New project takes life
I just finished a new PCB that I’ve been thinking of making. It’s general purpose car interface. I left a lot of headers on it so I can reconfigure aspects of it on the fly via jumper cables.
Some of the features it (can) support..
- GSM cell modem interface for data and voice (incoming only) – interfaces to car’s built-in cell phone system. The radio should mute and say PHONE on the display with built in hands free.
- GPS modem I salvaged from a car tracking device I had in my car. The company that ran it went out of business. I hope it still works.
- 12-Bit digital accelerometer, for measuring your acceleration of course.
- microSD card socket for recording lots of data
- switching power supply (5V + 3.5V) – much more efficient and less prone to drain car battery
- Valentine V-1 interface (detect radar and auto-mute when driving <25mph).
- iPhone interface (just a header for now, but a simple interface should be feasible)
- OBD2 & engine diagnostics (VAG protocol + CAN). Not sure why I added CAN given that my car doesn’t use a CAN bus.
I spent the day trying to debug a number of hardware problems. The first one a pain to debug – the power supply IC was freaking out and producing some high frequency noise which was audible on it’s inductor. It turned out to be a current sense resistor that I put right before it. It impeded enough current that it prevented the IC from working right. The reason it was a pain to debug was that I had tried shorting out the resistor with a wire earlier but it didn’t help. Many hours later and out of options, I tried soldering a wire to short it, and that worked. Guess it needs a really low impedance source
The next problem was my comparator that I use to drive the Audi’s K-Lines (for diagnostics). I used a preexisting package, which had poor labeling of pins and I ended up switching the +12V and GND supplies. Fixable, just ugly.
I also think I fried my accelerometer and have to replace it; while debugging the supply issue, I put 6V on the 3.5V rail and possibly fried it. Most other ICs populated at the time should have been able to handle that voltage. Unfortunatly, I think that was one of the most expensive components on the PCB, around $13.