summaryrefslogtreecommitdiff
path: root/hw1/README
blob: 9ac2b15c923673d6d25af1bd5f6429af34f052c0 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
This is a board designed to receive nucleo64 (and maybe 32) boards
to be test TARGETS.  The stm32l1 on _this_ board is reprogrammed via
one of the debug connectors, with appropriate software to be
master/slave/etc to correspond with the DUT.  All (most) communication
lines between the two boards are tapped to a 10 pin connector matching
cheap FX2based logic analysers, allowing automatic tests to be written
that use sigrok to check that the signals on the wire match expectations.


Notes on assembly hw1-a:
* 16Mhz crystal, 15pf load caps, and 3k7 i2c pullups
  (why don't I have 4k7 0603 resistors lying around? who knows!)
* A power LED wouldn't hurt.
* Adding a red/green user led wouldn't hurt for test pass/fail, even though
  it's meant to be autotest
* maybe a fet to turn on/off the i2c pull ups programatically?
  (maybe just a single jumper? is there really any need for individual control
   of the pullups?)
* move stlink connector 90° or just down and tuck it in, so we can pull the
  edge in.  It overlaps the buttons on the nucleos, and requires longer
  pins than necessary otherwise.
* Add some ground pads for test points, just in general.  Test points good,
  more test points better.