Dumped a bunch of random BIOS chips from old ISA/PCI cards/mobos- Need advice

[GearTechWolf]

Some ROMs might need a chip enable or output enable to be read too. For example the mask roms in my Compaq Portable and Deskpro both needed me to hook up a jumper wire to read them with a generic profile, because my TL866 didn't include a profile for the exact part.

A good thing to be aware of, I'll have to browse/search some datasheets (if available) and see what I find!

 

[GearTechWolf]

Also... others seem to have a single bit constantly stuck ON, which is apparent when you try view them as bitmaps. In these examples bit 3 is always set, which generates those vertical lines:

By the way, how does one view them as bitmaps? That's a useful diagnostic.

 

[VileR]

'IWS-4000 REL 2-0 EGA U55' (overdump? - 32K, should probably be 16K; the first 16 are junk)

I'll check what size that chip is/should be if I can find a datasheet, maybe they just used an over-large chip and filled the unused space with junk?

Yeah, I suppose that's a possibility. At least in this case all the data seems to be there, so probably nothing to worry about too much.

'64V 3DX BIOS' (too small)
'Cardex1 VGA BIOS' (too small)

Why are those two deemed too small? I'll check datasheets on them as well, if I can.

They look like partial dumps- I've just never seen a VGA video BIOS that's only 8 KB; that's not enough to hold the font bitmap data alone (in fact all 'Cardex1' contains is a part of that).

By the way, how does one view them as bitmaps? That's a useful diagnostic.

I use Crystal Tile 2 (the UI is crap, but you can load up any raw binary file, hit F5 for a bitmap view, and play with the Properties/zoom/etc). Or you can try this one - works well too: https://github.com/bbbradsmith/binxelview.

There's also resman's quick n' dirty method described here: prefixing a small header will turn a raw binary file into a valid .pbm (Portable Bitmap) image, so you can automate that with a batch file/shell script.

 

[GearTechWolf]

One last thing, how do I translate a stuck bit to a particular pin?
Also, a little data on the one chip that seemed like an overdump, it's an AM27C256, a 32K-byte chip, so they did just fill the excess with junk, weird!
More info to come after I attempt re-reads of the 6 chips that didn't come out right before, wish me luck! (the too-small ones will be easy, it's the stuck ones that worry me)

 

[GearTechWolf]

Okay, sanded some mildly corroded pins, dumped using larger generic profiles, and I *think* they're all good now!
I already glanced through them using the buffer-editor and they all verified clean multiple times.
Most of the 'stuck bit' chips just needed half their pins sanded to correct the issue, but then I needed to sand all the pins on one of the 'too small' ones to get a clean re-dump at the larger size!
It's no surprise some needed cleaning, now that I think about it, the ZIF-socket grabs the untouched sides of the legs, not the wide flats that typical double-wipe sockets keep polished and free of oxides! (plus, many of these were salvaged from boards/cards that were out in the weather a bit)
Let me know how they look, please!

 

[VileR]

All look good now! Didn't figure that the size issue would take sanding work, but all the issues seem to be sorted - thanks for taking care of that.

 

[GearTechWolf]

All look good now! Didn't figure that the size issue would take sanding work, but all the issues seem to be sorted - thanks for taking care of that.

Technically speaking the size issue wasn't fixed by the sanding, it just ended up needing the sanding to read properly after I selected the right profile.
Guess I got lucky on the pins making decent contact the first time around? Anyhow, you're welcome!
Thank you, that's great to hear! I'll add more ROMs as I discover them in my vast collection of various computer-parts and junk.
Got at least one I'll need to de-solder to dump, and I'm not confident I can do so without damage to the board as the through-holes are tiny.
That one might have to wait until I get a job and can afford a cheap hot-air station!

 

[GearTechWolf]

Third big dump! 19 more BIOS images. Had one that wouldn't verify properly, despite manually reading it returning the same checksum every time.
Had another that had to be put in a socket to get a consistently good connection due to corrosion on the pins.

 

[GearTechWolf]

BTW, I know one of the EV348 dumps is bad, the chip was uncovered and seems to have developed some bit-rot.
Thankfully, I had a second card of exactly the same model/version/everything! Just need to scare up my UV-eraser and burn a good copy.

