Engine Mis Match And Warning Light

[mhyphenl 10 1 Cars]

I've had my 1.9 Gti for 7 or so years now and despite all the work done I still have an annoying yet consistant issue with the engine Diag light, I do have a new theory though that might need considering by those who know their stroke lengths and compression ratios etc.

Brief synopsis of the issue. Start engine and yellow diag light comes on for a few seconds then goes off as it should, after about a minute the light comes on on tickover. If I rev the engine it goes off for about 10 seconds then back on, if I tootle around at 40 in 5th then it stays off, if I do 70 - 80 on the motorway it stays on. Every sensor has been changed and tested, continuity on the loom tested.!!

Now my theory is this, when I got the car the block was damaged on the rear mount and so I got a replacement block, also the head has been replaced, I'm sure neither of these are actually DKZ parts and while I know things are largely interchangeable does anyone think that the mismatch between a DKZ ECU and sensors but a D6B engine would cause the above??

 

By the way, I've put a wanted ad in the forum to see if anyone has a DKZ engine, happy to give them the D6B plus some cash as an exchange or buy outright!

 

[welshpug 1,627]

I would start by reading the diagnostic codes.

 

D6C is just a little higher compression and possibly more cam duration but will be marginal, I doubt this will bring up any issues as long as all the DKZ management is fitted in its entirety.

[mhyphenl 10 1 Cars]

Hi Welshpug,

 

Yes all the management is there and it's all new pretty much! It only ever brings up one code and that's 52 from memory which is air fuel mixture. Basically it doesn't like want it's getting from the AFM (new), Coolant temp sender (new), Lambda (new), the injectors have been professionally cleaned and tested. There are no air leaks that I can find, I've been through a couple of ECU's even. I just can't get the idea that the only thing I haven't or can't change is the output from the engine itself given what's being put in. As I see it it's taking the AFM / Temp/ Lambda info, adjusting the injection interval and isn't happy with the result except for under acceleration.

What is strange is that I can't get the light to go off at idle, i would have thought I could mess with the afm to compensate any issues but I've tried and the light is always on unless i pulse the fast pedal.

[dcc 855]

do you have a cat? is the correct set of injectors and AFM being used for the ecu? and lastly is the correct fpr being used?

[mhyphenl 10 1 Cars]

Yes the cat is installed, the AFM is definately right, the injectors are 1.9 yellow ones but beyond that I wasn't aware there were different 1.9 ones. The fuel pressure reg works and is delivering what autodata says it should!? The strange thing is that whatever I change including the ecu the pattern with the light is the same!

[Tom Fenton 1,526]

What does it drive like?

 

If its otherwise nice to drive and runs well then I would remove the bulb from the K light...

[mhyphenl 10 1 Cars]

It drives really well actually, it has a tendancy to stall if the revs drop suddenly, like coming up to a junction. The revs just dip too much and it stops. While I was trying to get the light to go off at idle by changing the AFM spring tension I found that a couple of teeth looser (richer) stopped this completely, so at the moment I've left it like this because now it drives really well if not a little less economically

The problem with me is I know it should bloody work and even if I take the bulb out I know it's still on if you get my meaning!

[Tom Fenton 1,526]

Easy solution to rev drop stall is to open up the throttle disc stop a little, then re adjust the idle bypass screw to reset idle speed, then finally reset throttle position switch.

[mhyphenl 10 1 Cars]

I'll give that a bash tonight, i marked the AFM so I know where it should be. Still doesn't get me closer to a solution though. Is it likely this mismatch between block / /head and ECU system could be causing the issue or am i barking up the wrong tree?

[Ryan 99]

There's no way for the ECU to know about any of that.

[mhyphenl 10 1 Cars]

I know the ecu doesn't know what block, head, crank etc I've put in, but the resulting exhaust gasses from different combinations may give readings outside what the ecu is expecting! That's what I think is happening.

For example, older engines wouldn't have needed to keep as tighter control over the CO etc as the near ones!

[mhyphenl 10 1 Cars]

I meant old ones

[Miles 331 1 Cars]

Most faults with these are the valve which is in line with the carbon Canister, All the CAT car's I;ve had have this fault

[petert 554]

If dropping revs at idle, it's most likely too lean. Remove the tamper proof bung from the side of the AFM and adjust the idle mixture.

[mhyphenl 10 1 Cars]

Miles, that is interesting, I tested the valves resistance which was ok and why it wasn't causung a fault code itself but it seemed I could blow through it both ways. There doesn't seem to be any info on what that valve does and when! May be its sticking open and letting unwanted air through!?

[welshpug 1,627]

should be closed till the ecu triggers it to purge the charcoal canister, sounds like you might have found a plausible cause, it would give effectively an unmetered air leak and a lean situation.

[mhyphenl 10 1 Cars]

Are they still available? And what is its proper name? I'm only on my phone so have my got access to service box!

[Miles 331 1 Cars]

You can from memory stick a resistor in line, Valves I have no idea now, when I looked last it was OK but it was a while back

[petert 554]

It's enough to check. Disconnect the canister from the inlet of the valve, so it vents to atmosphere. Leave the solenoid connector to the ECU, but plug the inlet, preventing any air entering that way. That will verify if it's leaking and the ECU shouldn't fault. Clear all the faults first. The 1.9L Mi16 has the same valve, or you can replace it with a 1200Ω resistor, which will keep the ECU happy.

[mhyphenl 10 1 Cars]

This is the first time in years I've got something new to try to resolve this. I'm actually excited but cant get on it till the weekend. I'll keep you all posted though!

[Bobbafett 49 2 Cars]

My 1.9 with cat had exactly same issue when I bought it. It was the lambda sensor.

 

EDIT: Yeah I read the bit where you changed that now....

Edited December 18, 2014 by Bobbafett

[mhyphenl 10 1 Cars]

To be honest , if the purge valve thing doesn't work then changing the lambda again may be an option. The car has done a fair few miles without working correctly so the lambda is probably s*itted up by now! Can they be cleaned in any way?

[Tom Fenton 1,526]

I would suggest the lambda is actually working properly as that's what is flagging the code, eg it's lean most of the time if the purge valve isn't shutting as it should. I'd block the line to the purge valve first of all and see what happens. Leave the valve plugged in.

[mhyphenl 10 1 Cars]

As an air leak makes it lean, I assume it's the lambda that says the mixture is too lean and tries to add more fuel? I presume the error is triggered because it can't bring it into line considering other factors like AFM and temp!? I have noticed a small rise in revs when the light comes on, does this hint at the attempted adjustment of the fuel/air mix?

[Tom Fenton 1,526]

At a guess when the light comes on it then STOPS trying to correct the mixture as it's identified the fault hence put the light on.

Only a guess there maybe someone who knows the Motronic system better who can advise.