Hi All,

 

Still having over rich problems (amongst others) on my recently rebuilt TT. It runs extremely poorly and is not really usable.

 

I've noticed an anomally, or at least i suspect it is, checking injector duty cycle on the AVC-r, it shows as 90%+ at idle and drops with throttle applied. Is this as expected? Seems to be completely the reverse of what should happen and makes me suspect a crossed connection or short somewhere.

 

I double checked with consult and sure enough the injector time drops as throttle is applied. Whether this is not the case when the engine is under load I haven't been able to test (not really driveable).

 

Can anyone check on their AVC-r, consult etc and let me know if they experience the same please?

 

Cheers

 

RobH

Given the running problems you're having I'd say that the reading is possibly correct and that's why it's overfuelling on idle.

May be a good time to try a known good 555cc chipped ECU and see if that works any better. Always a possibility that there's a problem with the program on the nistune NVRAM.

I've seen these symtons once before - Injectors running at near full chat without any throttle, could'nt get the car to start, tried everything.....could'nt find out why, in the end it was diagnosed with a fault with the dodgy old EFI loom...shorting itself out somewhere...new loom..fixed.

Hi guys,

It appears that I wired the AVC-r's optional injector monitoring connector into the RPM connection (this allows self learning mode) - it was in the dim distant past so I had forgotten. The downside to this option is that the injector duty cycle reads a load of b*ll*x. I made some other changes yesterday - including a gaping boost leak (tut!) and opening the fuel tank vent hose (previously connected to the carbon can which has been binned) - the original metal tube it connected to was blocked solid!! The carbon can was split hence venting was not a problem when it was in use ;-)

 

It's still running like a pig and sounds pretty awful but the black smoke has gone and I have been able to hit half a bar! Not bad considering it only seems to be firing on 4 or less most of the time!

 

Compression test and new plugs today, new injector and coil pack connectors are in the post (they're all fkd)....might be running again by next summer...

Cheers

H

might be running again by next summer...

 

Sure I said that about mine 2 years ago!!

Good luck with it.

I've seen these symtons once before - Injectors running at near full chat without any throttle, could'nt get the car to start, tried everything.....could'nt find out why, in the end it was diagnosed with a fault with the dodgy old EFI loom...shorting itself out somewhere...new loom..fixed.

 

I had the self same problem with mine. I took a plug out to see how much fuel was going in and it short fuel right over me and the engine bay. about a mug full and two eye as we'll. Not recomended it fooking stung like nobodies business. Mine was that bad that the car wouldn't start and was indanger of hydraulicing the engine.

 

I took the rail back out to make sure I hadn't split an injector sealing ring etc. Even went and got Andy Duff to put the injectors in Third time round as I was sure I was splitting the seals. Had JD send me a couple of newly burnt chips and tryed 2 different ECU's. In went a end new loom Vroom and away to go.

Interesting.

Only thing I can think of that might cause that on idle would be the cranking signal to the ECU being shorted?

Given that you can see MAF, TPS etc readings on consult / nistune then it's a fair bet that all that wiring is OK.

Not going to be shorted injector earths if you can hear them clicking.

Just thinking that the cranking enrichment will be huge so if the ECU thinks the engine is cranking then it will overfuel massively.

Don't have tuning stuff on my work PC so can't check but might be worth looking at, try disconnecting that signal??

Think I've sorted it, latest changes are relocating the engine block ground strap, refitting my bigger injectors, new plugs (again!!), put the Nistune daughterboard back in and programmed with a 550cc image, had to fiddle with one of the injector connector terminals which was being pushed out of its housing when pushing it onto the injector. Took it for a spin last night and got half a bar (that's all I'd set the AVCr for). Sounds good and idles quite nicely.

The downside was that a fkin RX7 pulled up behind me at the lights and was able to get past me due to my low boost setting :-( BAH! Bet he thinks his car is faster than a Z now - fker!

 

John, am I right in thinking I shouldn't be dicking about with the injection multiplier and latency until I've got a wideband O2 sensor installed? It's awfully tempting ;-)

 

Cheers

 

H

Yeah, prob best to leave it. On idle/ low throttle you won't do any damage but once on boost you could quite easily cause a blowup!

I'd try K=205 to start with and whatever latency venom recommend, see how that goes. Worth disabling the closed loop initially too to stop it trying to adjust out your changes.

Just thinking that the cranking enrichment will be huge so if the ECU thinks the engine is cranking then it will overfuel massively.

 

Yeah, injectors fire on every stroke, rather than just sequentially on cranking.

And with water cold, 11ms is added to the pulsewidth when cranking which is >10x the actual idle requirement.