I don't think that high voltage is making it out of the ADC 212/3, I think that the parallel to usb doesn't understand the voltage out of range signal that is being sent to it encoded as parallel data.
If I understand correctly how it works (and I probably don't
), the ADC 212/3 is sampling & digitizing the analog voltage input and sending it out as a digital signal in a parallel data format. When it hits the usb converter, the converter translates the parallel digital data into a serial digital signal.
From the looks of what is happening this conversion seems to be ok as long as the voltage stays within the +/- 20 V (or +/- 400 V with the 20:1).
Now, when the voltage goes out of range the ADC 212/3 sends out a digital signal that indicates an out of range voltage... which is correctly translated by the PicoScope software when it is sent directly as a parallel digital signal. However, when the usb converter gets this same signal it appears that it does not translate it properly causing the PicoScope software to lock up.
At this point the question is whether this incorrect translation of digital data is due to a defective converter or that the converter is working properly and just has no way to correctly translate the over voltage data signal that it is receiving.
And it looks like the answer may have to wait until I get the new usb connector to try out... Thanks Tom!