Bought a CQ4x4 and 054 adapter to program 27C1024s. Looking at waveforms on the adapter showed significant cross talk, voltage spikes noise and ground bounce. Bypass caps and hard wiring VSS on the socket to true ground got me to the point where I can program a full part. Or so I thought.
The program cycle writes a byte and reads it then moves onto the next byte. Even if I can make to the end of the buffer without error multiple program cycles are not always successful.
Sadly I am unable to verify any proms. Regardless of what is programmed I should be able to read the part into the buffer and verify it against itself. It fails consistently. I can find the error in the buffer and re read the device. The bye in question will switch from correct to incorrect and so on with sequential reads.
I examined the read timing waveforms on a scope. You can see glitches ever so often in data when the input to the bi-directional buffer begins to change ahead OE\ going high. There is nothing synchronous about how the read timing is implemented. The CE and OE on the 27C1024 are low the entire time and the address just scrolls past. There is no clock to qualify the data as is pours out of the prom. I can see why the read data in very inconsistent.
These problems are serious enough for me to give up on saying that the CQ-4x4 can actually program a 16 bit EEPROM. I found numerous problems and things such as software scripts that do absolutely nothing but not being able to read a part into the buffer reliably put the nail in the coffin for me.
I've gone out and order some OTPs and perhaps newer parts but don't expect this to make a difference.