| ??? 02/18/04 18:21 Read: times |
#65042 - RE: good atmel & bad atmel Responding to: ???'s previous message |
Dear Juergen,
you think the probability of PCB fault is greater that microcontroller fault(correct me if I misunderstood). well, I'm quit sensitive about PCB too. Actually I haven’t used any SMD, all of parts are through hole ( I think they are more reliable in this way), PCB has just top & button layers and soldering is done by hand. I used lot of sockets, first I thought about fixing parts to sockets with paraffin but I don’t think it is necessary because sockets are military grade).about checking strategies first I’d planned to check each circuit with a Debug version of microcontroller that checks all functionalities and report errors via a test point to computer serial port then I would replace the Debug version with release version ( If you see I’m going too far please stop me! I’m not sure how much necessary it is), second check method is visual using a magnifier(one by one!). you said about a thermal test, you meant this test should be done for each individual device didn’t you ? what about voltage overshoot noise, ripples,… do you think they are necessary too( for individual device or for the design). The device has stepper motors that as you correctly said can cause vibration but I don’t have necessary equipments to conduct shaking stress! But what about micro, shouldn’t I be worry about it ? you said It could fail, but do you think the risk of failure over 10 years is considerable, can I do anything to make sure, I especially like to use this chip because it snugly fit my needs, and it is cheap if the failure rate is not considerable, can I have any rough estimate about failure rate? Thank you very much for your attention. |



