Ron, Often times, ESD related failures can't be traced to the event because they can stress semiconductor junctions without causing an immediate failure. Then a short time later--days, weeks, even months-the stressed component fails prematurely for no apparent reason. So even if a specific case can't be named, it doesn't mean that I/O pin discharges in real life don't cause failures.
Concerning removal of the cover, the assumption is that one is removing the cover to connect a cable. At the time of removal, the charged individual will discharge to the D-shell or chassis, assuming the individual is not using a non-conductive stick to remove the cover. The pins will not be subjected to the discharge. Of course a cover would not do much if there is no alternate discharge path nearby or if the customer removes the cover just for the heck of it. We should consider that one can always find a way to break something. I would hope we design for the majority of events rather than for making things 100% foolproof. regards, [email protected] ---------- From: RON_WELLMAN To: emc-pstc Subject: Re: ESD testing on exposed connector pins List-Post: [email protected] Date: Wednesday, April 17, 1996 8:14AM Has anyone ever attributed a real Customer failure of a product to an ESD discharge at an exposed i/o connector? Also, putting a cover over a connector gives us a warm fuzzy feeling that we have protected the product from the Customer and met the requirements of the standard. What is there to prevent a Customer from removing the cover? Regards, Ron Wellman +============================================================+ |Ronald R. Wellman |Corporate Quality Department| |Hewlett-Packard Company |External Product Regulations| |Product Processes Organization |Voice : 415-857-6059 | |1501 Page Mill Road, MS 5UL |FAX : 415-857-6340 | |Palo Alto, California 94304 USA|E-Mail: [email protected] | +============================================================+ | "Common sense is the collection of prejudices acquired by | | age eighteen." - Albert Einstein | +============================================================+ ______________________________ Forward Header __________________________________ Subject: Re: ESD testing on exposed connector pins Author: Non-HP-owner-emc-pstc ([email protected]) at HP-PaloAlto,shargw3 List-Post: [email protected] Date: 4/16/96 12:47 PM re: "ESD testing directly on the pins of a bare connector." It seems reasonable to exclude testing into a bare connector which will normally be populated with a cable. In that case, testing should be conducted with the cable in place. I think where confusion exists is where there are optional connectors, one or more of which may not be populated in a customer's installation. As an example, many of our laser printers normally have both standard serial (RS-32) & parallel I/O connectors. The customer will choose one and leave the other vacant. That unused connector is now vulnerable to ESD. Isn't this a case where the connector must either be tested directly or protected somehow? For our products, we recommend that such connectors be tested & shipped with some sort of protection. I've found that simple plastic dust covers are usually sufficient. This is assuming, of course, that there is a real risk. We have found, as has been noted by someone else, that typical 9, 15 & 25 female pin D-shell connectors are safe without additional protection as we've never been able to arc directly to a pin during tests (air discharge, of course). But perhaps we're being unnecessarily conservative. I'd like to see more opinions from others who've been through this. Jack Cook, EMC Competency Center Xerox Corporation [email protected]
