Log in

View Full Version : Question Cabriolet 2.0e ABK ecu identification



Pexter
07-04-2025, 07:54 PM
Hello!

I have a cabrio with legendary ABK engine and it has got the problem running very very rich.
Have been replacing parts multiple times and so did the last owner and the one before that.
Got the original repair manual and started measuring all the components and one that seems to be off is the voltage from the ECU to the signal wire of the oxygen sensor.
Manual says:

"Checking wiring
- Connect up multimeter for voltage measurement between connector of signal wire to engine control unit and engine earth.
- Switch on ignition.
- Specification: 450 ± 50 mV
If reading does not match specification:
- With ignition switched off, connect test box V.A.G 1598/18 to engine control unit connector => Page 45 .
- Check for open circuit in the following wiring.
Signal wire to ECU pin17
- Eliminate open circuit in wiring if necessary. "

On my car the voltage is ~850mV (ignition on, engine off)
This is where i started to think that do i have incorrect ecu installed. It is 039906024D 5WP4 185 <- vag number is correct but what the 185 means in that siemens number? almost every other with the same vag number has 5WP4 174?
I also have 039906024A and 039906024F, but those put out straight B+ to the oxygen signal wire. These also had one burnt resistor/diode on the pcb, but what i read from some russian forum, that resistor isn't needed :D

If i measure the "mixture" straight from the oxygen sensor signal to ground, i get constant ~900mV indicating rich mixture and at the same time vcds shows 0.4-0.5V meaning that the ecu thinks everything's fine.
If i give the engine more air somewhere, those both dip down showing the engine going lean and the ecu "corrects" that by increasing injectors open time.

My car VIN is WAUZZZ8GZWN00****
And sorry for my messed up writing, I'm from Finland and it's been a while since my last time writing in English :)

EDIT: And if someone can confirm some vag partnumber + siemens partnumber that is 100% correct, that would help me buying one more ecu

Crasher
08-04-2025, 09:39 AM
We need the full VIN, it isn't GDPR sensitive data. Do you have a diagnostic system which shows measuring blocks?

Pexter
08-04-2025, 01:17 PM
Sorry, my bad.
Here it is WAUZZZ8GZWN002474

Yes i have almost genuine vcds. :)
Friend has genuine vcds cable and another friend has autocom or something like that.

Crasher
08-04-2025, 05:33 PM
You need ECU 039 906 022

Pexter
08-04-2025, 07:50 PM
OK, thanks!
Just ordered two of those from salvage yard, should be here in week or two.

Is my car some anomaly or how those part numbers go?
Looked from online etka and got very confused with all.
F >> 8G-P-015 000 'DIGIFANT' ABK 039 906 022
F >> 8G-P-015 000 'DIGIFANT'039 906 022 ABK 039 997 022 X
F 8G-R-000 001>> 8G-R-015 000* 'DIGIFANT'039 906 024A ABK 039 997 024 X
F 8G-R-000 001>> 8G-R-015 000* 'DIGIFANT' ABK 039 906 024 A
F 8G-S-000 001>> 8G-V-006 000 'DIGIFANT' ABK 039 906 024 D
F 8G-S-000 001>> 'DIGIFANT'039 906 024F ABK 039 997 024 FX
F 8G-S-000 001>> 'DIGIFANT' ABK 039 906 024 F

Mine has W in the vin code and it doesn't seem to go anywhere on those and i get these like that 039906022 was in the early models and 039906024F on later.

Crasher
08-04-2025, 09:20 PM
Remind me tomorrow as I have this info at work

Pexter
09-04-2025, 10:10 AM
I remind you 😁

Crasher
09-04-2025, 05:26 PM
Too late when I saw this, just shut ETKA down, prod me tomorrow.

Pexter
10-04-2025, 07:11 AM
New try

Crasher
10-04-2025, 07:06 PM
Nope, missed again; been working on staring up an S2 restoration.

Pexter
11-04-2025, 08:45 AM
One more question, what grounding points should ecu have? Should the ecu end plate "the connector end" be grounded and what pins should be directly to ground?
Did read something about ground for the gearbox to make these boxes go boom. Don't wanna fry the ecu's that i just ordered :)

Pexter
11-04-2025, 02:08 PM
This is interesting. I got one of the 039906022 ecu's i ordered and it has different connector. Only 38 pins and other differences too...
I'm pretty sure thats not going to work 🤔

Edit:
I saw writing on some other forum that there are little resistor packs that can alter the signal from the o2 sensor so that ecu gives more fuel for the engine. I'm not sure that i have this thing, but i measured the resistance of the o2 signal wire and it was 2.5ohm. sounds pretty resistive copper to me? Some other wires were 0.2ohm so it's not my meter resisting :) this was done ecu disconnected and from the wire to ground there was nothing, so maybe it's not short to ground at least.

Maybe i should have told this earlier, but i bought the car with this running rich problem and it has been like this atleast for 3 previous owners. Changed parts list is long and many parts have been replaced more than once. I don't know how many of those parts were new, cause they are pretty rare today. O2 sensor is for sure new and original bosch 🦾

List what has been checked/replaced:
Distributor + wires + plugs
Ignition timing
Engine timing
Ecu temp sensor
Fuel pressure regulator
Airflow meter 2x
Idle stabilizer 2x
pcv valve 2x
Vacuum leaks removed
ECU 4x
O2 sensor 3x
Throttle position sensor
Maybe something else, don't remember all.

There is slight possibility that some previous owner has done some magic to the wiring harness. Car has some newer vw/audi instrument cluster and steering wheel. Wiring looks very bad for those and there is many more bad connections made for speakers, interior lighta etc etc. Most of them don't affect the ecu, but isn't speedo somehow connected to ecu 🤔

Crasher
12-04-2025, 01:41 PM
Interesting, I need to be at work to dig deeper. Devices that linearly modify the Lambda sensor output voltage can have all sorts of nasty side affects so ones that apply mapped selective corrections that are plotted using a 4 gas analyser or an additional wideband sensor (the Bosch LSU 4.9 is popular) and there are various devices out there that incorporate a wideband display and the ability to map the narrowband output to give you the values required. There are few full “piggyback” systems now such as the Perfect Power SMT8 (they simply vanished) but there are plenty of wideband lambda controllers. I am rebuilding a 95 S2 Avant and I have incorporated an additional m18 sensor boss in the new downpipe I had made and will use a 4.9 here to tweak the signal from the OE 4.2 to cover the mild modifications.

Pexter
14-04-2025, 08:48 AM
What's the difference between 039906024A, 039906024D, 039906024F? Are they completely different controllers that one can't be replaced with another?

Pexter
14-04-2025, 05:56 PM
Damn, my car wiring is pretty messed up. Checked pin42 from ecu connector and that should be ground point and I had 12v on there :D
I dont't know what else is wrong, should i just fix that and hope everything else is fine :confused:

FrenchAudi
14-04-2025, 06:15 PM
I suggest you use an analogue voltmeter (one with a needle) with a low impedance rather than a digital one, which typically has a very high impedance, to ensure you are not reading misleading leakage voltages.

Pexter
15-04-2025, 03:25 PM
Case closed.
Fixing that ground point solved the whole thing... Almost. Slight idle issue remains, but i got emissions test cleared today and thats what i needed.
Thanks to all.

FrenchAudi
15-04-2025, 03:31 PM
Congratulations. Just goes to show that a bit of traditional troubleshooting and simple checking can get better results than those who just throw new parts at a problem.