Posted 2012.07.15 14.18 in Computers/Internet/Technology by Stephanie

I’ve been working on this on-and-off for the past couple days – a second revision of the custom circuit board for my ISEB-6. After learning from all my mistakes on the first revision, this one is a thing of beauty, if I do say so myself.

It’s got a great big gorgeous ground plane, on-board USB port and FTDI chip, all the same sensors and expansions as before, but I think it’s much more neat and tidy than the first attempt.

There’s three separate Vcc networks (VUsb, VBattery, Vcc) and all the supply traces have been beefed up to hopefully ensure there won’t be any losses to heat or resistance.

Just about everything has its own 0.1uF filter capacitor and the caps are as near their components as I could manage, while still keeping it possible to hand-solder it all.

The analog section is relatively compact and well grounded, the analog leads are all as short as possible.

As for the various problems I had identified on the previous board, I think I’ve solved them one way or another:

Better ground, shorter analog lines, and properly-positioned filter capacitors should solve all my fluctuating analog reads. I’ve already got the current version working acceptably by giving it a more stable power source and cleaned up the software.

I’ve decided to stick with the OLED screen. The Nokia LCD is nice but it’s just too darn big to fit. Staying with the OLED, I have corrected the footprint so the board will fit, and am working on a method to dim the display so at night it won’t light up my wrist like a beacon.

I’ve also opted to stick with the ceramic resonator rather than switch to a crystal. Yes, a crystal would give me a much more accurate clock, but I can work around the resonator. In the final build I will measure it’s drift and compensate for it in software. It wouldn’t be a big problem anyways because most of the time, the ISEB-6 uses GPS satellites for its clock source instead of the internal software clock.

Finally, I have also chosen to stick with the ATMega1284P. It’s a good microcontroller, lots of folks use it, my code is good, I’m already familiar with it, my bootloader works fine, and with the on-board FTDI chip, I have on-board USB, which was the only thing I thought was lacking in the ‘1284P.

At this point, I’m just reviewing the design to ensure I didn’t miss anything or make any mistakes, before I send it off to OSHPark for fabrication.

This time, I’ll try the rush shipping I think.

Leave a Comment