Quoting Andrew Greenberg <and...@psas.pdx.edu>:
Bdale and I spent the day reviewing chip foot prints and have quite a
few minor updates. Bdale stuck them all on his server:
This directory has postscript versions of the schematic and the pc board
along with gerber files and a cnc drill file.
OK, so, I see a couple things that don't look right but that's probably
because I don't know which parts you're actually using. Do you have a
bill of materials with actual part numbers. Not just "miniUSB", but
actual mfg (or even better Digikey) numbers?
I've been going blind trying to crank out 3 board designs in the last
2.5 weeks, so I may be missing something here--- BUT,
1) Aren't the N-FETs upside down? The way I see it, the source should
be connected to GND, otherwise the internal body diode will be forward
biased, causing C1 to discharge through the main and/or drogue squibs
as soon as you plug them in. Further, in order to turn the FETs on,
you need the gate to rise x volts *above* the source, which is
presumably at 5V.
2) I haven't looked up the gate charge on the FETs, but as a general
rule I like to use a resistor between the uC and the FET gate (keep
pulldown on uC side). It limits the current draw on the pin, controls
rise time, cuts ground bounce, yada yada. It may not be necessary
here, as I'm usually driving PWM, and you are (most likely) not;
however, it's cheap insurance to protect the uC output.
3) I like the pull downs (R2 and R3) on the FET gates. That's
important when the uC goes high-Z after a reset.
4) Does anyone know what the LTC1682 or LP2980 does when it sees
reverse voltage on the input? A schottky diode or a PFET might save a
bunch of parts if someone connects the battery backward.
That's all I've caught so far. Wait... anti-aliasing? I'll look up
some of the part numbers and think about it.
psas-avionics mailing list