Diagnosing possible DV problem...

[grayham]

Hi all, I'm in a slight pickle! Sorry if this is a bit waffly, but...

A couple of weeks ago my CEL came on - first time ever, didn't even know what it was. So I took it to the nearest VAG dealer (VW dealer in Ruislip) sharpish, in case it was something serious. They charged me £65 to plug their laptop in and read out the fault codes! Anyway, turned out to be 17705 and 17545. They quoted me ~£300 to fix it, by cleaning the throttle body and replacing the MAF.

17545 P1137 Long Term Fuel Trim Add.Fuel,Bank1 System too Rich
17705 P1297 Connection turbocharger - throttle valve pressure hose

I then started searching around on here and elsewhere, bought VAGCOM and a lead, and started logging things. Looks to me like the MAF is fine. My car is a stock 2001 Cupra (AUQ), and I'm getting up to 158g/s on the airflow reading, which sounds like it's OK to me, unless there's a very intermittant fault that I'm just no capturing. Plots of it look fine - no unexplained jumping around.

So currently my uneducated money is on either a dodgy DV, an air leak somewhere, or a dodgy pre-cat O2 sensor. Logging the O2 sensor shows it jumping about a fair bit, and spending a lot of time below the -10% normal threshold, but I don't know if that's because the sensor is knackered or because something else is causing the O2 levels to go outside the normal range.

So I've taken the DV out, and taken a good look at it. Thing is, I can't get the top off to check the diaphragm. The bugger is on very tightly indeed - if I try any harder then I'm sure to just crack it. I've even filed a couple of little flat spots on it to get purchse in a vice. I read somewhere that you should be able to push the diaphragm up, seal the top hole with your hand and see if it returns or stays up, but having poked around in the bottom hole, I can't get anything to move at all. I've not seen inside, so am not 100% of how it's supposed to work.

Can anybody give me any further clues as to how best to proceed? I cleared the fault codes when I first got VAG-COM, to see how quickly they came back, and the fuel trim parameters haven't yet drifted back out of range. Not sure how quickly the adaptation should take place.

Cheers,
Gray

 

[h100vw]

Gray, if you had an alloy DV then the top would come off easily. By un-screwing it. I take it therefore that you have the std plastic one fitted, you can check the diaphragm by sucking on the vac connection on the top of the DV.

My advice would be to replace it for an alloy one from the usual places. If you are running a remap you might need an uprated spring if you are using a Forge DV.

Gavin

 

[grayham]

Hi Gavin,

Gray, if you had an alloy DV then the top would come off easily. By un-screwing it. I take it therefore that you have the std plastic one fitted, you can check the diaphragm by sucking on the vac connection on the top of the DV.

Yup, I've got the standard 710N fitted. If I suck on the small connector on the top of the DV, am I right in thinking that if it's working correctly then this should just lift the diaphragm inside, but not allow air to come sucking into my mouth through the other orifices?

My advice would be to replace it for an alloy one from the usual places. If you are running a remap you might need an uprated spring if you are using a Forge DV.

I'm not running a remap. I'm slightly loathe to spend £100 on a new DV (even if it is a better quality component) if I don't know that the current one is faulty. Might have to though, since I'm just as unkeen to spend £20 on a replacement standard one, since there's a fair chance that I'd have absolutely nothing to show for it.

I've now been logging a few measuring blocks, and my fuel trim values are showing a consistent ~-0.2% additive and ~-7% multiplicative. The multiplicative one is a bit low, but that's not what triggered the original fault codes, and it's within the +/- 10% 'normal' range.

I've also looked at my lambda values - the first is mostly around 1, but occasionally it blips right up to 2...! The second lambda value is right where you'd expect it - around 1, but blipping down to about 0.75-0.8 during prolonged accelerations. My oxygen sensor reading from block 001 seems to average about -10% (right on the borderline of acceptable) but seems to oscillate rather wildly between about -25% and +5%... so I'm guessing that it's likely that either my pre-cat O2 sensor is up the creek, or I've got a leak somewhere that's randomly letting air in / out. Sound like I'm on the right track?

Cheers for your help.

