Turbo problems....

[warren9022]

This is my first post so be gentle.....

My S3 has been suffering with a lack of boost for a few months now, changed the usual suspects (N75, MAF etc). Pressure tested and no leaks. Did the rocker cover and cam tensioner gasket today so decided to strip everything back to get at the turbo. Turbo spins freely and has no play. As a last resort I disconnected the hose from the N75 to the waste gate actuator and went for a (very cautious) drive. Put my foot down in the higher gears and the boost started to surge but then stopped and started to judder like the boost was coming and going.

No fault codes show on the VAG COM.

Can anyone tell me what this points to please?

I've had the car for 13 years and she's like one of the family but like most family members at some point she's really testing my patience now!!

Thanks in advance.

 

[<tuffty/>]

As a last resort I disconnected the hose from the N75 to the waste gate actuator and went for a (very cautious) drive. Put my foot down in the higher gears and the boost started to surge but then stopped and started to judder like the boost was coming and going.

Uncontrolled boost at least suggests that the turbo is capable... I assume you don't have a boost gauge so no telling what boost you are running...

You ideally need to log boost and a few other blocks to work out whats happening...

Not really the best way to test without a gauge if I am honest as boost onset like that can bend rods...

<tuffty/>

 

[Stuart B]

is it only when the snowflake is displayed in the DIS ?

is it exactly the same with the N75 electrically connected and without the N75 electrically connected - that is about 6.5 psi you wont notice it, but there is a thread on here with electrical faults going to the N75

 

[warren9022]

Uncontrolled boost at least suggests that the turbo is capable... I assume you don't have a boost gauge so no telling what boost you are running...

You ideally need to log boost and a few other blocks to work out whats happening...

Not really the best way to test without a gauge if I am honest as boost onset like that can bend rods...

<tuffty/>

Thanks Tuffty,
Have got a boost gauge here ready to go on as I've read all the similar threads and realise that they are a crucial piece of kit when trying to diagnose. Hoping to get it installed next week so I'll see if it sheds some more light on it. Only had the VCDS a day so I'll try to work out how to log boost on it and post results as soon as I do.
Yeah I've read all the horror stories about running ungoverned boost but I was careful and wanted to reassure myself that the turbo was still in working order.
I've thrown that much money and time at it lately and I just want to get it fixed.
Thanks again.

 

[warren9022]

is it only when the snowflake is displayed in the DIS ?

is it exactly the same with the N75 electrically connected and without the N75 electrically connected - that is about 6.5 psi you wont notice it, but there is a thread on here with electrical faults going to the N75

No it's all the time, feels like the mini I had when I was a kid....looks good but no power!!

I'll try unplugging the N75 and see what happens. If it performs the same connected or not connected I'll find the thread about electrical faults and start diagnosing.

Thanks Stuart.

 

[Stuart B]

if you have a ODB2 Bluetooth Reader and Android Torque Pro you can record the requested boost (VAGCOM / VCDS) can also do this (but that is not a fiver) - at least then you can compare the boost to see if it is zero or as if always unplugged 6.5psi or so. I think they are about 120 BHP with the N75 unplugged - which sounds like a lot for a small hatchback but feels like its dragging an anchor.

when I bought my S3 the N75 was tucked away and the previous owner had "stupidly" put on a manual boost controller between the charge pipe and actuator - if you are going to do non electrically controlled turbo testing then I would consider this option (they are frowned upon on this 'ere forum due to the dangers of stressing the engine internals if badly adjusted)

a turbo essentially turns an engine into a motorcycle 2 stroke "powerband" - it is either on or off

I loved my mini 1100 special with walnut dash and weller wheels, it felt like it was really quick but my girlfriend overtook me easily with her 1979 1.1 L fiesta

 

[<tuffty/>]

No it's all the time, feels like the mini I had when I was a kid....looks good but no power!!

I'll try unplugging the N75 and see what happens. If it performs the same connected or not connected I'll find the thread about electrical faults and start diagnosing.

Thanks Stuart.

Electrical fault on N75 will show a fault code mate... you will need to log boost request vs actual (block 115) and N75 duty (block 118)... log block 003 too with this lot to see what the ignition and throttle plate is doing..

List of logging blocks...
http://www.ross-tech.com/vag-com/m_blocks/

Info on how to log is in the stickies (the "read before posting thread")

<tuffty/>

 

[warren9022]

Thanks Tuffty,

Didn't know about the stickies so apologies for not looking there first.

Will do some logs and post the results soon.

