John, Excessive voltage gradients are what causes failures in this test. I'd suggest looking at each chip and finding the worst case voltage gradients between pins. You can start with your supply voltage (battery or AC) to ground separations. If you have a voltage gradient in excess of 11V/thousandth of an inch, this would be a good place to start looking for issues.
The best, but most painful short-term solution, would be a design re-layout of the offending separation. I've never used conformal coating due to other issues (reliability, thermals) but that may be a way to go as well depending on your product. You may also want to go to a better air filter if you have the space and/or thermal margin. Good luck, Marko Radojicic email: ma...@caspiannetworks.com phone: 408/382-5206 fax: tbd -----Original Message----- From: Kretsch, John [mailto:john_kret...@adc.com] Sent: Monday, September 18, 2000 8:06 AM To: 'nebs@world.std.com'; 'EMC PSTC'; 'TREG Newsgroup' Subject: Hygroscopic Dust Troubleshooting I thought I would try to contact the collective on this one... We have a shelf system that is failing GR-63 hygro dust with excessive bit errors (unit did pass Gaseous contaminants no problem). Air is filtered. Anyone have experiences that they would like to share (non-proprietary of course) about how to trouble-shoot and solve this problem? This was the only GR-63 test to have a failure. Regards, John K.