A ROM constructed using logic will suffer from the same statistical reliability issues as the components used in constructing it. The net reliability of a ROM constructed in this way will be lower than that of the same ROM implemented as a bipolar PROM, simply because of the summation of all the probabilities. Bipolar PROMs are subject to damage in programming/verification, but, if not abused, say, by reversing the power polarity, or the like, once properly installed in a functional system, they are as reliable as any similarly complex TTL component.

What Mike hasn't yet realized is that he's building a much less flexible and much more failure-prone ROM than he'd have if he were to use an EPROM, FLASH, EEPROM, or a bipolar PROM. What's more, he's completely overlooked the benefit of using a battery-backed RAM that's downloaded from his parallel port. The hardware for the latter would be much simpler than his hard-wired ROM, use fewer components, and provide much more flexibility. Moreover, it would be considerably less prone to failure. The driver for it would live on the PC, just as would the driver for what he wants to do. Further, because it wouldn't involve the 80C31 (the 80C51 in that mode), the entire system would be simpler.

It's HIS hobby project, however, and I think we should let him do what he wants, however much we may disagree with his approach.

RE