Board bring-up

[zdun8]

• R125 is 52R3 and should be 52K3
• All power rails are up and have correct output voltage
• Connecting CMOD and ARDUINO works, no shorts, no increased current consumption
• PROBLEM CMOD is powered from 3.3V on the board, but USB is 5V, so we can't program it while it's connected to the board. I will try with a USB cable without power line

[gkasprow]

the last problem might be caused by back-supply of the FTDI chip and has nothing to do with USB power. If you connect power to such back-supplied chip, you can damage it

[zdun8]

I haven't connected any 5V power to the CMOD while it's powered from 3.3V as I didn't want to risk damaging it.
@gkasprow do you think removing USB 5V line from the cable will work, while CMOD is attached to the board and powered from 3.3V?

Here is also a diagram of the power circuit of CMOD:

[gkasprow]

this LTC chip requires more than 3.3V. It should work from 5V rail.
With 3.3V FPGA module won't work
FTDI is probably supplied from USB so it needs own power

[zdun8]

CMOD spec says we can supply it from 3.3 V (min).

[gkasprow]

OK, with very little power drawn from 3.3 that would work. The buck would have to run at 100% duty cycle. Note that when you plug USB, you get 4.3V instead of your 3.3V rail on the DIOT board

[gkasprow]

4.3V will probably damage LVDS translators and many other chips.
It seems you have to get rid of diode from CMOD module

[zdun8]

Let's define the problem: how can we debug the CMOD through USB while it's connected to the board where it's powered from 3.3V. My view is that the only sensible way at the moment is to remove the 5V line from USB cable, USB differential should work with just GND? Another option is to change the power to CMOD from 3.3V to 5V, but that requires some tinkering, but should be doable.

[gkasprow]

No, removing 5V line will disable FTDI chip.
Just remove the diode from CMOD