S3 doing my head in, still not running properly

[idnan]

Ok so from the beginning. I bought a 132k on the clock 2001 s3 about 6 weeks ago with a cracked manifold. New manifold was fitted a long with headgasket, cambelt, water pump and a full set of gaskets for the turbo and exhaust etc. I also fitted new vacuum hoses. Mods to the car are a full PCV valve delete, n249 delete, SFS oversized TIP, 3" DP, Forge recirc valve, upgraded catback and S2000 filter.

Now the car has had a stage 1 map fitted and only produced 230hp and 260lbft on the dyno and was accompanied by thick black smoke. I did notice a kinked hose from my simplified PCV setup which has since been rectified.

So symptoms are black smoke, low power and an unsteady idle. The PCV hoses have all been tidied up and anything unnecessary has been removed as per the guide. Currently venting to atmosphere.


Doing diagnostic work I realised the following:

Rear o2 sensor has been unplugged and removed completely by PO.

Considering this is the hottest day of the year, I noticed that the IAT didnt go below 50C no matter how I drove it.

I did a log and noticed up to 10 degrees of timing retardation among all 4 cylinders under WOT pull in 3rd

Mass airflow sensor only showing 170g/s at redline and 4.5g/s at idle.

Removing the MAF sensor plug the car runs exactly the same. Plugging it back in whilst idling shows no change at all.

o2 sensor ageing degradation (dynamic factor) was 0.002 when it is supposed to be between 0.5 and 2.0

Boost achieves target and seems to hold there

Smoke has appeared to have cleared up now, on last few pulls I noticed no smoke at all.

Turbo has slight up down play in shaft, approx 1mm

Compression test with throttle closed (forgot to press accelerator) dry was 145, 120, 150, 145 with a very cheapy am tech tester. Waiting for my decent one to come before repeating.

Emissions do smell very much like petrol (possibly because of decat?)

Completely unrelated but the brakes are very spongey

Any ideas?? I'm really pulling my hair out. Car seems to be making boost but not the power from what it appears to be timing pull.

 

[ayfreetee]

Full brake and clutch fluid replacement would be first for the brakes, needs doing every two years anyway, compression seems a bit low, especially no 2, may need some new rings and honing the cylinders, did you check the head for cracks when it was off?

 

[Stuart B]

cam belt out by a tooth?

I think mine is advanced by one tooth and I am now only getting 220 HP after timing belt, head gasket, water pump and 3inch downpipe. it was 243 before - I also get a co2 fault occasionally and a cam shaft position sensor error every 100 miles or so - need to move it back a tooth when I get a chance. that might explain the black smoke and timing ****** ? also I cannot get more than 180 g/s MAF and that was 190(ish) before

 

[idnan]

Full brake and clutch fluid replacement would be first for the brakes, needs doing every two years anyway, compression seems a bit low, especially no 2, may need some new rings and honing the cylinders, did you check the head for cracks when it was off?

Worn rings is one of my worries, hopefully getting the new compression gauge today to re do the rest. Yep head was checked thoroughly and couldnt find any issues.

cam belt out by a tooth?

I think mine is advanced by one tooth and I am now only getting 220 HP after timing belt, head gasket, water pump and 3inch downpipe. it was 243 before - I also get a co2 fault occasionally and a cam shaft position sensor error every 100 miles or so - need to move it back a tooth when I get a chance. that might explain the black smoke and timing ****** ? also I cannot get more than 180 g/s MAF and that was 190(ish) before

Seems we are in pretty much the same boat. Ive got no codes except the post cat lambda which is obviously missing. My brother checked the measuring blocks on vcds after the cam belt and said the timing is fine.

 

[idnan]

As a very long shot I've noticed that in some cars upgraded engine mounts can cause knock sensors to act funny. Currently I have polytekd mounts and the prop is touching the exhaust on full boost and scraping. I was thinking maybe this was causing the engine to hear knock when it wasn't actually. Maybe I'm clutching at straws?

