Thanks Mike,
I should know better. I've worked on these types of bugs in our computer servers. You turn something off and the circuit load doesn't decay the voltage fast enough to get it below the trip threshold so it doesn't actually turn off. Or half the parts turn off and the other half don't and the system hangs. Sometimes you give it more time, sometimes change the circuit, sometimes add a load resistor. Using the digital voltage meter gave me that answer after talking to the service rep. The guy says 0V, but I know from the meter that the trip voltage is somewhere around 3V, which is very reasonable for a noisy automotive environment. I'm not certain why changing from a transistor to a relay would make a difference here since there seems to be some current bleeding from somewhere. Hmmmm, I've got a few theories, but can only confirm with an oscilloscope. Ok, now i've got to look for a new o-scope to get the waveform. What's an EE without a scope? Always an excuse to get one. Wow, just looked them up, there are lots of options reasonably priced. Will get some pictures and post in a few weeks.
Mark