Club 80-90 Forums

Hi fellow owners!
I was hoping to find advice on an engine issue.

I own an Australian VW T3 Syncro 1990 2.1l petrol fuel injection which was imported to the UK before I bought it.

It starts running extremely rich as soon as the engine warms up and then dies off after about five minutes. I tried and tested different temp sensors (or temp sensor disconnected). Also it has a lambda sensor (which seems to be aftermarket?) - I also replaced it and disconnected it but no difference. Had another ICU installed and tested, also no difference.

It was running fine but only started doing it after I had all the coolant pipes and tubes replaced.

In addition it also is running strange under idle. It is idling normally before moving up to 3.000 rpm. Had the ICV replaced and tested, no difference.

I‘m a bit lost now on where to look next - and so are the mechanics around here.

I don’t have the deep knowledge of the van, so apologies if there is something obvious that I missed.

Thank you so much in advance!

If it was fine before the water pipes being replaced , either something has been knocked a wire etc or there’s an air lock

download the 2.1 manuals from Syncrosport.com or the Sir Adventure website and when doing maintenance/diagnostics follow the procedures - these engines are simple by comparison with modern engines. I'd be checking the idle switch on the throttle body and also that the vacuum take off spigot on the plenum chamber is clear and the hose that goes to the fuel pressure regulator is good, over pressure on the fuel rail combined with no lambda or the wrong lambda will cause over fuelling, as will failure to switch to idle mode when the idle is controlled by the little box behind the right rear lights rather than the ECU which is only supposed to run the engine when not idling.
The only lambda in my experience that works with the Digifant system is the Bosch, part number 0 258 003 957; the Beru is supposed to work but we've had numerous incidences where it didn't. The cheap aftermarket ones off Ebay do not work - I don't exactly know why, but suspect it is something to do with the signal hysteresis or waveform. The original was made by Bosch as was the whole ignition system, it's a VW specific version of Motronic ( I think ver. 2.x or 3.x), same as Digijet is

and post up your location in case someone who knows what they are doing can help locally

Thank you for the quick replies. I wasn’t aware of the Lambda sensor specifications - seems like the first thing to change. All I tested were likely aftermarket though did work for some time before.

I’m looking deeper into it over the next few days.

I’m based in South Manchester and have two garages specialised on old VW (not only T25/T3 though).

Thanks so much again, will look into the things and report back.

way back in 2004/5 three of us fitted Lambda probes from GSF to our new group buy JP stainless exhaust systems, all of them worked at first, all failed inside 700 miles, when tested before replacement the probe was cycling readings between high and low just as it should but it was like the ECU was sticking its fingers in its ears and going no I can't hear you, we all fitted Bosch and were fine thereafter.
Subaru ECUs on the other hand are quite happy with £20 Lambdas in my experience

Aidan wrote: ↑24 Jan 2024, 20:52 way back in 2004/5 three of us fitted Lambda probes from GSF to our new group buy JP stainless exhaust systems, all of them worked at first, all failed inside 700 miles, when tested before replacement the probe was cycling readings between high and low just as it should but it was like the ECU was sticking its fingers in its ears and going no I can't hear you, we all fitted Bosch and were fine thereafter.
Subaru ECUs on the other hand are quite happy with £20 Lambdas in my experience

The MV ECUs and most stuff of that era really do use very primitive lambda sensor signal conditioning stuff - they were designed in an era when the bosch sensors were the only ones and the technology was pretty new, so they're built to expect the exact "foibles" of the original sensors. As time went on, other manufacturers got on board with O2 sensors etc etc, the signal conditioning stuff evolved a lot and newer ECUs can make sense of a muddier signal. The aftermarket O2 sensors are designed with a much more slack set of characteristics then the ones supplied in the late 1980s, and they work fine in most cases. 
I think there is a level of ECU degradation too - power supplies become noisy because capacitors wear out etc.