Below are the responses I received on this topic and a copy of my original e-mail.
Thanks to everyone who replied - your responses are insightful and bound to be helpful, and are much appreciated. Jim Eichner Statpower Technologies Corporation [email protected] Any opinions expressed are those of my invisible friend ---------- Jim I have quite a few products with the same basic user interface you describe and have tested extensively to ESD. To answer some of your questions: If the metal plate on which the pushbutton overlay is applied is earth grounded it should attract the ESD discharge. I assume that your device is panel mounted and there fore the dischrage path would be to the panel. If for some reason you must float the panel, the overlay can be designed with a grounded internal shield ring around the perimeter of the overlay. This would be an internal trace just like those to the pushbuttons but positioned near the edge. ESD applied to the edge of the overlay will then be attracted to this sink path. Be careful about routing and the design of the drain lead. If you have a plastic bezzel that the plate floats on you could consider some conductive plastics. They act as a good sink path for ESD but will most likely give you some fits with safety engineers because it acts as both an insulator (@ power frequency) and a conductor (@ High frequency like ESD). The bezel usually is treated as a floating piece of metal in safety review. After all of this ESD concern you may find that you actually have a bigger problem with other common mode immunity tests - Fast transient and RF immunity (especially the RF conducted). You see, if the overlay panel is grounded you have a good common mode path through the overlay circuit. Common mode High frequency noise will find its way to ground through the push button circuits and can cause false push button action or actually thrash the data bus. This can also be an emission issue since the path allows radiated noise to escape too. A simple approach is to have hefty pull up resistors, but you most likely will need some form of filter. You may want to consider ferrite filter beads . On the display placement you have to way two issues. If the LCD is pushed up against the overlay membrane you may have a better looking display. The image will tend to be crisper and the viewing angle optimum. However, if you back the LCD off a quarter of inch or so then you have some level of impact resistance. Note that the LCD glass can be fragile in some designs. This separation would allow the overlay to absorb user impact. If the LCD is backlit then the image may be fine with the separation. I believe the best resolution is mock up a display and take a look and try the impact resistance as well. I believe our overlays are basically polly ethylene and the adhesives are a 3M safety recognized material (tested not to fall apart under temp or time). If you want to get into more detail let me know. Like I said I have been through this design ringer before. If you have any additional questions feel free to email me Regards Chris Wells Sr Design Eng. Cutler-Hammer [email protected] __________________________________________ Having done some recent ESD testing with membrane keyboards I can perhaps pass on some things I learned. Membranes are typically polyester or polycarbonate. Polyester is the best. 10 mil. sheets of this material will give you around 15-20Kv of protection. Embossing the membrane will reduce this thickness. At 8Kv you most likely would not have a problem. The thru-surface arcing is usually where the energy gets into your product. The energy will arc through any opening, or crack, however tiny, it can find when attempting to find a path to ground. I have read that 1 cm spacing is sufficicient, but this has not proved out in the lab. We found that for 12-15Kv 0.5 inches of surface spacing was an absolute minimum, with 0.75-1.0 inches being best. We found that the best solution was to attempt to re-route the ESD energy to a large chassis. This is not always available. A "Static Shield" between the membrane layer and the can help if you can bond it directly to chassis. The worst situation is when the ground path that the ESD pulse takes is through the ground traces in your ribbon cables and PCB. Once it gets into your circuits there is almost always a problem. Darrell Locke Advanced Input Devices ________________________________ Jim, AFIK the LCD display manufacturers don't have a problem with ESD (although I don't think they test to the level to which you are subject). Most notebook computers don't even have a plastic window covering their LCDs. Most calculators or clocks don't either. How thick is the top layer of your membrane switch/window? I believe that if you have sufficient thickness in this part, you won't have an ESD problem. Then again, I haven't a lot of experience with either membrane switches or LCD displays, so all I am going on here is a hunch... Steve Chin StreamLogic Corp. Menlo Park, CA, USA [email protected] _________________________________________ In my experience, if you use a 'normal' proprietary membrane switch/panel, with the half inch dead zone at the edges that you describe, you wont have an ESD problem, at least not by breakdown of the membrane at 8kV. However, ESD discharges at the periphery of your membrane panel, if mounted on a conductive (metal) panel, will cause highly intrusive spikes to be generated within the membrane conductors. It would be wise to decouple the ultimate destinations of these conductors with a small capacitor to avoid spurious operation of the related electronics, particularly if you have high speed latches etc. being driven by the membrane switches. LCD panels are temporarily visually affected by the 8kV potential present on an ESD probe while it looks for a discharge path, happily the front glass of the LCD panel is easily capable of holding off 8kV, and I've never had an LCD driver fail due to induced energy from adjacent discharges. (YET!) Another tuppence worth... Regards, Chris Dupres Surrey UK. ----------------------------------------------------------------- Jim, I had the same problem (design issue). The key here is to have the longest pos ible path length between the ESD source and the potential circuit trouble poin . Several things will help reduce or eliminate the problem. 1. Add a guard band, in etch, inside the switch layer. This guard band should over the entire perimiter of the layer and should be connected to chassis grou d (the metal panel). This can be done with a tail that connects the band to a ounting screw for the panel. It needs to be adequately spaced from the active ircuit. Your membrane switch vendor should be able to help with this point. 2. The window in the label will satisfy your safety problem with LCD's. That i , the user cannot come into contact with the leaking liquid from a broken LCD. 3. You don't necessarily need an air space between the window and the LCD, but it helps. In one case, I added a 1/8" spacer between to keep the discharge fro getting coupled through the LCD to the driver circuit. 4. Many LCD's come with a metal band around them tpo hold them to the circuit oard. This metal band may or may not (usually not) be grounded to chassis. In his case, I designed a beryllium copper spring that attached to the mounting s rew and made spring contact with the metal band. With the above changes, our design now takes 16 kV and won't fail. Occaisional y we see a momentary darkening of the LCD, but there are no operational failur s or damaged parts. Hope this helps some, Scott Douglas Principal Compliance Engineer ECRM Incorporated Telephone: 1-508-851-0207 Facsimilie: 1-508-851-7016 e-mail: [email protected] _______________________________________________________ From: [email protected] on Fri, Jun 6, 1997 7:59 PM Subject: ESD - Membrane switches and LCD Displays To: [email protected] I am looking for some guidance on the following construction: We have a remote control panel that contains an LCD display (approx. 2" x 1") mounted on a circuit board behind a metal panel. An opening in the metal panel allows the user to see the screen, and is only marginally larger than the display. Over the metal panel is a laminated label containing some membrane switches. All the circuit traces for the membrane switches are approx. 1/2" or more from the edges of the label, and are contained in the bottom layer of the label (ie one layer of plastic between them and the metal panel). The ribbon cable from the membrane switch to the pwb is routed through a hole in the metal panel. There is a clear plastic window in the label allowing viewing of the LCD and that window is more or less touching the glass face of the LCD display. We don't want to recess the LCD because of viewing angle problems. We are subject to 8kV air discharge and 6kV contact discharge. 1. Does this construction sound like an ESD test failure waiting to happen? 2. Does it make any difference whether or not there is an air space between the clear plastic window in the label and the glass face of the LCD display? Should we recess the LCD, and if so, by how much? 3. Are there preferred materials for any of the layers of the membrane switch label? 4. Any hints anyone? Any help will be greatly appreciated. Thanks, Jim Eichner Statpower Technologies Corporation [email protected] Any opinions expressed are those of my invisible friend