 

[warren9022]

if you have a ODB2 Bluetooth Reader and Android Torque Pro you can record the requested boost (VAGCOM / VCDS) can also do this (but that is not a fiver) - at least then you can compare the boost to see if it is zero or as if always unplugged 6.5psi or so. I think they are about 120 BHP with the N75 unplugged - which sounds like a lot for a small hatchback but feels like its dragging an anchor.

when I bought my S3 the N75 was tucked away and the previous owner had "stupidly" put on a manual boost controller between the charge pipe and actuator - if you are going to do non electrically controlled turbo testing then I would consider this option (they are frowned upon on this 'ere forum due to the dangers of stressing the engine internals if badly adjusted)

a turbo essentially turns an engine into a motorcycle 2 stroke "powerband" - it is either on or off

I loved my mini 1100 special with walnut dash and weller wheels, it felt like it was really quick but my girlfriend overtook me easily with her 1979 1.1 L fiesta

I've bitten the bullet and ordered the registered version of VCDS LITE, not the cheapest thing in the world but figure I will get my money's worth out of it the older the car gets!! Just waiting for the activation email and I'll start logging.

The car feels like your girlfriend's 1979 1.1 Fiesta at the moment mate, I've had it from 3 years old (2004) so I know exactly how it should perform and it's really frustrating not being able to access the performance that I know is hiding somewhere under the bonnet!!

When I disconnected the hose from the N75 to the wastegate actuator it threw up an overboost code so I'm guessing if the turbo is capable of 'overboosting' there can't be much wrong with it and I need to be looking in other areas?! Hopefully the VCDS and boost guage will confirm this.

A couple of things that worry me......When I had the charge pipe etc stripped off to change the valve cover and camshaft tensioner gaskets it gave me good access to the wastegate actuator. I could just about pull the rod with a pair of mole grips and all my strength, do you know if this is normal? It sprung back into place quickly with an audible click so (again) I'm guessing the spring and diaphragm are doing what they should. I just never thought it would take that much force to move the rod. I've got a pressure hand pump on order to hook it up and test it properly but was just wondering if you'd had any experience with the actuator?

Also the coil packs have only ever been changed on the recall when it was 2 years old, is it pushing it to think that coil packs would last this long? (Car has now covered 104,000 miles). There's no fault codes showing up for the coil packs (but on the other hand there's no fault codes showing for anything and I know there's a problem somewhere!!). Could the coil packs be breaking down when put under heavy acceleration and not throwing a code?

Sorry for the essay but I don't get to talk 'engines' with like minded people very often!!

Thanks in advance.

 

[<tuffty/>]

Thanks Tuffty,

Didn't know about the stickies so apologies for not looking there first.

Will do some logs and post the results soon.

no probs... just re-read what I posted and it reads more like a telling off lol... I have edited it to make it read less 'get on the naughty step'

When I disconnected the hose from the N75 to the wastegate actuator it threw up an overboost code so I'm guessing if the turbo is capable of 'overboosting' there can't be much wrong with it and I need to be looking in other areas?! Hopefully the VCDS and boost guage will confirm this.

If you are getting more boost than the ECU expects at anyone time that will throw an overboost code... doesn't necessarily mean the turbo is good though as overboost is 200mbar over request for a given time... you still haven't said what boost you have seen when you tried this though...

Logging with VCDS will help to gives clues as to whats going on... it will be the data the ECU is seeing so will be able to determine whats actually happening and go from there... just be aware that VCDS will only see MAP sensor readings and is limited to 2550mbar (MAP sensor reading is absolute pressure so ambient pressure of around 1bar + boost)... what you are looking for is what the ECU is asking for boost wise and if the actual boost can get close to it... when logged with N75 duty you can see how hard the ECU is trying to meet the boost request... also note that ME7 ECU's don't have a map for boost... boost is derived from load request so the ECU calculates how much boost it will need to make the engine produce the load being asked of it from the driver input (throttle pedal)... this is why you often see more boost in the summer vs the winter due to air density differences...

A couple of things that worry me......When I had the charge pipe etc stripped off to change the valve cover and camshaft tensioner gaskets it gave me good access to the wastegate actuator. I could just about pull the rod with a pair of mole grips and all my strength, do you know if this is normal? It sprung back into place quickly with an audible click so (again) I'm guessing the spring and diaphragm are doing what they should. I just never thought it would take that much force to move the rod. I've got a pressure hand pump on order to hook it up and test it properly but was just wondering if you'd had any experience with the actuator?

