Just taken my DV off as I suspected it was split due to flat spots/hesitation around 2500rpm.  Turns out it's already a rev G (same as the replacement I bought) and isn't split or punctured anywhere that I can see.  Shall I fit the new one anyway?

 

The one that I've just taken off is branded pierburg, and there is no threadlock on the three bolts that I can see so I strongly suspect that it has been replaced already at some point...

 

There was as lot of oil around it so where should I look next? Can a DV still fail even with an intact diaphragm and no codes?

Edited 6 November, 20169 yr by planehazza

Yes, I made a post similar a while back. Mine has got quite a bit of oil around it, the mounting bolts in particular. I've cleaned it up and waiting to see how quickly it comes back. Mine is an April 2011 manufactured engine and mine is rev G in there as well. 

 

As to whether it could be faulty without being split I suppose unlikely. Maybe if the spring mechanism is not resistant enough it may not sustain pressure required.

 

Data logging on VCDS would help tell for sure or just replace it for piece of mind. 

 

Mine doesn't have the hesitation but given the oil present, the oring or gasket must be not sealing that well.

Yes, I made a post similar a while back. Mine has got quite a bit of oil around it, the mounting bolts in particular. I've cleaned it up and waiting to see how quickly it comes back. Mine is an April 2011 manufactured engine and mine is rev G in there as well. 

 

As to whether it could be faulty without being split I suppose unlikely. Maybe if the spring mechanism is not resistant enough it may not sustain pressure required.

 

Data logging on VCDS would help tell for sure or just replace it for piece of mind. 

 

Mine doesn't have the hesitation but given the oil present, the oring or gasket must be not sealing that well.

 

Yup, same here mate.  It was ****ing it down and I only had a crappy torch, but the oil that I can see appears to be coming from the seal between the DV and turbo on mine also.  It was noted on my MOT that there was a slight oil leak, but I was already aware it of.  It was one of the reasons I suspected my DV.  I've cleaned up all the old oil from the DV so I should be able to see it return.  I just refitted the old part and will return the new one as even the bolts on my DV were new so I'm almost certain it has been replaced in the past, probably for the same reason that I planned to. I'm going to go out with the laptop and VCDS now.  What logging should I enable aside from boost pressure etc?

I don't have VCDS yet but there have been a few posts fairly recently with turbo logging data to capture. You need requested boost  and delivered boost, against engine rpm.. Often use 3rd or 4th gear flat to the floor for the logging.

 

Edit: here is a post with a data log. It wasn't the one I was looking for but shows the plot anyway.

 

http://www.briskoda.net/forums/topic/412280-turbo-log-and-remap/

Edited 6 November, 20169 yr by TheClient

I don't have VCDS yet but there have been a few posts fairly recently with turbo logging data to capture. You need requested boost  and delivered boost, against engine rpm.. Often use 3rd or 4th gear flat to the floor for the logging.

 

Cool, I'll do that.  I'll take up the A1 ha.

Cool, I'll do that.  I'll take up the A1 ha.

Have fun.  Let us know how it looks. 

OK, so I have some data, but I'm not sure how helpful it is.  I logged the following:

• Accelerator position
• Throttle position
• Boost pressure (mbar)
• Misfire counter
• MAF

Here is the data from driving normally at 30 in forth, to sudden 100% accelerator, then back to normal again:

 

 

It peaks at 1620mbar, which is around 23PSI. Is this normal? It seems high to me?  Also, when I had my foot flat on the floor, I could hear a constant, quiet psssh sound, as if air was venting. It certainly sounded like it was coming from the turbo area.  Doing some reading, it could be an n75 failure? Any thoughts?

 

I'm not really sure how to use VCDS in great depth yet, and I think it would be easier if I had someone with me to start and stop the logging rather than just have one massive file! Ha

 

Here are the full logs in excel file type:

https://drive.google.com/drive/folders/0B9KSwu5pc5lzUlk1SGtENTl4Rlk?usp=sharing

Edited 6 November, 20169 yr by planehazza

Yes, it would be easier with someone to start logging but it is not a problem just filtering or cutting out the bit you want from your capture.

 

What you really need is RPM, Actual Boost and Requested Boost (at least) in the one capture.  I don't know the map well enough to know what would be requested at full loads at different RPMs on a standard map to compare against the delivered boost. Someone else might.. 

