BrianC72gt

The power output stages don't suffer abuse well. When my ecu swam, it also cooked a bunch of coils (40 valve engine, no POS), but I bet a short due to water could saturate the POS bits and cook em. How certain are you that they are OK?

Hey can you measure fuel injector duty cycle? Ross tech's VAG scope does it. That will tell you if they are being triggered.

hillpc

It seems that the power output stages aren't even being driven by the ECM (no pulses coming from the ECM to the disconnected POS connectors). I was thinking of triggering them manually with a jumper wire from +12V to see if they'll fire the plugs, but that would only prove they'll fire if triggered, and doesn't address the ECM not even firing them. What's the 40 v you refer to?

I'll do that if it doesn't start once I have spark.

fsrowsell

hillpc, I changed out my G28 sensor today but, unfortunately still no start! She turns over fast and wants to start. She sounds like a flooded engine. Any luck on your end?

hillpc

Nothing yet on my end. I've posted on the Ross-Tech forum as well. They make the VCDS diagnostic tool/software. how to check what's keeping ECM from driving the power output stages (97 Audi 4.2)?

All - is posting to another forum (with a little different focus, maybe) an OK thing to do, etiquette-wise?

fsrowsell

I took all the plugs out, to check them and of them were clean except #8. It was black!

hillpc

maybe a leaking fuel injector on #8. But I don't think that would stop the car from starting.

fsrowsell

Just finished checking my relays inside the cabin and found 381 to be not working. (see picture) I don't know what this relay is for,but I will replace it. One on order now.

Kneale Brownson

Get all the help you can ANYWHERE.

BrianC72gt

I forget where we left this...seems like 2 cars with similar problems intermingled in this thread, so I'm not sure. Limited Ross Tech access on this one? OK, Simple test, get one noid light. Summit racing sells an assortment for $14, harbor freight probably cheaper than that.

Key in, engine in RUN position. Remove one injector plug wire. Two wires. One should show CONSTANT 12 - 13.5 volts ish on your cheap-*** multimeter. The other will show, nothing, not even continuity to ground. This dead wire gets switched to ground by the ecu very rapidly while running to open and close injectors. Too fast for any digital multimeter to detect. Stick a noid light in there, it should flash on/off while engine is cranking.

If no flashing, then the ecu is not swicthing the injector to ground. Why not? 1) Bad ECU ground wire(s); 2) bad ECU power in; 3) ECU is not getting good signal that the engine is cranking (flywheel sensor), or; 4) ECU is ****ed after swimming. Were you able to record engine speed sensor with RCDS? How fast? Perhaps someone with a good car can unplug their fuel pump and measure their Engine speed sensor while cranking for comparison. Does a bad ESS sensor die altogether, or just start failing intermittently by missing some and counting low. If your count with a well charged battery is way low by comparison, then maybe the sensor is bad after all.

If the noid light is flashing, check the other 7 for laughs. Is the fuel pump good. (Yes?) Are the plugs sparking? No? All 8? Swap in one known good coil and plug? Will it spark a grounded plug while Cranking? If the wires are so short that grounding is awkward, flip the coil over, plug pointing up, put small jumper cable on plug thread and other end to a good engine ground. Heavy speaker wire and an alligator clip ought to do it. Crank it. Spark? Hold a piece of black paper behind it so you can see. Test them all. If they are all dead, BUT the NOID light flashes, then likely ECU OK, but water in cowl shorted power to the coils which sat, saturated & hot, without discharging. Death by blue *****. I hate to throw parts at it randomly but at that point try one good POS, one good coil, and one good spark plug. My swim cooked all 8 coils either immediately or shortly after everything else got sorted.

Again, if the noid light doesn't flash, FULL STOP, nothing else matters, don't spend another penny. Confirm good crank signal to ECU, correct power in/out to all ECU circuits (switched and constant - sorry no wiring diagram), and good ECU ground for F/I circuit to the body. If all that checks out, then your ECU is toast.

Forgive me if I'm reading you wrong, but I'm seeing a lot of maybes, not sure and I think so for the bits you are testing thus far. I'd suggest you run through what I gave you, track down a 98 S8 wiring diagram like your life depended on it, and get definite answers on each step of the process, and you will get it figured out.

Good hunting

hillpc

Yes, we haven't heard from fsrowsell (1998 S8) in about a week, since he said he was changing out a relay.

I'm chasing a no spark condition before I test any fuel system parts, on my 1997 A8.

This is a really good point (All 8? ) that hit home with me yesterday. My earlier work only checked one or two cylinders for spark, leading me to say "no spark". But yesterday I checked all 8 coils for supply voltage (terminal 15 on each 3-wire cable coming into the coil; all good) while cranking, and then checked for all of the coils being switched to ground briefly by the Power Output Stages (terminal 1 being grounded) while cranking. Did this with a test light (incandescent, not LED; this may be an important detail later in troubleshooting the ECM output), hooking it up to terminal 15 (+12v all the time with ignition on) and terminal #1 (grounded when the POS pulls terminal 1 to ground ["charging" the coil] and then quickly opened again by the ECM to discharge the coil into the spark plug. So it should flash. I got 2 out of 8 being switched while cranking! This kind of makes sense because the car sounds like it almost wants to start.

Then I swapped the POS modules for each other, and the flashing test lamp moved to different cylinders, on the other POS, but same output terminals. This implicates the POS inputs from the ECM POS output terminals, I think, but I have to think this through some more. It sounded like it wanted to start more eagerly, but I don't know what to make of this. Maybe firing order-related; I don't have the details of which cylinders with me now.

Now I want to go back and work more methodically, do an actual spark test on each coil, triggering them by hand (grounding #1 terminal while #15 is hot) instead of letting the POS's do it. If they're all OK, then walk upstream and trigger the POS's by hand to see if they will trigger the coils. For this I need to know the terminal assignments for input and output to/from the POS. The outputs I know, but need to study the manual to see if I can figure out which input terminals on the POS go to which outputs. I also need to know exactly what kind of signal the ECM sends to the POS. I assume it is some kind of positive voltage that triggers the POS to ground the coils. How many volts? If all the POS channels can be fired manually and make actual sparks at plugs, then I suspect the wires from the ECM to the POS (I have checked them for continuity and no shorts; they're good), or the ECM itself. That would be expensive, I presume. Repair outfits in the US?


I can scan and send pages, fsrowsell, if you can use them. I don't know if the S8 diagrams are any different from the A8. I'm assuming they're all the same for the D2 generation A8/S8.