Regarding motor control circuits..

I've been having my own, different, issues here. I'm using a FT232RL USB breakout board to interface to the H-bridge circuits, all controlled from my MythTV box (for Frankenswitch V2).

The problem I see, is that when the motor is under high load (aka. "stalled"), sometimes the FT232R disconnects/reconnects on the USB side, and consequently gets reset by the host. The host is using Linux (of course), and I looked into the low level code there. It's simply the UHCI hardware reporting a "connection change" event, meaning a disconnect/reconnect.

So I figured maybe a voltage brownout on the USB cable was the cause, and then went through a progressive series of increasing the circuit isolation.

Now the FT232R breakout board is still bus-powered, but simply connects to some opto-isolators, using internal (5V) drivers through 10K resistors to the O-I to ground.

The motor controller side of the house is now triggered from the far side of the O-I, and runs on a totally independent PSU, all the way to the 120VAC wall outlet. Not even a ground wire is shared.

The USB now behaves much better, but I can still cause it to disconnect/reconnect with repeated stressing of the motor.

Why?

The obvious guess here, for me anyway, is EMI from the motor, since I'm doing these tests with 6" or less physical distance between the motor and the FT232R.

I suppose a nice grounded metal/foil shield, or simply more distance, should cure this. Or is there something better (yet easy) to also try?

EDIT: Information on the FT232R Breakout Board is here. I have tied pin VccIO to Vcc.

Cheers