| ??? 02/12/08 11:46 Modified: 02/12/08 11:48 Read: times |
#150640 - I just cant believe Andy replies for Malund? Responding to: ???'s previous message |
Andy Neil said:
Ap Charles said:
Mean Time Between Failures has nothing to do with design Oh yes it does! If the design overstresses the parts, that will lower the MTBF - both in the sense of Mean Time Between Failures, and Mean Time Before Failure. Though not a big deal what I want to say is there are certain discrepancies that we dont know and cannot gaurantee anything . True: it's all about statistics, and statistics can never predict individual events - statistics can only ever refer to a population as a whole. MTBF=50000hr does not guarantee that no failures will occur before 50000h - just that the number will be acceptably low. do you just ridiculously design and throw away it in the market? I think you were the one "ridiculously" designing by putting in buffers when not needed. Remember: adding additional unnecessary components also reduces MTBF! as stated earlier by Neil , It does cause wear . That sows the purpose of wear-leveling algorithm You are quoting out of context: that's not the general running of the software that I was talking about. The thing that causes wear in flash memories is erasing, and that issue was separately addressed in that thread. though not with the port pins Which is what I was specifically referring to but with the memory Not any memory - just (E)EPROM (including Flash EEPROM), and just when erasing. This reply is in context to this: http://www.8052.com/forumchat/read.phtml?id=150602 |
