[Cryptech Tech] FMC Arbiter

[Fredrik Thulin]

On Thursday, August 27, 2015 12:17:05 AM Павел Шатов wrote:
> Hello!
> 
> I uploaded initial version of FMC interface arbiter for Novena and
> corresponding demo program for the dev-bridge board:
> /user/shatov/fmc-test
> /user/shatov/novena-fmc-arbiter
> 
> It would be great, if someone else could test them on his set of boards.

Great work! I'll test with four different dev-bridge boards and then send one 
to Joachim.

Paul, I think you want one soon-ish, as in before our next f2f meeting? In 
that case I will ship you two boards and you can get one to Rob.

> Speaking of hardware issues, I got the following:
> 
> 1) One of the expansion connector pins was not soldered well (see
> attached photo), as a result, one of the address lines was always read
> as 0 by the FPGA. In case you observe bits stuck at 0, inspect expansion
> connector for bad solder joints.

My bad. Bottom side is hand soldered and I really miss good magnification. 
That's why I was waiting for this code to be able to test the boards before I 
send them to people who also lack fine soldering tips ;).

> 2) Expansion connector turns out to be very sensitive to misalignment.
> Sometimes the demo program starts getting readback mismatch errors, that
> are all in the same bit. After I gently push the bridge board to align
> it in parallel with the base, everything starts working fine again. I
> also think, that such connectors are rated for a limited number of
> mating cycles, so I try not to separate the boards without reason.

That's... disappointing. Good thing we're working on the Alpha board.

> 3) My Novena somehow doesn't work properly, when bridge board is powered
> from it. I removed diode D2 and added solder bridge instead to power my
> bridge board from wall Micro-USB phone charger.

Interesting. I don't remember ever seeing any power ratings for the 5V rail 
from the Novena, but we're not consuming very much at all with the dev-bridge 
board IMO.
 
> 4) 25 MHz HSE crystal on my board doesn't work, that's why I temporarily
> configured it to operate from 16 MHz HSI oscillator. Fredrik gave me a
> hint on how to fix this, but I haven't time to get to an RLC meter so far.

My QA department did a rather poor job in the hectic last few days before the 
Prague workshop it seems.

/Fredrik