More on the MLCC capacitor cracks, I have made further searches and found i= nteresting documents which I'd wish to share: Board Flexure Damage Can Be Hazardous to the Health of Your MLCCs: http://www.vishay.com/docs/49096/technicalpaper.pdf Flex Crack Mitigation: http://www.kemet.com/kemet/web/homepage/kfbk3.nsf/vaFeedbackFAQ/683DFC61AF8= D7CA18525745900743238/$file/Technical%20Overview%20of%20Flex%20Mitigation%2= 0Solutions.pdf Crack types: http://www.niccomp.com/MLCCcracks.asp MLCC cracking: http://www.gideonlabs.com/posts/mlcc-cracking/ PCB poor solder lead to EOS MLCC Fracturing: http://www.gideonlabs.com/posts/pcb-poor-solder-lead-to-eos-mlcc-fracturing= / MECHANICAL FLEXURE CRACKING: http://www.niccomp.com/help/techinfo/MLCCMechFlex0400.pdf When cracks occur in a chip monolithic ceramic capacitor, what mechanism ca= uses the capacitor to fail? http://www.murata.com/products/capacitor/design/faq/mlcc/quality/29.html I noticed that both failures I had (on two separate boards) were on caps ne= ar a hole, and that on those 2 boards I had tightened that screw to hell, e= vidently causing serious mechanical stress. I remember I was mean also while soldering.. I'd like to ask You this: given that the boards work although soldering was not like I will do it in = the future, would You recommend anyway to desolder those MLCC's and resolde= r (with care this time) new ones, although the ones I'm gonna remove work a= nyway? i.e. did they become more fragile for the future, or is it a "either die or= survive" thing and as they survived my soldering and the tests perform OK = then I can assume that if they haven't died during soldering they won't eve= n now, providing I don't stress them mechanically? Thanks! Mario PS: As a final note, what I learnt from this lesson is that MLCC capacitors= (not resistors or other components) do become indeed short given sufficien= t board bend or thermal stress. It's a reality and you should follow the gu= idelines you find on some of the links above, i.e. use as small as possible= SMD sizes, keep the MLCC's parallel to the board edges and as far as possi= ble from connectors and holes, reduce thermal stress during soldering as mu= ch as possible, don't use too much soldering on pads. Silicone won't help, because the mechanical stress reaches the capacitor fr= om the soldered pads. Using less soldering will help, as said, and also kee= ping the capacitor well down on the PCB (not lifted by the solder on its pa= ds). I think that with the due care MLCC's can be considered reliable, but = one must always keep in mind that if they crack they will fail short, so be= careful about high currents, and about what other components are placed in= their surronding, which may then catch fire, etc.. Lead-free soldering causes less stress than Sn-Pb soldering, because it's s= ignificantly stiffer than the "older" soldering [ see http://www.dfrsolutio= ns.com/pdfs/2006_Cracking_Pb-free.pdf ]. My application is Sn-Pb (automotiv= e, RoHS exempt) and thus is "weaker" (for once!). Through hole multilayer ceramic capacitors, if one has the space to use the= m, are mechanically much more reliable, not suffering from stress coming fr= om the PCB flexure. --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .