I would expect the forces in general use of a key ring would exceed 10's of Gs under normal use. People drop them on counters or drawers, they collide with keys, shoved around with other things in the pocket. I wouldn't want to live my life as a key ring. There is a logarithmic relationship between fatigue life and stress. A small decrease in stress increases the part fatigue life by a large amount. Typically, when you are less than 1/3 the ultimate breaking load of a nonferrous material you will be into the very high cycle life range (>10^6 cycles). I'm not familiar with electronics failure modes under fatigue so YMMV. 3G's isn't that much unless it is causing resonance problems. As a first pass test I would load the components to 9 Gs and see what falls off. You could spin it or attach an accelerometer to it and drop it onto a compliant surface to generate the force (soft surface better than hard surface). If that passes then go to 12 G's and if it survives you will have a decent safety factor. If it is safety critical or loosing it means big $, then it is off to the shaker table lab. Regards, Gordon Williams ----- Original Message -----=20 From: "RussellMc" To: "Microcontroller discussion list - Public." Sent: Friday, October 22, 2010 11:40 PM Subject: Re: [EE] circuit board under vibration > We are developing a controller board that will be mounted on a > machine, the machine will vibrate (shake) heavily, but should not > exceed 3g ... (balance at at end) ... I don't know the mechanism for the IC debonding mentioned above but forces in a pocket would not be expected to be large or especially frequent. Perhaps keychain vibration is the main mechanism? Loss of ICs invariably seems to happen. Solder would probably be tinlead as these were older SIMs some yeas ago. Maybe I should try it with a newer RAM. Russell McMahon --=20 http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .