BrianC72gt

I'm not an elelctronics guy, but when I disassembled the cluster there appeared to be a power supply (assuming 5v) surface mounted to the inside of the back cover. Easy enough to get open, to quote Slim Pickens,"We're gonna need a ****load of dimes." Shim a dime in at each retainer prong and it will come apart. Been a while, maybe a few screws at the back as well. PaulW walks you through it on Audipages.

And take a long look at the pins at the back of the cluster, they bend easy and can give weird results. Doubt it here, but it happens. Pull and reseat everything a couple of times to get some friction cleaning going on. Did I mention spray electronic contact cleaner recently?

BrianC72gt

The CEL coming on is interesting, Usually key in, turn to RUN, crank and CEL stays on until you start. Start is determined when flywheel sensor confirms crank speed is above a few hundred rpm. Engine is running, so CEL switches off. Either your ECU choose that moment to register a fault code - doubtful - or the cluster decided the engine wasn't running. But the engine WAS running just fine. If the ECU thought the engine stopped spinning, it would have cut fuel. It didn't, so the sensors are telling the ECU that the engine is running.

Either the Cluster is getting erroneous info from the ECU or the Cluster is glitching. If it was an ignition switch fault, then the engine would have stumbled or dies. Also, that previous IMMO code is strictly cluster based.

The one error code your reported, P1185, is that a ground fault on the O2 circuit? Can your software datalog sensors, or monitor them in real time. I have an old HEX Key Vag Com cable, and can monitor output on sensors which can be very helpful in chasing intermittent faults.

Keep at it, you'll get it for the price of your time and not much else.

Brian

mz250

Ok, my mistake (again). Code is P1850, powertrain data bus missing message from ECU or something to that effect. Odd thing, when I try to clear it, I hear a distinct CLUNK from behind the console somewhere, but I could just be hallucinating because its about 8 degrees out side. I did take the whole cluster apart, checked every pin, cleaned all the connections, put it all back in, no change. When i start it the first time after a long wait, I get no IMMO light, no CEL. The after the cluster dies (engine running) and restarts, I get the CEL and IMMO lights again, and they come back as the cluster repeatedly resets. I'm thinking this is beyond my diagnostic abilities at this point. I'm just going to send the cluster to one of those guys that tests and fixes such things, there are at least two I've found. I have to at least rule the thing out. I appreciate your input, and I'll update on the results as they come in.

a8gd

Any way to just disable the IMMO? I know it is possible on 2000 ish vw's

BrianC72gt

Before you send out the cluster, I would unplug the radio entirely from the harness. You said the radio was running in safe mode? I can tell you two items that did get wet, the radio, and ecu, but not the cluster. A dead radio with a blown fuse will shut down your coms entirely. Who's to say a gimpy one wouldn't twitch intermittently. The other possbility is the signal to the cluster from the ECU is dodgy on the ECU side.

BrianC72gt

I p'm d you about a fellow enthusiast on another forum about a world of D2 quattros parting a 2000A8L project that isn't happening. Maybe you could strike a deal for a key, cluster and ECU. Plug those 3 into your car, use the cut key blade from your key, the key head with pellet from his, and you should be off to the races.

mz250

Hey, thanks for the link. I'm going to pursue that option, if he does take about $150.00. As far as the IMMO delete thing, I had a donor ECU I sent off via an Audi specialist (not dealer mechanic) to get the IMMO deleted, or so I thought. He said he'd had it done on his S8 and worked great, but mine sure didn't. With the supposedly IMMO-free computer in, the car won't even start, and the cluster does the same narcoleptic thing it does anyway. I'll try the radio tomorrow just to see, but I'm really thinking the cluster is f**ked. This is likely the first literal "clusterf**k" I've ever really seen. Anyhow, I appreciate your time, fellas, and I'll keep this updated as things progress.