Close

First prototype report

A project log for GPS Clock

A simple desk clock that gets extremely accurate time from GPS

nick-sayerNick Sayer 12/11/2016 at 09:170 Comments

I had an earlier log that spoke of abject failure, but that needs to be corrected. I got the prototype to work.

The problem, indeed, lay in the fact that the legs of the QSOP package of the LED driver could barely reach the pads, and thus positioning was a fiddly and critical thing. I've already ordered new boards to correct this.

Also, I found a software bug that kept the self-test from functioning properly - I failed to notice in the datasheet that the register numbers skipped from 4 to 7, so I was writing the self-test bit on register 5, which doesn't exist. With that corrected, the display lit right up.

Although the display did light, the leftmost digit wasn't working. Long story short, the D0 line of the charlieplexing matrix was shorted to ground. This, I believe, was a PCB fab error. I cut the trace in a couple of spots and reconnected the cathode line of the left digit past the cuts and it all worked beautifully.

The project isn't 100% there yet, though. I suspect as we go through the days ahead there will be firmware bugs to find and fix, but right now, I'm going to bed.

Discussions