 

[Stuart B]

mine was stage 1 mapped "pre" Badger5 TIP, PCV Delete with catch can, N249 delete and also pre 3 inch downpipe with sports cat - I am still floating the idea that adding the 3inch downpipe may have lost me some power (if unrelated to timing belt) and perhaps it needs remapping to cater for easier breathing.

Do you have a dog bone engine mount upgrade? I think you need those for most 3inch downpipe upgrades - I only changed the easy bush on mine and it seems there's plenty of clearance. (although it does have a little vibration at idle and slow 1st gear)

 

[idnan]

mine was stage 1 mapped "pre" Badger5 TIP, PCV Delete with catch can, N249 delete and also pre 3 inch downpipe with sports cat - I am still floating the idea that adding the 3inch downpipe may have lost me some power (if unrelated to timing belt) and perhaps it needs remapping to cater for easier breathing.

Do you have a dog bone engine mount upgrade? I think you need those for most 3inch downpipe upgrades - I only changed the easy bush on mine and it seems there's plenty of clearance. (although it does have a little vibration at idle and slow 1st gear)

Powerflex red Dogbone is in the post and should be with me today or tomorrow. I've tried adjusting the DP as much as possible and it must a cheapy chinese that the PO bought as I can't get it more than about 8mm from the prop. It seems on full boost it causes the flexy to compress and touch the prop. It feels the only way to move it away from the prop is to get an extra gasket and spacer for the dp flange.

 

[scotty_24]

Ive heard of this too, but many including myself , @Prawn etc use polytek engine mounts and have no knock issues

 

[badger5]

Powerflex red Dogbone is in the post and should be with me today or tomorrow. I've tried adjusting the DP as much as possible and it must a cheapy chinese that the PO bought as I can't get it more than about 8mm from the prop. It seems on full boost it causes the flexy to compress and touch the prop. It feels the only way to move it away from the prop is to get an extra gasket and spacer for the dp flange.

if its moving that much your mounts are too soft.... Not just the dogbone to do.

 

[scotty_24]

if its moving that much your mounts are too soft.... Not just the dogbone to do.

Polytek time!


Ive got a spare set of mounts you could polytek idnan pm me

 

[idnan]

if its moving that much your mounts are too soft.... Not just the dogbone to do.

Cheers Bill, appreciate your help. You built my brother's LCR engine when he went hybrid, Imran from Birmingham but he works not far from you.

I've used that Polytek stuff on the gearbox and the engine mount to try and stop the movement. Hopefully that and the powerflex dogbone when installed will help

 

[idnan]

Polytek time!


Ive got a spare set of mounts you could polytek idnan pm me

Already polytek'd cheers mate. Just need to do a slight bit more on the engine mount as I ran out!

 

[scotty_24]

Awesome mod I just hope
It doesn't start giving knock issues.
So far so good, my car went round Croft like a bat out of hell a few weeks ago with no issues

 

[vag-tech]

Even if knock sensors have been out putting them back in and over tightening can cause serious issues.

Otherwise, I'm thinking basics here:

• Stick a vacuum gauge on and see what it pulls at idle
• Confirm no intake tract air leaks
• Confirm timing (even with no cam sync errors)
• Lambda is critical for closed loop fuelling so if reading low - i.e. lean "go rich", it won't take long to hit rich stop

 

[<tuffty/>]

Rear Lambda (Bank 1 Sensor 2) is for CAT efficiency... these are commonly coded out when modding cars with decats etc... they are not used for main fuel control, thats the pre cat (Bank 1 Sensor 1) but they need to be electrically connected for adaptions to work... the map software expects there to be a sensor connected even if the DTC's and efficiency threshold values have been coded out... this can cause over fuelling if there is another issue that would cause adaptions to not be able to take fuel out..

A 2001 S3 is a wideband ECU so it will be a sensor issue if its overfuelling as wideband ECU's are very capable of achieving requested lambda even up to its 25% adaption limit... black smoke is representative of an over rich AFR...

Compressions seem a little low...

