| ??? 07/27/07 15:57 Read: times |
#142443 - Misunderstanding cubed and baked Responding to: ???'s previous message |
Lynn: "if you bake a chip for 1 hour at 100°C (or whichever the maximum temperature given by manufacturer is) as a screening for early failures, it shouldn't affect the reliability afterwards (assuming today's standard silicon processes)". A bake won't provide any screening at all. We bake all PLCCs and QFPs for 24 hours at 125C to remove moisture from the plastic packages prior to dry-pack. The dry-pack specs also allow the use of 150C bakes for shorter periods of time. A burn-in actually runs the chips at an elevated temperature to get rid of infant-mortality problems. Intel used to burn-in 8051 micros (optionally) for 168 hours at 125C. Freescale also does that to their 8-bit micros. I don't know about NXP. That is why I said 1 hour at 100C isn't a problem. |
