8-bit acorn hardware • Re: WE BBC EPROM Programmer not copying all data

EDIT: If we can read a chip OK, then the 8255 is probably sound. If bits were sticking low or high, that would show up in the dump files as the same data bit being wrong at all addresses and/or addresses with a consistent bit pattern always having wrong data.

Hmmm, yesterday it was reading EPROMs ok, today it appears it is not....I'm not sure at what point it stopped reading them correctly. It could have been after swaping the wires on the transformer. However in read Pin1 and Pin15 are both maintaining 5v (4.9v). It can read enough to know the title of the rom to display in *roms, but not enough for *help as today it just gives "METIMECTRUETRUENTILUNTILUSRUSR", where as yesterday it was giving the full list and those commands were functioning.

I've tried to find an old image of the ROM I took yesterday to compare against (WYS(OG) in the WYS archive). Address 0100 is different (F0/A6) compared to the image today (WYS1). I've then taken subsequent images which change slightly each time WYS2 compared to WYS1 is different at 0601 (07/12) and WYS3 compared ti WYS2 is different at 0201 (C0/00).

And just to clarify, before you tried programming any of these EPROMs, did you test them to see if they read as blank (all memory locations reading back as 0xFF)? Did you erase them with a UV light box/unit?

Yes it was a rom I'd previously programmed which had been erased using the UV light box and was checked to be blank before I started. The voltage on Pin28 during writing is 4.1v

will generate a test pattern EPROM image. Locations &0000 to &00FF will be programmed with values &00 to &FF respectively, and everything else left unprogrammed. (I've also made it easy to change the filename, and left plenty of room to add more tests.)

Knowing what this should look like and seeing how it reads back might provide some information about which addresses other than the desired one are being programmed.

That's a great idea (although it might now not be reading back correctly). Even so it does something odd...either when reading back the eprom or programming the eprom it is duplicating at 1000.

Statistics: Posted by vela025 — Mon Dec 23, 2024 8:13 pm

---