Was your chip a new or used one? Experienced weird behavior when reusing salvaged UVPROMs a couple of times, saw that a few bits seemingly randomly flipped back and forth after programming by doing a few dumps and comparing them by content. No idea if the bit flipping had to do with light conditions in my case, it's an interesting theory though.

I believe it was caused by underexposure when erasing them the first time. The solution was re-erasing them, giving them longer time under the UV lamp, and reprogramming them anyways. Might be worth a try for you too 😀

It's a used one, well I guess I'll try to reprogram it, but the case is that 55ns eproms are hard to get here, so I don't want to make it worse, by damaging my only 55ns capable eprom

Reflashing it didn't help

how did you manage to use the eprom when the dead chip had four enable lines ?

Maybe the light on the chip charges some diode junctions just enough so they get a little boost when switching.

other solution, add a resistor and an led to the eprom window and use it like that.

So, M20 uses 3 out of 4 CS. But these 2 more are for using expansion cards I do not own ( example 8086 processor card). I guess I need to dissasembly that led bulb and desolder one of that SMD LED and indeed put it above that eprom

I only wonder.. Original IC was 55ns, that uveprom is 45ns. So this behaviour is a weird failure of that eprom or what?

Are all the unused pins either grounded or pulled high, dependent on their logic?

Yes, the only NC pins are three CS that PROM had and this EPROM does not. But I'll recheck them.

Indeed one pin wasn't grounded, and now it works fine 🤣

Bernkastel7734 wrote on 2022-11-12, 21:13:

Indeed one pin wasn't grounded, and now it works fine 🤣

well, if we ever meet, you owe me a beer ! 😁

Sure thing. I just wonder what had light to do with it?

Bernkastel7734 wrote on 2022-11-13, 03:49:

Sure thing. I just wonder what had light to do with it?

Not claiming to know the inner workings of EPROMs, but you can probably assume that the photoelectric effect was involved somehow 😀

Yeah, I guess. Anyways I'm glad that it started working normally, and that my M20 is alive.

Bernkastel7734 wrote on 2022-11-13, 03:49:

Sure thing. I just wonder what had light to do with it?

(Bipolar) transistors can not only be turned on by base current, but also by photons. That's why transistors are shipped in non-transparent cases. If a transistor is optimized for photons driving it, and shipped in a transparent case, we call it a "phototransistor", but it basically still is the same PNP or NPN silicon construction as on a regular transistor. So any old NPN or PNP transistor is influenced by light!

In your case, you had an open input. This is a bad thing to have on CMOS chips (the C in 29C1001 is for "CMOS"), because those input randomly read as 0 or 1 or flicker between 0 and 1, maybe influenced by what's going on inside the chip. With enough light irradiation, you likely turned on transistor-like structures (or just the ESD clamping diodes) in a way to provide enough leakage current so the input got pulled down to ground as required.

It makes sense, thanks for explanation.