You haven't posted fault codes or that you have checked for any so a scan with VCDS should be top of the list

There is also the possibility the map you have is causing the issue... was the car logged while mapping? what was the AFR?

If you have access to full VCDS then logging fuelling would be worth while to see what requested AFR is

<tuffty/>

 

[vag-tech]

Rear Lambda (Bank 1 Sensor 2) is for CAT efficiency... these are commonly coded out when modding cars with decats etc... they are not used for main fuel control, thats the pre cat (Bank 1 Sensor 1) but they need to be electrically connected for adaptions to work... the map software expects there to be a sensor connected even if the DTC's and efficiency threshold values have been coded out... this can cause over fuelling if there is another issue that would cause adaptions to not be able to take fuel out..

That makes sense...I wondered if they were mapped out.

 

[idnan]

Even if knock sensors have been out putting them back in and over tightening can cause serious issues.

Otherwise, I'm thinking basics here:

• Stick a vacuum gauge on and see what it pulls at idle
• Confirm no intake tract air leaks
• Confirm timing (even with no cam sync errors)
• Lambda is critical for closed loop fuelling so if reading low - i.e. lean "go rich", it won't take long to hit rich stop

Rear Lambda (Bank 1 Sensor 2) is for CAT efficiency... these are commonly coded out when modding cars with decats etc... they are not used for main fuel control, thats the pre cat (Bank 1 Sensor 1) but they need to be electrically connected for adaptions to work... the map software expects there to be a sensor connected even if the DTC's and efficiency threshold values have been coded out... this can cause over fuelling if there is another issue that would cause adaptions to not be able to take fuel out..

A 2001 S3 is a wideband ECU so it will be a sensor issue if its overfuelling as wideband ECU's are very capable of achieving requested lambda even up to its 25% adaption limit... black smoke is representative of an over rich AFR...

Compressions seem a little low...

You haven't posted fault codes or that you have checked for any so a scan with VCDS should be top of the list

There is also the possibility the map you have is causing the issue... was the car logged while mapping? what was the AFR?

If you have access to full VCDS then logging fuelling would be worth while to see what requested AFR is

<tuffty/>

That makes sense...I wondered if they were mapped out.

Thank you very much for your feedback guys.

First of all only fault code present is:
17526/P1118/004376 - Oxygen (Lambda) Sensor Heating; B1 S2: Open Circuit

I told the mapper to code out the o2 sensor but the mil and fault code was still present. The car was logged quite a few times but with only 12 channels at a time I've only been looking at boost, timing, iat etc. My next logs are going to be looking at fuel pressure, afr etc. The mapper is highly regarded by nikki@r-tech, they send each other work when they're busy on a regular basis.

Does the iat seem high for moderate driving at decent speed, even though it was the hottest day of the year? 50c seems really high.

Thankfully my brother has broken 30 cupra r's and a single s3 so has millions of bits left over. And a rear o2 sensor is something he has and will be bringing it to me on Friday.

I'm going to give the map, maf and iat sensors a clean as it seems the turbo has been replaced for one without a egt sensor (new sensor in dp). I've also seen a little oil in the ic pipe joining the smics which also points to a previously blown turbo.

I'm also pricing up a wellycooler bar and plate type intercooler and will get it as soon as I have sorted this issue out.

Thanks again guys

 

[Stuart B]

I think 8mm clearance sounds about right down in the gully with my 3in DP, I had to cut off the downpipe hanger but put a new rubber on the post cat idle section and changed the dog bone mount bush - don't know how long it will last though as it feels a bit "vibratey" at low speeds, it makes the whole engine feel "dry" as if there is not enough oil in even though its a mount. best of luck with your other bits - can I ask whether it was mapped post the upgrades as you have more than stage 1 components and I asked wondering whether these cars should be mapped again if this amount of changes are made.

 

[<tuffty/>]

Thank you very much for your feedback guys.

