| ??? 12/21/06 16:23 Read: times |
#129915 - Dangerous extrapolations Responding to: ???'s previous message |
I fully admit that it is dangerous to extrapolate from a small sample. Unfortunately, it is my job to do so.
Look at it this way: if someone bought one of your chips which happened to be somewhere at the upper end of the performance spread, "tested" it, then made a design based on those tests, would you honour your warranty on the chips from the lower end of the performance spread that failed to perform to his "test" results - even though they were still within your datasheet specs...? Would I change a chip to make a customer happy? Yes. And would I honor a warranty for parts within the specs? Most likely yes. "Customers are always right. Even when they are wrong." |