Sounds normal... using a mity vac type pump you should see a std actuator 'crack' open (start to move) around 0.5bar... if there was 'slack' in the rod that would cause lag and low boost... you would need to check the preload of the actuator then...

Also the coil packs have only ever been changed on the recall when it was 2 years old, is it pushing it to think that coil packs would last this long? (Car has now covered 104,000 miles). There's no fault codes showing up for the coil packs (but on the other hand there's no fault codes showing for anything and I know there's a problem somewhere!!). Could the coil packs be breaking down when put under heavy acceleration and not throwing a code?

Faulty coil pack would be fairly obvious... misfires under boost and multiple hits on the misfire counters... you would get a misfire code if nothing else

<tuffty/>

 

[warren9022]

no probs... just re-read what I posted and it reads more like a telling off lol... I have edited it to make it read less 'get on the naughty step'

The view from that step was starting to get boring.....

If you are getting more boost than the ECU expects at anyone time that will throw an overboost code... doesn't necessarily mean the turbo is good though as overboost is 200mbar over request for a given time... you still haven't said what boost you have seen when you tried this though...

I'm learning every day.....I didn't realise that overboost was measured in that way.

Logging with VCDS will help to gives clues as to whats going on... it will be the data the ECU is seeing so will be able to determine whats actually happening and go from there... just be aware that VCDS will only see MAP sensor readings and is limited to 2550mbar (MAP sensor reading is absolute pressure so ambient pressure of around 1bar + boost)... what you are looking for is what the ECU is asking for boost wise and if the actual boost can get close to it... when logged with N75 duty you can see how hard the ECU is trying to meet the boost request... also note that ME7 ECU's don't have a map for boost... boost is derived from load request so the ECU calculates how much boost it will need to make the engine produce the load being asked of it from the driver input (throttle pedal)... this is why you often see more boost in the summer vs the winter due to air density differences......

I've received the activation code for VCDS today so will try and get my head round it and do some logs tomorrow, I'll post them here and see what you think. Whats the best format to post them in, Excel or graph? Sorry if that's a dumb question!!

Sounds normal... using a mity vac type pump you should see a std actuator 'crack' open (start to move) around 0.5bar... if there was 'slack' in the rod that would cause lag and low boost... you would need to check the preload of the actuator then.........

I've got a Mityvac 8500 coming this week so will test the actuator properly when it arrives....

Faulty coil pack would be fairly obvious... misfires under boost and multiple hits on the misfire counters... you would get a misfire code if nothing else.........

I thought as much, clutching at straws a bit now I think. I'm pulling out what little hair I've got left .

Thanks again Tuffty, hopefully next time I'll have some hard facts to show you instead of just speculation.

 

[<tuffty/>]

Whats the best format to post them in, Excel or graph?

Excel is the best bet... see here...
http://www.audi-sport.net/xf/thread...lp-please-read-before-posting-updated.168407/

<tuffty/>

 

[warren9022]

Excel is the best bet... see here...
http://www.audi-sport.net/xf/thread...lp-please-read-before-posting-updated.168407/

<tuffty/>

Hi Tuffty,

Thought I'd take the opportunity on my lunch break to do some logging!! It's not as easy as it sounds to find a road where you can go full chat in third gear .

It might as well be in chinese to me but could you have a look and see if it makes sense to you please?

Many thanks.

 

[<tuffty/>]

Boost request looks standard (not mapped)... but actual boost is around 200/300mbar less...

What are the fuel trims like ? (block 032... no need to log them, just read off the values at idle for now)

<tuffty/>

 

[warren9022]

Boost request looks standard (not mapped)... but actual boost is around 200/300mbar less...

What are the fuel trims like ? (block 032... no need to log them, just read off the values at idle for now)

<tuffty/>

Lambda 0.2% Lambda 0.0% (Car is as standard as the day it rolled off the production line apart from PCV simplification, N249 deletion, catch can and silicone TIP)

 

[<tuffty/>]

ahh... have you recently done a code clear? as that resets fuel trims...

If you have then it might take a few miles of driving to pick up again if there is an issue...

Basically speaking I am trying to determine where the boost is going... could be an air leak (smoke test would be a good idea), could be a weak actuator (would expect the N75 duty to go up quite high) but...

Airflow is low (assuming the runs were foot to the floor) and the indicated load in block 115 seems low too plus timing looks quite high although in fairness its been a while since I have seen a completely std cars log file so I have no further frame of reference... std cars normally flow around 180gs... your highest figure is 142gs.... 142/0.8 gives 177.5hp based on airflow

It does look like the MAF could be under reading... this would affect all of this as the ECU is load based and the MAF is used to determine airflow/load..