Gray

 

[MarkE]

Wouldn't like to be putting my mouth anywhere near a DV, to be honest.

You say you've tried to check the DV but without success. If you look inside the bottom hole (ooh-err missus) then you can see a small metal plate. If you push on this (and you have to push reasonably hard) then you can move it upwards (towards the small vac pipe connection). Whilst holding it in this position, put your finger over the vac pipe connection and release the metal plate - it shouldn't pop back down again until you release your finger from the vac pipe.

 

[h100vw]

You should experience resistance to your sucking. If you are shy about sucking engine parts, MarkEs method achieves the same result.

http://cgi.ebay.co.uk/Piston-Closed...ryZ72205QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

I bought one of these for a Cupra that I was selling. I didn't want to leave a hundred quid one fitted in the car as it was only a few days old. I needed one in a rush.

If you speak to the guy he is expecting some more in that will be coming with alternative springs. The one that comes with them would be fine for std boost.

Having had an A/F meter fitted on a G60 Golf. I know that the lambda probe output fluctuates at tickover between rich and lean. It does the same thing during the cruise. On the G60 this was normal and I would think the same of a 20vt. The ECU monitors the mixture and as it senses a weakening mixture will richen it up. As it gets richer the ECU will weaken the mixture. It will keep doing this many times a second. If you were to rev the engine then let the revs fall the mixture would go fully rich as you blip the throttle, then down to zero on the over-run as the ECU cuts the fuel until engine drops to 1500 revs or so. Once idle is achieved, the cycling will start again.

You would need a moving coil meter to see this properly, the sample rate of VAG-COM is too slow to log it perfectly.

Having spent a bit of time at JBS, James told me that he expects to see readings in the .7s during acceleration. This indicates a rich mixture. Readings of .9-1 would only be seen during the cruise or idle. Readings over 1 would happening while coasting down.

I think you should log your boost pressure, requested and actual. If actual doesn't match requested chances are you have a leak, it may be very slight.

Gavin

 

[ForgeMotorsport]

Please be aware that the " similarity " of these valve are causing some BIG problems , they are not Forge and do not come with our guarantee , I have had some sent back to me for service , and have had to inform disapointed customers they are not made by Forge ....there are many Forge " Forgeries" on E bay

You should experience resistance to your sucking. If you are shy about sucking engine parts, MarkEs method achieves the same result.

http://cgi.ebay.co.uk/Piston-Closed...ryZ72205QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

I bought one of these for a Cupra that I was selling. I didn't want to leave a hundred quid one fitted in the car as it was only a few days old. I needed one in a rush.

If you speak to the guy he is expecting some more in that will be coming with alternative springs. The one that comes with them would be fine for std boost.

Having had an A/F meter fitted on a G60 Golf. I know that the lambda probe output fluctuates at tickover between rich and lean. It does the same thing during the cruise. On the G60 this was normal and I would think the same of a 20vt. The ECU monitors the mixture and as it senses a weakening mixture will richen it up. As it gets richer the ECU will weaken the mixture. It will keep doing this many times a second. If you were to rev the engine then let the revs fall the mixture would go fully rich as you blip the throttle, then down to zero on the over-run as the ECU cuts the fuel until engine drops to 1500 revs or so. Once idle is achieved, the cycling will start again.

You would need a moving coil meter to see this properly, the sample rate of VAG-COM is too slow to log it perfectly.

Having spent a bit of time at JBS, James told me that he expects to see readings in the .7s during acceleration. This indicates a rich mixture. Readings of .9-1 would only be seen during the cruise or idle. Readings over 1 would happening while coasting down.

I think you should log your boost pressure, requested and actual. If actual doesn't match requested chances are you have a leak, it may be very slight.

Gavin

 

[grayham]

Thanks guys. I'll try testing my DV again - I pushed the metal plate fairly hard and detected no movement so, not wanting to break anything, didn't push any harder. If I find that it's not working properly then I think I will just get a Forge one rather than something off eBay with questionable support etc.

I've logged requested versus actual boost pressure, but wasn't really able to get any nice sustained accelerations - (roads are a bit busy round here!). In general though, the actual boost pressures were pretty much the same as the requested ones. In stop-start conditions, the requested boost readings quite often stayed at baseline whilst the actual blipped up and down.

