Help me to revive this dead ISA floppy disk and serial controller card

[jscipione]

I have had a catastrophic failure on my ISA Floppy Disk and Serial Controller card in my micro 8088 machine.

Something went very wrong with the card erasing every ROM on my system as it went out in flames. I've managed to restore the rest of the ROMs on the rest of my system but traces on the floppy controller PCB have burned up, see attached images. Since there's not too many components and the traces are as damaged as they are I've ordered a new PCB, desoldered the components from the old PCB and re-soldered some passives to the new one. In the process I've discovered at least one 10uF capacitor and one 74LS138 have gone bad. Given the damage shown, I'm assuming that the main PC8477bv-1 floppy controller chip must also be bad but I have no way of testing it. Any ideas here?

 

[Chuck(G)]

Only way that I know of testing a suspect controller chip is to try to use it. I don't understand what happened to your card, however. A single-point failure shouldn't cause that much collateral damage. It almost looks as if you had 12 volts on your 5v power rail.

 

[sergey]

That damage looks pretty bad. It appears that 5V was short circuited on the far side near the floppy connector.
Are you sure you didn't install JP1 jumper with a standard non-PS/2 floppy drive?
By PS/2 here I mean an actual floppy drive that came from a PS/2, that uses pin 3 on the floppy interface connector to supply +5V to the floppy drive.
That shouldn't have damaged the components though...

 

[jscipione]

No I did not install JP1 and I don't have a PS/2 floppy drive just regular old 3.5" and 5.25" drives. The two of you (ChuckG and Sergey) have identified a short on 5V as being the culprit, and I believe you. My best guess is that the short is inside the PC8544bv-1 chip given the damaged traces surround the IC. I'm a bit reluctant to try again with that chip I have and I'm looking for a replacement. All I know is that my system stopped booting one day with the floppy drive lights coming on solid. I flipped the floppy cable over a couple times thinking it must have been installed upsidown but that didn't help. That's when I noticed the burnt out traces on the PCB. Every single ROM chip on my system got erased, my BIOS ROM, microcontroller that controls the PS/2 port, UMB GAL is burnt out. I reprogrammed the chips and thankfully that got me booting again. You can see my current system in my profile pic. I plugged the floppy controller in one more time briefly and the system would not boot, there's definitely a short somewhere on the board but no other damage thankfully. C12 cap in the upper left was definitely shorted, you can see I replaced it with a blue cap on the new PCB, but I'm guessing that's not the main issue.

 

[jscipione]

I finally got the floppy/serial controller rebuilt and put a new (real NS) PC8544bv-1 chip in it. I made sure not to install JP1. On the bright side plugging the card in no longer results in burnt out traces. Unfortunately, plugging the floppy controller in seems to have killed my micro-8088. No visible damage to it but it won't boot any more, no startup jingle, no codes on the rom code analyzer. This is a similar symptom to what I experienced before rebuilding the floppy controller. However, I was able to get things working again by reprogramming the BIOS ROM last time, no luck there this time. I suspect that the AT29C010A ROM chip I am using has gone completely bad. I tried reprogramming it several times, putting it back in the board and booting (it fails) then when I verify the chip on my TL866II+ it fails to verify. Perhaps because the micro-8088 is writing BIOS settings to the ROM but I doubt that because it doesn't appear to be executing code in the first place. I have ordered some new AT29C010A ROM chips. Any other suggestions in the meantime? I tried a different 8088 cpu, no dice.

 

[sergey]

A few questions and comments:
Did you test your FDC in another system? Or you are assuming it is good since you don't see any new burned traces?
When you reprogram your AT29C010A BIOS ROM, does it verify successfully in the programmer? (And then it fails to verify once you put it in Micro 8088 and try to boot it?). If it does... most likely your Flash ROM is fine. But it is highly suspicious that it gets corrupted... There is a very specific sequence that is required to erase and write the Flash ROM... The other option would be high voltage... maybe?!
8088 BIOS will reprogram the Flash ROM to save settings only when told to do so from the setup utility. It will not try to reprogram the ROM by itself. If the system doesn't print any POST codes, it doesn't run the BIOS...
Can you verify that you're programming the 128 KiB ROM image, and not the 32 KiB one (that is intended for upgrades using xiflash utility)?
Your RAM chips look suspicious to me... They say Alliance AS6C4008, and the date code is week 38 of 2017. But they totally look relabeled. The package looks like older Hitachi packages... The logo is off... One of the chips has a nick. I'd actually recommend getting SST39SF010A flash ROM and original AS6C4008 from Mouser or DigiKey...
But before that, I don't think we know why your system failed in the first place. Looks like some short circuit or wrong power supply voltage, polarity, etc. Did you verify that your power supply is good? An OK test, is to connect an HDD (preferably one that you don't mind damaging), turn on the PSU and measure all the voltages.
Also, before trying the system with the FDC, I'd try running bare configuration - just the CPU board, and check for the boot melody.


Older Hitachi SRAM:


Newer Alliance SRAMs - Original ICs ordered from Mouser (on my board):