Have you changed the MAF recently? if so is it a genuine Bosch one? Try unplugging the MAF and see if the car feels any different?

<tuffty/>

 

[warren9022]

ahh... have you recently done a code clear? as that resets fuel trims...

Yeah sorry, checked for codes and cleared them after I did the logging!! Will drive it for a few miles tonight, check again and post results.

Basically speaking I am trying to determine where the boost is going... could be an air leak (smoke test would be a good idea), could be a weak actuator (would expect the N75 duty to go up quite high) but...

I've pressure tested with this ..... (Can you tell I work in a plumbers merchant?)

From the TIP to the intake manifold upto 15psi and couldn't find any leaks, have either replaced or checked every hose to do with the turbo and can't find any leaks/split hoses etc. Do you think it's worth a smoke test?

Airflow is low (assuming the runs were foot to the floor) and the indicated load in block 115 seems low too plus timing looks quite high although in fairness its been a while since I have seen a completely std cars log file so I have no further frame of reference... std cars normally flow around 180gs... your highest figure is 142gs.... 142/0.8 gives 177.5hp based on airflow

I will do another run tonight and make sure I'm foot to the floor, I thought I was but don't want to give you any false readings so it's better to be on the safe side.

It does look like the MAF could be under reading... this would affect all of this as the ECU is load based and the MAF is used to determine airflow/load..

Have you changed the MAF recently? if so is it a genuine Bosch one? Try unplugging the MAF and see if the car feels any different?.

After getting a small mortgage....the MAF was brand new from Shrewsbury Audi a few weeks a go (it was the first thing I changed) along with the MAP sensor (and the very expensive plastic pipe that it sits in as the bolts just spun in the old one!!).

I will unplug the MAF on the drive home from work tonight and see if I can feel any difference.

could be a weak actuator

Hopefully the Mityvac will arive in the next few days so I can test it, is it a simple case of pumping pressure into it until it opens and then watching to see if it creeps back?

I have noticed a miss when idling these last couple of days but could be totally un-related.

Thanks for taking the time to help Tuffty, really appeciate it.

 

[<tuffty/>]

Do you think it's worth a smoke test?

Yes... best thing in the world for 1.8t's

Most of the leaks will be in the pipes under the inlet manifold...

I have noticed a miss when idling these last couple of days but could be totally un-related.

Hummm... compression test and leak down test would be next thing to do if the smoke test is clear but if you have tested up to the inlet manifold then my money is on all the pipework under it having a leak somewhere... 95% of leaks are there..

I post this up fairly regularly but this is all the pipework in question..

V pipe (orange) splits inside the V, all the other points it normally goes are marked on the pic... the hard plastic pipe shrinks from the heat then splits where it joins... most people bin the lot off in a PCV 'delete' (in the FAQs) or you can replace them with expensive aftermarket hoses or buy a new complete kit from the dealer (use ASN parts request, genuine parts but really good prices)

<tuffty/>

 

[warren9022]

Most of the leaks will be in the pipes under the inlet manifold...>

I didn't make myself very clear sorry, when I said a PCV simplification I meant a delete, all the pipework for the breather system and vacuum lines under the inlet manifold are now silicone and not very old. I had them all off when I stripped everything back and checked them for leaks.

Here's the VCDS results for the run I just logged, definitely WOT:

I also looked again at the FuelTrim again after about 5 miles of driving and the readings were 0.2%/-1.6%.

I removed the electrical plug from the MAF on the way home from work, it drove fine but didn't have as much power as with it plugged in.

A weird thing did happen though, for a while now the car has been idling at 1100RPM on cold start and dropping down to normal (800RPM) when it's warmed up a bit. When I started it from cold with the MAF unplugged it idled at 800RPM straightaway and stopped there.

Any thoughts Tuffty?

 

[<tuffty/>]

When I started it from cold with the MAF unplugged it idled at 800RPM straightaway and stopped there.

Any thoughts Tuffty?

Maybe a default in alpha-n mode when the ECU can't see airflow... can't say I have ever tried it

Compression and leak down test might be an idea...

<tuffty/>

 

[warren9022]

Maybe a default in alpha-n mode when the ECU can't see airflow... can't say I have ever tried it

Compression and leak down test might be an idea...

<tuffty/>

Thanks Tuffty, will get it booked in for a compression and leak down test.

 

[badger5]

its stuck on actuator pressure..?
any activity on block 118 for n75 duty cycle?