I think I'll go out one evening this week and try to get some cleaner data.

Anyone got any good ideas as to how to test my pre-cat oxygen sensor without just replacing it and see if anything changes?

Cheers.

 

[flashbsd]

where abouts are you from? you could have my standard DV whats in the loft to try if its any use?

 

[grayham]

where abouts are you from? you could have my standard DV whats in the loft to try if its any use?

Hi, that's really kind of you - thanks. I'm in Richmond, Surrey. I'll have another play with my DV, and if I suspect it's dodgy then perhaps I could take you up on your offer to see whether it sorts things out?

Cheers,
Gray

 

[PhilW]

Did i just read right? Your not running a remap but seein 158g/s?

 

[flashbsd]

leon cupra not an ibiza possibly? MAF readings are different

Just let me know grayham and i could drop it in the post for you Annoying when you get these niggly faults, i ponder for the day mine runs 100%

 

[grayham]

Did i just read right? Your not running a remap but seein 158g/s?

Yes, that's right. 2001 Leon Cupra, no modifications whatsoever.

Gray

 

[h100vw]

Please be aware that the " similarity " of these valve are causing some BIG problems , they are not Forge and do not come with our guarantee , I have had some sent back to me for service , and have had to inform disapointed customers they are not made by Forge ....there are many Forge " Forgeries" on E bay

Thanks for your input. Forge DVs should be instantly identifiable as they carry the company logo.

I don't believe the guy on ebay is trying to pass these off as Forge DVs. How different could you make them look?

At the end of the day I bought mine as an experiment and it worked fine as far as I was concerned.

Being an aircraft engineer with an inquisitive mind I couldn't not try one. As to how they'll last, compared to a Forge I'll never know.

@ Forge. In the interest of science, would you be prepared to elaborate on the problems that you have seen with the imitations. Over and above people trying to get you to service ones you didn't supply.

Cheers
Gavin

 

[grayham]

leon cupra not an ibiza possibly? MAF readings are different

Yes, it's a Leon, not an Ibiza. The Ross-Tech guide suggests a ballpark figure of around 0.8*stated max bhp. Mine should be 180bhp, so the air flow I'm seeing is a touch on the high side, but not crazy.

I didn't get the car new, so for all I know the previous owner *could* have had it remapped, though I doubt it.

Just let me know grayham and i could drop it in the post for you Annoying when you get these niggly faults, i ponder for the day mine runs 100%

It is annoying, yes! To be honest, since I cleared the faults, there are only scant clues that there's anything wrong at all. I'd far rather find something broken and fix it than just forget about it and wait for something to stop working completely though!

Cheers,
Gray

 

[flashbsd]

@ Forge. In the interest of science, would you be prepared to elaborate on the problems that you have seen with the imitations. Over and above people trying to get you to service ones you didn't supply.

Look very similar to me, very interesting. did the internals look any different?

 

[h100vw]

flashbsd, was going to pm you but can't. Not got permission, is that's because I am a newboy? I didn't want to stir up the hornets nest any more.
Can you pm me an email?
Gavin

 

[h100vw]

Nevermind the hornets then.... The bit I could see that was different was the vac connection on top. The rest was........

 

[flashbsd]

cant pm you either mate not got enough posts?

 

[h100vw]

need to go cap in hand to someone

 

[grayham]

leon cupra not an ibiza possibly? MAF readings are different

Just let me know grayham and i could drop it in the post for you Annoying when you get these niggly faults, i ponder for the day mine runs 100%

Hi again, I took the DV off again over the weekend and had a play - it appears to be working fine. Pushing a bit harder from beneath shifted the piston, and it stayed up with my hand over the top intake - no sign of leakage at all.

So I don't need to try another, thanks very much.

Does mean that I'm still at a bit of a loss for what caused my original fault codes though. Car seems to be running fine though, and hasn't thrown any codes lately. Also the fuel trim parameters are pretty consistent and within-range. So I'm inclined to just leave it for now, and see what happens!

Cheers,
Gray