Saturday, October 6, 2007

Attempted debugging

In a last-ditch attempt to get the car going for the Homecoming Parade, I stayed up late one more night. I hacked apart a spare RS232 cable, and soldered insulated quick-connects on the end in preparation for hooking it up to the controller:


Here is the cabling, plugged in. This is a real royal pain - you cannot easily see the top three pins due to the position of the controller in the rear trunk. I had to use a flashlight and a mirror to get the pins plugged in correctly:


After much fiddling and cursing, I finally got HyperTerminal to listen to the controller:


Azure has been extremely helpful during this - responding quickly to emails, and sending me files to try. Apparently, my controller is older and has pretty ancient firmware. They recommended using the older ccShell 2.0 with the controller - but it does not seem to work. I wonder if my firmware is even older than they realized. They have offered to upgrade the firmware, but I'll have to send the unit to them to have that done. I'll do it if I have to, but only as a last resort. I did order the official serial cable that plugs into the data port. I anticipate many future debugging sessions...

Since the car was not running, I walked the parade. Not a terrible conclusion - but a little disappointing. However, it was useful to have a deadline so that I would actually get the car basically done (I'm going to change the subheading on my blog to delete the "So far, just destruction, but will chronicle the build in coming weeks" tagline).

I took the rest of the day today to do a variety of miscellaneous things, including installing the hoods, putting the side windows and outside door handles back in, and installing some exhaust fan ducting - I found the smallest flexible heater ducting that I could and compressed it down with the cable clamps. Here it is in place, with the car plugged in and charging:


So, next steps - get the official serial cable from Azure, try one more debugging session, and, if that fails, send the controller back to Azure for a firmware upgrade.

4 comments:

TimK said...

That's a great looking car with the silver paint job and all! Sorry you didn't get to drive in the parade.

I'll be interested in seeing if your CCPower DCDC converter fails like mine did. If you're waiting for Azure Dynamics, you could turn on all the accessories for 20 minutes and see how well the CCPower module works. While this seems like a stress test, it's typical to have the lights, turn-signal, brake lights and fan running at night.

Ross Cunniff said...

Thanks, I'm pretty proud of how it looks :-)

I ran the highbeams, turn signal, blower (on high), and wipers for about 20 minutes last night, and the little red light on the DC-DC converter stayed on the whole time. I'll double check this morning.

Joe said...

I think Tim indecated that the LED lights when there is 12V from the battery or from the device. So it may not be 100% acurate.

Ross Cunniff said...

Joe, good point. I pulled the fuse from the DC-DC converter, and it still looks good. So, either the converter blew without killing the fuse :-) or mine seems more robust for some reason.

On the other hand, I've never subjected it to a regenerative braking session...