it

for the common Shared info pool I might add

 

[BENJAMIN]

After hearing about the legendary 'karltech v1" exhaust system. We made Reesy's car sound less g*y

 

[beachbuggy]

CATR deletes EGT... Problem solved.... I guess this was the missing byte. I would also stick my neck out and say as its 0 on all the non egt ECU's that setting it to 1 and setting the other bits accordingly, along with adding wiring to the ecu you can add EGT if you wished.

 

[<tuffty/>]

CATR deletes EGT... Problem solved.... I guess this was the missing byte. I would also stick my neck out and say as its 0 on all the non egt ECU's that setting it to 1 and setting the other bits accordingly, along with adding wiring to the ecu you can add EGT if you wished.

Nice work... some testing needed maybe to double check what's going on and if normal egt models still work as intended

<tuffty/>