First of all only fault code present is:
17526/P1118/004376 - Oxygen (Lambda) Sensor Heating; B1 S2: Open Circuit

I told the mapper to code out the o2 sensor but the mil and fault code was still present. The car was logged quite a few times but with only 12 channels at a time I've only been looking at boost, timing, iat etc. My next logs are going to be looking at fuel pressure, afr etc. The mapper is highly regarded by nikki@r-tech, they send each other work when they're busy on a regular basis.

Does the iat seem high for moderate driving at decent speed, even though it was the hottest day of the year? 50c seems really high.

Thankfully my brother has broken 30 cupra r's and a single s3 so has millions of bits left over. And a rear o2 sensor is something he has and will be bringing it to me on Friday.

I'm going to give the map, maf and iat sensors a clean as it seems the turbo has been replaced for one without a egt sensor (new sensor in dp). I've also seen a little oil in the ic pipe joining the smics which also points to a previously blown turbo.

I'm also pricing up a wellycooler bar and plate type intercooler and will get it as soon as I have sorted this issue out.

Thanks again guys

There are (apparently) ways to map the rear sensor out completely but seems very hit and miss from what I have seen and read so generally prefer to leave them electrically connected... you only have to code the the efficiency value and DTC for the most part anyway which fixes it for decats etc...

Fuel pressure will need a mechanical gauge to measure... fuel pressure on 1.8t's isn't logged through the port... one issue 1.8t's have is from blocked FPR's... the rubber in the fuel hoses rots off and collects in the FPR bowl blocking flow and causing higher fuel pressure... this is normally associated with a fault code indicating 'system too rich' (I can't recall the exact code off hand) and is where the ecu is trying to adapt out the excessive fueling and gone over its 25% limit... but seeing as adaptions may not be working on your car anyway there is no way for the ecu to adapt the increased fuel out and so it ends up out the exhaust hence the black smoke...

Assuming you are on standard SMIC's then 50 deg or more sounds normal in this weather for a mapped car... the std SMIC's can just about manage std power... with mapped cars they are next to useless

Having the EGT probe in the downpipe is ok, its how you have to do it on big turbo builds but obviously needs map tweaks if its to do its job properly as EGT's will be lower there than in its normal place which will mess up fuelling strategies to a degree (in particular component protection)... R-Tech's mapping should account for this assuming they knew about the EGT probe being in a different place... if not its probably worth mentioning it

<tuffty/>

 

[idnan]

There are (apparently) ways to map the rear sensor out completely but seems very hit and miss from what I have seen and read so generally prefer to leave them electrically connected... you only have to code the the efficiency value and DTC for the most part anyway which fixes it for decats etc...

Fuel pressure will need a mechanical gauge to measure... fuel pressure on 1.8t's isn't logged through the port... one issue 1.8t's have is from blocked FPR's... the rubber in the fuel hoses rots off and collects in the FPR bowl blocking flow and causing higher fuel pressure... this is normally associated with a fault code indicating 'system too rich' (I can't recall the exact code off hand) and is where the ecu is trying to adapt out the excessive fueling and gone over its 25% limit... but seeing as adaptions may not be working on your car anyway there is no way for the ecu to adapt the increased fuel out and so it ends up out the exhaust hence the black smoke...

Assuming you are on standard SMIC's then 50 deg or more sounds normal in this weather for a mapped car... the std SMIC's can just about manage std power... with mapped cars they are next to useless

Having the EGT probe in the downpipe is ok, its how you have to do it on big turbo builds but obviously needs map tweaks if its to do its job properly as EGT's will be lower there than in its normal place which will mess up fuelling strategies to a degree (in particular component protection)... R-Tech's mapping should account for this assuming they knew about the EGT probe being in a different place... if not its probably worth mentioning it

<tuffty/>

Redid the compression test and got:

181ps 173psi 180psi 190psi

Slightly low on cylinder 2 but much more consistent and what is expect for the compression ratio of a BAM engine. So can hopefully don't have to worry about the block/head when going forwards in diagnosing this nightmare car.

