In a message dated 04/01/02 19:31:51 GMT Standard Time, [email protected] writes:
> >The trick, I believe, is not to be in that position in the first place. > >Design your products using the latest safety knowledge and test them well > to > >discover if they have any weaknesses you did not address. > > How do you decide what tests to do **for weaknesses you don't suspect**? > Isn't that fundamentally impossible? I actually said "for weaknesses you did not address" not "for weaknesses you did not suspect" - quite a different matter. Mind you, if a designer is not very competent in safety matters there might be quite a number of things that he/she did not suspect, much less address, that he/she should have done. But I was thinking of tests such as bump and vibration, thermal extremes, etc, that reputable companies do to test the reliability of their products. Also safety tests such as simulating faults in components (such as shorting or opening power transistors and capacitors, disconnecting connections to resistors and ICs, etc.). These tests tend to reveal many safety issues that were overlooked in the heat of the design process, or through lack of knowledge of the design staff, and sometimes even reveal things that safety experts familiar with the product type would not have expected. Because such tests can be done and are available from many suppliers (if you don't do them yourself) I understand from UK safety enforcers (Trading Standards) that a manufacturer would have a hard time proving compliance with the LVD or PLD directives if using them would have revealed a safety problem that contributed to an actual safety incident. Regards, Keith Armstrong
