Solution:

After quality review meeting(2/10/1998), I would like to address clearly some points which is common understanding shared by all attendee: 1. Triden Chip is an old design and may not have a "strong" protection against the surge from VGA monitor. We don't see any protection circuit proposed by Trident. We also notice in the Chip design, some clamp diodes build in for protection is common and should be included in the chip, buy semiconductor solution, not by extra circuit outside the chip. 2. The possible reason of destroy is reverse voltage. The leakage current from monitor to PC will destroy the chip, if too large current detected. Normally, we treat monitor as a current sink and should not have any voltage detect when PC off and monitor on. However, a lots of monitor do have pull high circuit to make sure the signal level is good enough for long distance connection. In this case, you can easily measure a voltage level at VGA connector. Now, the question is how large the voltage and current will backward input to PC. Again, if chip is designed well in advance, this kind of backward risk is acceptable. 3. Advantech's background record on this issue Last year, we notice there are some reports from the field that monitor screen will black out when operate in high temperature. After studying and testing result verified by QA lab. we implement a A-204 version change during last summer. Some of customer, we also suggest they had a simple rework themselves on old A-203 version boards, to solve the problem. The rework instruction send to the customer only if they report the problem. But, all the CPU board with A-204 PCB implement the solution completely. After September, we didn't receive same complaint again. 4. As for chip destroy, our total failure rate in the field is around 5%. If you get rid off so call "test OK" items, it is around 3% in whole year. Not much but still higher than the average . Most of the failure is cau