[nathusius]

Hi from Kent, UK!

I've had a Golf Mk4 TDI for about 6 months and getting limp home mode due to positive deviation almost every time I drive it now - seems to occur at high throttle openings in top gear, usually going up a hill. VNT rod appears to move freely and the capsule isn't leaking; have cleaned out the EGR valve (although it didn't look too bad) - that didn't help the limp home, but did seem to clear up a slight stumble at light throttle around 2000rpm. I've just changed both valves (N75 and EGR) but again no difference, so it is looking more like the VNT, but hoping to find something else before investigating that, as from the outside it appears to be working OK!

I'm using a registered version of VDCS-Lite, so have been looking at logged data - actual intake pressure is going up to ~3000mBar when spec value is ~2350! (and actual MAF is 1250 when spec value is 850). Limp home mode is triggered after 5 or 6 seconds of overboost, but can be avoided by not using full throttle; sometimes the fault code is described as intermittant, so I guess that is triggered after several shorter periods of overboost? (rather than one long one!). With a vacuum tester on the capsule (with turbo still on car) the lever appears to move as freely as in your demo video and the capsule seems to hold vacuum OK.

I have a couple of questions to try and understand how the VNT works (sorry if already answered elsewhere!):

1) which way round does the lever work? (i.e. does vacuum to the capsule give low or high boost?) When the engine is started the lever moves, so it seems that full vacuum gives max boost and venting it to atmosphere gives low boost, but just wanted to confirm this

2) what returns the VNT vanes when vacuum is removed? (i.e. is there a spring in the capsule)

I'm an electronics engineer and used to design motorsport ECUs, so I'm quite comfortable using VCDS and looking at logged data; not so keen to take my turbo apart though! :ugh

Any help gratefully received and of course I'm more than happy to post whatever help I find

 

[Seatman]

There's a spring in the capsule that puts it back to its starting point, I have heard of the actuator spring becoming weak and causing overboost, I'd expect stickty vanes to cause more of an underboost because they're sticking, as the car needs more power it increases the vacuum to give more boost but that's about all I know about the vnt, I'm still learning. There'll be a way to measure how much vacuum is required to move the actuator though which you might be able to compare with specifications if you can find the info which should be able to tell you if it's weak or not, if I find anything I'll post back but in the meantime welcometomyturbodies

 

[chittychittybangbang]

Hi there, read this FAQ Article: http://www.myturbodiesel.com/1000q/a4/VNT-wastegate-adjust.htm it explains many of your questions. Disconnect the arm to isolate any sticking to the actuator or inside the turbo. Once you do that you'll have a better idea of where to look. It's possible that it's sticking and then pops free, giving a sudden spike of too much boost which is triggering the limp mode.

 

[nathusius]

Hi again and thanks for the replies - seems like I may have fixed this, although maybe not a permanent fix?

I measured the vacuum needed to move the VNT actuator and it seems too low - the rod moves as soon as I apply any vacuum (0~1 inHg) and seems to reach full travel at ~15 inHg, so could I have a weak return spring in the actuator? The tandem pump was producing 25 inHg vacuum at idle, so more than enough to move the actuator through the full range; there were no leaks in the vac system and I have already changed the N75 valve. With it all still on the car the VNT rod travel appeared to be smooth with no obvious stiction seen or felt by hand, but maybe things are different when its all good and hot? Gave it good soak with 3-in-1 professional penetrant and left it running with VCDS basic setting operating the VNT in test mode, then repeated with 3-in-1 professional lubricant with PTFE. Took it out for a good hard run with VCDS logging - the problem was still there, but now only really bad if I put my foot down hard in 6th gear - boost would go up to 2950 and stay there even with VNT duty cycle at 94%; doing the same in lower gears would not do this and the boost would be controlled to the spec value (2350), so it seemed to be the vanes or actuator rod sticking at high temps or max load.

I read on a UK TDI forum about Forte Advanced Formual Diesel Fuel Treatment being used to help with stuck VNT problems:

http://www.forteuk.co.uk/product.php?id=51/Advanced_Formula_Diesel_Treatment

The description says that it helps to "maintain EGR valve and variable vane turbo cleanliness", so decided to give it a go and got 2 bottles on eBay UK. One bottle should treat 100 litres, but I put a whole bottle into half a tank and went out for some more hard runs as recommended here, using high revs and high load, especially up some long hills. It was noticeably better right away - boost was being controlled in 6th gear, although response time was still quite slow, with a large boost spike followed by gradual decay down to spec value. I've used most of that tank over the last few days, so decided to put the second bottle into an almost empty tank this evening - it now controls boost in all gears with hardly any overshoot!

It seems to good to be true and I'm not getting my hopes too high yet, but it has definitely made a big difference; maybe my actuator spring has got weak and isn't (wasn't?) always able to moves the vanes back quickly enough? So I'm thinking that the turbo probably does need removing and the vanes cleaning and possibly the actuator replaced, but at the moment there doesn't seem to be a huge need to do this.

I could probably post some before and after logged VDCS plots, if anyone is interested?

 

[chittychittybangbang]

Sure, I'd like to know after a few weeks or few months if it's still working. If so, I'm sure many more people would like to give the product a try.

 

[nathusius]

Sure, no problem, will let you know (I'll be pretty interested myself!), although I am thinking that there is probably still an underlying problem somewhere.

Any thoughts on the vacuum figures I got to move the VNT actuator? Mine seemed to move full range between 0 and 15 inHg, which is about 5 lower than the figures in the VNT adjustment article, so maybe a "soft" return spring together with slightly sticking vanes gave me the problem, but the diesel treatment has helped to clean the vanes and make it all much less noticeable?

Next step would be to disconnect the VNT rod I guess, but wasn't feeling quite up to that task - happy to give it a go, but it looked very tricky to remove and replace the E-clip with it all on the car (didn't want to get stuck half way through the job!)

Anyway, more than happy with it for now, even if it's only a temporary fix!