The parts were actually stored in my basement for almost all of the 9 years. The supposed "new" old parts were in an anti static tube in the bottom of a huge box full of DIP chip tubes. The average year around temp of the storage area has been around 65F. I would say that in summer the relative humidity is quite HIGH and in winter very LOW.

In all of my experience, parts of this type were always leaving the factories in an erased "ready-to-program" state. If Cypress had left a test pattern in the parts then I would have expected to see the same pattern in multiple parts. Another interesting thing to note is that on the Bytek programmer when I read the parts into the edit buffer the binary fuse data showed as mostly 1's with a few 0's sprinkled in. If I erase the part and then perform the read and look in the fuse map edit buffer it shows as all 0's.

The customer is apparently online again now with the replacement parts I sent to them via FEDEX. They called today and are now asking for some additional spare parts made with new chips. (Maybe I'll supply the TI fuse link style Erik mentioned).

In the same big box full of chip tubes there are also some tubes of AMD PALCE16V8's. After this whole experience with the Cypress parts for this client I want to look to see what is in these empty AMD parts. (I'll report back here with what I find out). I suspect that the AMD parts are going to be perfect. I have several board designs that were done for clients that used AMD PALCExxVy type GALs. One of these boards has been in continuous production since 1992 and has never ever shown problems of GALs dying.

Michael Karas