1. USB connection works fine after being plugged in. I can remove the connector and insert it back. It works without issues.

2. One way to simulate the problem is by changing the serial port data. Once in a while, when serial port data changes, chip hangs. FTDI guys said that it could be because of 'errors in the pipe', or 'current flowing into the ground'. We are using a pipe reset attempt count of 300 against the standard 50. Not sure whether the baud rate mismatch (percentage error) is causing problem. But, for a standard data, I can keep it on for along time without any freeze-ups. At least, I haven't noticed a freeze-up till now. Most of the time data will be almost same.

3. Haven't checked the current flowing between two grounds. Will check it, but this issue is a common occurrence, I guess. If two floating grounds have a voltage difference between them, (meaning, a current can flow between them), still I can short them as long as the current is within limits. Current is limited by the total resistance of the path, and in this case, it would be limited by the current through RX/TX lines. Right?

4. Prolific chip based serial to USB converter PCB and moulded cable, bought off the shelf, are working fine in the same situation. But we do not get prolific chips locally.

Please comment...