 

I'll have another look later on and see if I notice anything but actual vs delivered would be the acid test. Or if you can find a log of a standard vRS for requested boost you could infer from that.

This shows data and some inferences for a MK1 but is a good thread for a read and there is a good link to a PDF for data logging and plotting within VCDS within the threads (also pasted below for you).

 

http://www.briskoda.net/forums/topic/233329-crazy-boost-pressure/?p=2747119#entry2747119

 

http://www.vaglinks.com/OBDII/Vag-com_Data_Logging_And_Graphing.pdf

 

Your boost levels don't look a long way different (despite being for a different series car).  When looking at the boost levels you've recorded,  I don't immediately see a problem of lost boost. However it would be better with demanded boost (specified boost in VCDS).   Were there any runs with full throttle, most seem partial throttle and don't cover all the rev range to see how boost is behaving through the spectrum.

 

Others may offer more guidance, as i've said, I'm saying it how I read it and I don't have practical experience of VCDS...yet.

 

 

This shows data and some inferences for a MK1 but is a good thread for a read and there is a good link to a PDF for data logging and plotting within VCDS within the threads (also pasted below for you).

 

http://www.briskoda.net/forums/topic/233329-crazy-boost-pressure/?p=2747119#entry2747119

 

http://www.vaglinks.com/OBDII/Vag-com_Data_Logging_And_Graphing.pdf

 

Your boost levels don't look a long way different (despite being for a different series car).  When looking at the boost levels you've recorded,  I don't immediately see a problem of lost boost. However it would be better with demanded boost (specified boost in VCDS).   Were there any runs with full throttle, most seem partial throttle and don't cover all the rev range to see how boost is behaving through the spectrum.

 

Others may offer more guidance, as i've said, I'm saying it how I read it and I don't have practical experience of VCDS...yet.

 

 

Sadly I was using a crappy old laptop that kept restarting itself.  The log2.xlsx is actually the start of the journey.  In Log1, after mark 5 and 6, I was doing 100% accelerator (pinned it to the floor!) in 4th at 30mph.  When I hit 50mph I let go entirely, which is where it appears some of the misfire counts occurred.   I'll see about setting up another log tonight where I can compare the speed, rpm and boost figures.

Yes, saw the misfire counts. Have the usual suspects been eliminated. Plugs, coil packs?

 

Scanning through the different runs within the data, some of the boost levels peaking at up to 1570mbar at 2640 rpm seem quite high. ~23psi in old money.   And is higher than boost referenced in those other threads I've read, but without the demand curve (specified boost) for the map, I'm just not experienced enough to say it is ok or not.

 

Is it a standard map or remapped?

Yes, saw the misfire counts. Have the usual suspects been eliminated. Plugs, coil packs?

 

Scanning through the different runs within the data, some of the boost levels peaking at up to 1570mbar at 2640 rpm seem quite high. ~23psi in old money.   And is higher than boost referenced in those other threads I've read, but without the demand curve (specified boost) for the map, I'm just not experienced enough to say it is ok or not.

 

Is it a standard map or remapped?

 

Coils have never been changed as far as I'm aware, but I changed the plugs myself ab out 5k ago for ngk 8s IIRC.  The car is standard, and I don't think it has ever been mapped.

 

I'll get some logging tomorrow night

Is the hesitation the same as the misfires? Given your initial description, it seems they are occurring at the same time and are directly connected probably causing the hesitation?

 

Does VCDS record any misfire fault detail i.e. cylinder banks etc.?  You could then buy a single coil pack to swap out initially, or if you know somebody local with a similar car, you could ask to swap out for a short time to go round the block.....

 

Having read a bit more on boost pressure, it would seem you need to deduct atmospheric pressure, 1000mbar (14psi) from your actual boost data to give the real force pressure.  So, true peaks of about 8-10 psi and sustained of about 5-7psi is what you should expect from a similar motor from my research anyway.  The range depends on ambient temp and lower end for cooler temperatures.

 

Edit: PS measuring Block 11 is where you should find specified and actual boost per the referenced guide:

 

"TURBOCHARGER

Known as charge pressure control in measuring blocks 010 and 011.

 

Most of the important turbocharger information is found in block 011

E.g.

 

11. 1 Engine Speed, (G28)

11. 2 Boost Pressure (specified)

11. 3 Boost Pressure (actual)

11. 4 Charge Pressure, Control Duty Cycle %

 

At idle the boost pressure specified and boost pressure actual should be the same at about 1000 mbar."