Good point about the SMIC, will be looking to do a custom wellycooler style setup as I have done on my 335i when this car is sorted.

335i is 500bhp/500lbft with custom prop and m3 differential as well as a few other bits....

 

[ayfreetee]

Wow, a tyre shredding monster I bet.

 

[idnan]

Wow, a tyre shredding monster I bet.

Just a bit, makes me grin every time I take it out

 

[idnan]

Another day, another problem.

Immediately after putting the car back together from the previous compression test I didnt start the car or move it. I did however fit a balljoint, track rod end and also a powerflex dogbone mount.

Took it out for a test drive and as soon as I hit 2000rpm and some boost, the whole car stutters and shakes. It does this each and every time. This is a bit strange considering the plugs are less than 400 miles old and the TFSI coils have done about 15k on them. I replaced the fuel pump for a 265lph which again has made no difference. With the n75 valve harness plug taken off the car runs fine as it doesn't come onto boost.

Any ideas where to start looking?

 

[Stuart B]

does the MIL flash occasionally flash (hard fault) ? You have described how mine behaved when it was misfiring due to a faulty coil which blew because of a faulty spark plug, but you said N75 disconnected doesn't have any problems - could it still be related to plugs and coils but only with massive power increase? maybe if the spark is simply not good enough to ignite the fuel?

 

[superkarl]

Who mapped it?

Sounds like it's misfiring. There might be a code. Or you can monitor misfires real time.

Also, even tho it's making the right airflow for the power it's made, you said there's no difference in the way it drives with maf plugged in vs unplugged. They're very sensitive things and might be worth getting a new one

 

[badger5]

my bets on coils

 

[idnan]

does the MIL flash occasionally flash (hard fault) ? You have described how mine behaved when it was misfiring due to a faulty coil which blew because of a faulty spark plug, but you said N75 disconnected doesn't have any problems - could it still be related to plugs and coils but only with massive power increase? maybe if the spark is simply not good enough to ignite the fuel?

No flashing MIL at all, have read it on other forums. Spark plugs are only 500miles old and are the ones recommended for mapped 1.8t's. Saying that I didn't gap them and they have been out a couple of times so will give do that tonight.

Who mapped it?

Sounds like it's misfiring. There might be a code. Or you can monitor misfires real time.

Also, even tho it's making the right airflow for the power it's made, you said there's no difference in the way it drives with maf plugged in vs unplugged. They're very sensitive things and might be worth getting a new one

TRP Developments. Nikki at R-tech rates him very highly and he is literally around the corner from mine. To be fair he thinks its plugs or coils so seems to be in line with what others are saying.

my bets on coils

Im leaning that way too.

My brother has a Cupra R and 1.8t Golf Gti so will pop to his after work tomorrow and start swapping bits. Think I'll try coils, MAF and n75 just for completion. I'm going to steal his ultrasonic cleaner and clean my injectors and maybe the FPR too.

 

[idnan]

OK so now we are getting somewhere. First of all, Bill was right and it was a coilpack so that was sorted fairly easily. Also managed to borrow a nearly new MAF and the datalogs were showing 200g/s plus and peaked at 217 g/s with that one so it's making healthy power. I have ordered one from TPS and will be getting it today. Managed to quickly plumb in a temporary boost gauge and it shows 20psi peak.

Unfortunately though I still have one problem and it feels like its a biggie. Under boost i am getting a horrible scraping and rattling noise from the turbo area. Also coming off boost I get a rattlesnake like noise. I tried taking videos but with the 3" DP and no cats the exhaust noise drowns it out. I'm thinking its the wastegate flapper or the DV. Having opened my Forge 007P, it has an orange spring inside it? I have googled everywhere to see what it is but to no avail. The piston is moving freely and when i push it with my finger and cover the vac port at the top the piston holds in position then pops back to its resting place when released. I've decided to get an OEM one anyway but TPS were out of stock on the DV so have ordered one from Audi today. Having already taken the head off once for the manifold I'm dreading that the turbo is fudged.