XT-IDE toasted - how to fix?

[kishy]

I've debated posting this because of "egg-on-face", but I figure nobody else is more qualified to answer my question:

I did an incredibly stupid thing in my haste to accomplish something, and put my XT-IDE (ver 1, from the very beginning way back when) into an ISA slot backwards and powered on the machine. The mounting plate wasn't attached and the motherboard wasn't in a case, so this was not as hard to do as one might expect. Yes, very stupid. No, the system did not POST, in case you were wondering.

The card is now dead, unsurprisingly (motherboard is fine though, go figure). It still seems to "take" (for lack of better understanding) the memory address it is configured for, as certain BIOS complain about a ROM checksum error at the address the card is configured for (and those errors will follow configuration changes on the card).

Following my error, I visually inspected the card and found no components or traces were burnt. No "magic smoke" was visibly released, nor did I smell any. I felt each component, and only one had become warm: the EEPROM itself.

I tried to use the flashing utility to put firmware back on the chip (long shot, I figured) and it complained about the chip not responding in the right way. I don't recall the exact phrasing but it was something to that effect.

What are my chances that it's only the EEPROM that got toasted? I imagine someone who knows the design of this card pretty well might be able to comment to that effect.

Additionally, if it was just the EEPROM, if I install a new one, can the flashing utility write that chip in the card given that the new chip will not already contain a previous version of the XT-IDE BIOS?

 

[glitch]

Someone else here did the same thing with a card, and IIRC was lucky enough to only have the EEPROM fried. Hopefully it's socketed (if not, it really should be!) so try replacing that first -- you'll likely have to replace it anyway.

Yes, the config tools will write blank EEPROMs.

 

[konc]

I got away with only a damaged EEPROM as well.

 

[kishy]

Someone else here did the same thing with a card, and IIRC was lucky enough to only have the EEPROM fried. Hopefully it's socketed (if not, it really should be!) so try replacing that first -- you'll likely have to replace it anyway.

Yes, the config tools will write blank EEPROMs.

It is socketed. Big thanks to whoever assembled it, off the top of my head it may have been hargle.

New EEPROM is in the mail, fingers crossed. Great news about being able to program it in-place.

I got away with only a damaged EEPROM as well.

This is reassuring...both that I wasn't the only one to make this error, as well as that the fix was just the EEPROM.

Seems searching would have found these references from the past, but I didn't expect anyone had made the same error

 

[Malc]

I did exactly the same thing with one of my cards a few months ago, As soon as i flicked the switch i realized what i'd done but it was too late, The eeprom was toast, A new eeprom and it was as good as new again.

In my own defence m'lud i would just like to say i was watching TV at the same time as messing with my computer and this scantily clad lady appeared on the screen, It was her that made me loose concentration and i inadvertently plugged the card in the wrong way round Resulting in the untimely death of the eeprom, So it really wasn't my fault. :lol:

 

[kishy]

Almost as quickly as it was toasted, it has been un-toasted. Count me as yet another who seems to have gotten away with just replacing the EEPROM

Relevant photos here: https://imgur.com/a/CdMrv
No drive was attached so the output as displayed is a working result.

 

[glitch]

Lucky break, indeed! Wonder if the EEPROMs are acting as a sacrificial element and protecting the rest of the TTL or what.