 

Edited 7 November, 20169 yr by TheClient

Is the hesitation the same as the misfires? Given your initial description, it seems they are occurring at the same time and are directly connected probably causing the hesitation?

 

Does VCDS record any misfire fault detail i.e. cylinder banks etc.?  You could then buy a single coil pack to swap out initially, or if you know somebody local with a similar car, you could ask to swap out for a short time to go round the block.....

 

Having read a bit more on boost pressure, it would seem you need to deduct atmospheric pressure, 1000mbar (14psi) from your actual boost data to give the real force pressure.  So, true peaks of about 8-10 psi and sustained of about 5-7psi is what you should expect from a similar motor from my research anyway.  The range depends on ambient temp and lower end for cooler temperatures.

Thanks mate. This is where it typically goes over my head. I understand the basic concept but understanding the data and what I need to monitor is a bit above me, so I appreciate your help greatly.

I did wonder about the ambient pressure mind as I remember reading when people getting boost gauges fitted, they panicked when they saw 30psi and didn't realise that was including atmospheric pressure also.

I'll go out tonight with a better laptop (the perks of working in IT) and log all the necessary data.

Regarding misfire, I originally had all 4 cylinder misfire counts checked, but I didn't realise I was limited to 12 logging items, so I removed them and did the overall misfire count. I'm using the advanced measuring blocks but there is a lot of repetition so I wonder if I'm picking the right sensor...

Regarding misfire, I originally had all 4 cylinder misfire counts checked, but I didn't realise I was limited to 12 logging items, so I removed them and did the overall misfire count. I'm using the advanced measuring blocks but there is a lot of repetition so I wonder if I'm picking the right sensor...

I think what you're doing is ok as far as I can tell. Misfire counters are in Group 15, 16.

 

If there is a limit to data capture, you'll just have to concentrate on what is important or what your present focus area is and if necessary repeat later with alternative views.

 

So as a minimum, Engine RPM, Actual Boost, Specified Boost, Charge Duty Cycle. That is 4. The four misfire counters make 8.[ Edit: I suppose time sequence is compulsory so actually may leave 3] Then you've got up to 4 more for anything else you are interested in. If you can't fit it all in you'd like, you'll have to do multiple captures I guess. 

 

If you do a full diagnostic scan after misfires does VDCS show any fault codes stored and descriptions?

Edited 7 November, 20169 yr by TheClient

I think what you're doing is ok as far as I can tell. Misfire counters are in Group 15, 16.

 

If there is a limit to data capture, you'll just have to concentrate on what is important or what your present focus area is and if necessary repeat later with alternative views.

 

So as a minimum, Engine RPM, Actual Boost, Specified Boost, Charge Duty Cycle. That is 4. The four misfire counters make 8.[ Edit: I suppose time sequence is compulsory so actually may leave 3] Then you've got up to 4 more for anything else you are interested in. If you can't fit it all in you'd like, you'll have to do multiple captures I guess. 

 

If you do a full diagnostic scan after misfires does VDCS show any fault codes stored and descriptions?

That's the plan

I did one last night after I got home and not a single error.

I don't have a clue how to get the graphs out of my data - I'm supposed to be an IT techy, but it's a LONG time since I've had to make charts in excel and the RossTech instructions are for Office 2003!

 

The data that I've got isn't really helpful.  I had 12 items checked to log so the sampling rate is too low to see any accurate, usable data :(  From the obvious though, it looks like actual boost is close to specified.  Here's a glimpse at a 100% accelerator in 3rd up to about 80mph, then off again:

 

 

I can graph it for you but back to the start of this trail for a moment, it doesn't look like holding boost is going to be a problem.  The graph will also look a little peculiar, at the start because of the gap between specified and actual boost.  300 vs 1000 mbar.  In other plots I've seen the, ECU didn't ask for less than 1000.  

 

Anyway, keeping on point. I don't think your DV is leaking assimilating all of what is here...

 

Link the whole data and I'll look through and plot something for you....

 

PS. have you run a complete scan after the misfires?  It should retain fault codes until cleared.  I'm a little surprised no fault code or misfire detail is stored?

