I've been where you are at. You need to get the idea that everything is working correctly out of your head. There is a problem and you are failing to find it. The tests you are doing or the parts you are replacing are faulty. It's your fault and not the car.
Don't take this personal. I've been down this road more often than most. It was my job for 20+ years to fix Mercedes that the others [including their engineers] could not. If everything was good, there would not be a problem. Everything is not good and you are not testing it right. Once you wrap your head around this, then you can get back to testing. Every time I got painted in the corner, I had to look at myself and ask myself why I cannot fix it. Telling myself that everything is working right but the problem is still there is wasted time.
There was only 1 Mercedes I could not fix. The factory support was stumped too. After 20 hours of diagnosis, we had to shot gun parts and since it was customer pay, the client pulled the plug. They pulled the plug since I could not assure them if any part I replaced would fix it and the parts I wanted were coded for their car and they could not be put back on the shelf.
If I have a chance to play with my wife's car, I will. I will post my results and try to get some pictures.
I'm trying to get a recently purchased Vanagon ready to go on a vacation for a client. They're leaving in a few days. So I have not had time to check my wife's car.
You are performing a test against the positive but that does not mean much since you do not have a reference to what it's suppose to be. So is it a good test or not. Right now, it's worthless information.
The test Bentley [again, I absolutely hate to be wrong and I rarely assume since it just leads down the wrong path] says is to use the ground pin. The test is measuring the power signal and you are not getting one. Sounds like that is the issue. Your test is showing it's getting a ground signal but the computer does not care, it needs a power signal. Does the connector need to be plugged in? Your car is old, but with the newest cars, most test must be hooked up or the computer may remove power/ground because it 'sees' it unplugged. If Bentley says to have it unplugged, maybe it's true but I would perform the test plugged and plugged in.
I'll do my best to make those tests today, but I do not want to look at this young family and say you're not going on vacation. I certainly do not want to get a call saying the Vanagon died too because of my hastiness.
So what makes you think that the ECU puts power on the line? I would think that there is a slight bias voltage on that line and the ECU pulls it to ground. I'm certain you will agree, it's got to be one or the other but it's possible your thought is wrong. Yes, I know I could be wrong too. If there is a bias voltage, I would suspect the ICU in the coil.So I am unsure why the ECU does not send the positive voltage down the wire, the only explanation I think of is that the wire is grounded somehow?
The reason I think that is because Bentley says so and it very clearly fails that test( with the wires "floating" as in coil completely unplugged). The LED would only light up if its get positive from signal wire and the waveform info from Autodata and Vivid confirms it too.So what makes you think that the ECU puts power on the line? I would think that there is a slight bias voltage on that line and the ECU pulls it to ground. I'm certain you will agree, it's got to be one or the other but it's possible your thought is wrong. Yes, I know I could be wrong too. If there is a bias voltage, I would suspect the ICU in the coil.
On the bright side, I replaced my wife's valve guide seals [which I did 2 years ago] because of a misfire and blue smoke on cold start ups. It's fixed, she just drove out and no blue cloud.
Does the engine lock up?
Not certain what you mean by half a key.
Does the engine lock up? If not, I would think there is an ignition timing issue. A spark at the wrong time will stop a cranking engine.