Tuesday, September 9, 2014
DSP Troubleshooting and Integration
The car has been at the body shop for a bit over three weeks at this writing so I've spent some time integrating the PCBs in the enclosure, building the wiring harnesses and coming up with a method to connect the system with the vehicle's harness.
Since I hadn't touched the boards in some time I attempted to duplicate the overcurrent condition I experienced earlier. I began by asserting the Remote Input signal associated with the first MiniDC and that worked normally. I then asserted the Remote Input signal to the other MiniDC and that too worked as expected. I then asserted both at the same time (or what I thought was the same time but was in fact perhaps a half-second apart) and much to my surprise it worked. I also noticed at this point that the supply indicated a very brief current peak of 575ma after which it settled down to 525ma. I then tied the two Remote Input signal wires together (which would guarantee that both boards received the signal at the exact same time) and like clockwork my supply went into current limiting when set to 650ma.
That's when it hit me. The update frequency of my supply's LCD was probably not high enough to display what I estimated to be a very brief peak current draw exceeding 650ma but the current sense circuitry was still able to detect and respond to it. I still wasn't sure why the current didn't settle back down after the initial peak and allow the power supply to exit constant current mode but I threw caution to the wind at this point, set the current limiter to 800ma, and tried again. With the remote input signal wires still tied together I asserted the signal and watched both boards come up and the supply settle down to 525ma as expected. This convinced me that the problem was an artifact of my power supply all along and the simple result of the boards drawing much more current on startup than I expected. Despite the hair pulling, better safe than sorry. The perk is that I've now characterized the startup behavior so I know what fuse value to use to protect the assembly.
After reviewing the E36 schematics I discovered that the amp harness and the nearby telephone interface both provide a switched 12 volt signal (what the manual calls “HOT ON ACCY, RUN, AND START”). Fuse 43 (5A) protects the affected circuitry, and given that several items on that circuit are either not installed (a convertible top control module) or no longer functioning (makeup mirror lighting) it's unlikely I'll exceed the available power. I'll probably take the switched input off the amp harness because that is closest to the vehicle harness junction box I'm developing.
I also began work on the wiring harnesses, but the solder cup d-sub connectors have proven more difficult to prepare than I anticipated. Next time I'll buy the crimp versions. Had I done that I would have completed all the wiring in less time than I spent soldering a single d-sub connector and I'd have the ability to pull and reinsert individual wires far more easily if necessary for troubleshooting. Oh well. This is what I get for being frugal.