Edited 8 November, 20169 yr by TheClient

I can graph it for you but back to the start of this trail for a moment, it doesn't look like holding boost is going to be a problem.  The graph will also look a little peculiar, at the start because of the gap between specified and actual boost.  300 vs 1000 mbar.  In other plots I've seen the, ECU didn't ask for less than 1000.  

 

Anyway, keeping on point. I don't think your DV is leaking assimilating all of what is here...

 

Link the whole data and I'll look through and plot something for you....

 

PS. have you run a complete scan after the misfires?  It should retain fault codes until cleared.  I'm a little surprised no fault code or misfire detail is stored?

 

Yeah I thought that to be a little odd too.  

 

Thanks, here it is:

 

https://drive.google.com/open?id=0B9KSwu5pc5lzR3hLWUxWZUdhTjQ

 

I did, and no codes showed up from an autoscan.  

Can anyone tell me why my car is only asking for 300mbar at times? I don't quite understand why it isn't asking for at least 1000 at all times...

 

 

That is a chart of your full throttle run. However, clearly you only commence proceedings from 3,400 rpm so you don't see the entire spectrum, and as you say the sampling frequency is quite spread out, but it seems to be holding boost ok.

 

I'm not sure why it's calling for 300mbar boost at times. Maybe something to do with overrun on release of throttle and vacuum...Hope someone can clarify.  

 

If you logged a full throttle run from 1,000 RPM I'd expect to see boost rise sharply and slightly overboost compared to specified boost for a short time at start and then settle to a sustained boost more or less in line with specified. 

 

Your, misfire counts are very low, only 1 on each of 3 different cylinder banks. Are you feeling or noticing the misfires? Clearly they are not severe enough or sufficient frequency and duration for them to log a fault and CEL.

 

Interested to see what other suggestions are offered by forum members to narrow down fault finding.

Edited 9 November, 20169 yr by TheClient

Boost.png

 

 

That is a chart of your full throttle run. However, clearly you only commence proceedings from 3,400 rpm so you don't see the entire spectrum, and as you say the sampling frequency is quite spread out, but it seems to be holding boost ok.

 

I'm not sure why it's calling for 300mbar boost at times. Maybe something to do with overrun on release of throttle and vacuum...Hope someone can clarify.  

 

If you logged a full throttle run from 1,000 RPM I'd expect to see boost rise sharply and slightly overboost compared to specified boost for a short time at start and then settle to a sustained boost more or less in line with specified. 

 

Your, misfire counts are very low, only 1 on each of 3 different cylinder banks. Are you feeling or noticing the misfires? Clearly they are not severe enough or sufficient frequency and duration for them to log a fault and CEL.

 

Interested to see what other suggestions are offered by forum members to narrow down fault finding.

I've not directly spotted a correlation between the misfire timestamp and when the car stutters/hesitates at ~2500 but it does almost feel like a slow misfire if such a thing exists.

I'll go out again tonight. The problem is that the roads are very wet, so it's difficult to get data from 100% accelerator without wheel spinning etc. I haven't got my laptop with me, but I'll put her in third and slow to 1000 rpm, then floor it and see if there are any misfire DTCs logged.

Urgh, this problem is seemingly getting worse.  Seems to stutter badly at ~4k RPM now.  Since my last post in this thread, I've replaced plugs and coil packs.  I'll get some more data.

I've got no codes, nor can I see any visual damage/faults on the existing rev G DV, but I'm still going to try fitting the replacement I bought a while back. I've got the car going in for troubleshooting with my local VAG indy who know fathoms more than I do.  I'll take my logging data (I forgot that I'd hit the 'Turbo!' button so my 10 min drive equals about 2500 lines in the log file!  It's clear from the RPM and misfire counter (enabled/disabled) if I have the pedal engaged or not.  Specified and actual boost seem to match from what my uneducated eyes can see, but I can't work out what this annoying stutter is :/

Some great info here chaps and although All of this is way above my skill set I'll add my findings. I bought my vrs late last year and on the drive home was very underwhelmed having just sold a Leon FR with the same engine. A few searches here came up with a faulty DV. Slightly different to the op here though but mine was still on the original unit ( B or C iirc ). I was fully expecting the diaphragm to look like a tetley tea bag but no rips or damage anywhere. I debated leaving it and looking elsewhere for the problem but as the car was jacked up and the low cost of the valve I just replaced it anyway. The car was transformed night and day. I was lucky really as I didn't run any logs for boost etc just gambled. So even though these units can look ok it could be still causing problems.