How exactly do PAL chips get programmed?

Teo Zenios wrote: <excellent replies snipped> If you have a chip which you doubt is functional, it may still read out it's fuse map. The fuse map of say a 22L10 or such may be perfectly readable even if the device is not playing in the circuit. One thing that went bad on them was the capability to fully drive the loads in circuit to ttl valid regions. IIRC this would be < .7 for low, or > 2.2 for high. (for the sake of argument). When you pull the part and put it in a programmer, the loading is much more friendly, and the programmer may well pull a copy of the fuse map unless the security fuse(s) are blown as mentioned by tony D. For historys sake, the PAL's were programmed originally with a tool called PALASM, and later by lots of other tools. The fact that the security fuse could be blown meant that you had to have a way to determine if your device was functional. This was supplied in a proper environment by supplying a "model" which would flip the lines according to function of the device and test it, after the device was programmed, to show that it was good. Most engineers I ever worked with, and in the organizations I saw, few things were ever done but to save the PALASM source, (or whatever source) and the binary fusemap. If the engineer ever bothered to use the model and simulation features, and specify that, I never saw it. They usually would blow a pal, test it in the circuit, and when it worked, save the source and ship the fusemap out as the only thing to manufacture by. Anyway to connect back up, what is malfunctioning, and why do you think the pal's are bad? I didnt see a reason why that is suspected, or how you could have found that out. I must assume you have a schematic to have figured that out. Jim