 

[GearTechWolf]

A small dump this time, as I found that I hadn't dumped some chips in a collection that were salvaged via use of a propane-torch.
Just a couple Award BIOSes, and a Quadtel VGA BIOS.
Plus, I also found three tiny 8-pin serial EEPROMs, but I'm not at all sure they were being read properly, unless they really were mostly empty.
Thoughts on the serial EEPROMs?

 

[JohnElliott]

With some advice on using the generic profiles in my EPROM-reader software, I got all the ROMs I failed to read the first round (excepting the DIP40 ones that don't fit, of course) and at least eight more I dug out of my clutter. including a huge old (IBM?) 486 ISA-only board with a full-length memory riser-board, a Venus Virge S3 DX PCI VGA card (with memory-doubling daughter-board installed), a Promise Technologies EIDE PRO ISA card, a massive full-length triple-decker Vermont Microsystems 48KHZ Image Manager 1024 Model 1B1 CGA ISA card with 1MB of video ram (capable of running VersaCAD!), a Willow Peripherals VGA-TV dual-card (one 16-bit EISA, one 8-bit ISA, the BIOS is from the EISA card), a SIIG EIDE Master ISA+I/O CN2424 16-bit ISA card, and an Adaptec AHA-2940 PCI SCSI CARD. Oh, and another one from my dad's work, labeled 'FEED'.
Enjoy perusing their data! (I'll take pictures of whatever cards folks show interest in, also need to archive the 5-1/4 floppies for some of them)

Having just got round to looking at these: Which dump(s) come from the Image Manager 1024?

 

[GearTechWolf]

Having just got round to looking at these: Which dump(s) come from the Image Manager 1024?

The A0100PL U5 V2-21 VM c 1987 and A0100PM U22 V2-21 VM c 1987 dumps in the BIOS-chip dumps-2 ZIP-archive.
That was the labeling on the labels on the chips themselves.

 

[JohnElliott]

The A0100PL U5 V2-21 VM c 1987 and A0100PM U22 V2-21 VM c 1987 dumps in the BIOS-chip dumps-2 ZIP-archive.
That was the labeling on the labels on the chips themselves.

Thanks very much. Looks as if it's 80x86 code, which is helpful if I want to point a disassembler at it. I've also attached dumps of the two fonts I found:

 

[GearTechWolf]

Oh cool, thanks! Looks like same characters in each just different sizes/boldness? Nice wide range of characters.

 

[VileR]

That is the standard IBM PC US/Latin charset (later codified as Code Page 437). The reason I'm always looking for dumps like these is to expand my compilations of PC text-mode fonts (1, 2) - I haven't updated either of those in a while, but your additions will sure be part of the next batch.

Curiously, the IM1024 has an 8x12 charset (the first image in JohnElliot's post) which looks like a condensed version of IBM's default EGA/VGA font (8x14) - and it's a very close match for the BigBlue Terminal font which I put together a few years ago by doing basically the same thing.

 

[JohnElliott]

The 12x18 font is also found in the Windows 1.x driver for the IM1024 (V1DSK.ZIP at http://cd.textfiles.com/carousel/030A/ ).

 

[GearTechWolf]

Dug through one of my many boxes of 'Misc Electronics' in an effort to label it a little better, found a board from some kind of HP test-equipment, labeled '05005-60003 A3 D series 1928 03L', and containing an Intel P8085A, a D8155, and Intersil 1820-2132, misc logic, and three 40-pin UVEPROMs, all Intel D8755A.
Can't dump those until I upgrade my EP1132 or get something else to dump 40-pin DIPs with, but I also found something else...
The disassembled remains of an old Postal scale made by ASCOM, with a daughter-board containing two ST M27C256B UVEPROMs!
So I dumped those and included a text-document with a bit of additional information in the ZIP-file.
One is mostly-empty but for a small sampling of nonsense and the other has some info (plain-text copied to the text-file) along with being a third or more empty.
These two were a bit of a pain to read, as they were mounted on these plastic quick-swap caddies with their pins bent around to hold them in place.
The Quantum water-purifier controller used the same quick-swap ROM-socket with the removable caddy. Odd system.
More to come as I run across them!

 

[GearTechWolf]

Okay, I've been sitting on this one hoping to have more than just the one ROM for this update, but I haven't gotten around to digging more out of my shed.
So! This is the ROM from what I thought was a weird back-plane-mounted full-length NIC,
but it's actually a Serial-interface card from a KVM, used to control Terminals remotely.
Made by Cybex (later acquired by Black Box) it was used in the 'BLACK BOX KV150A-R2 MULTI SERVSWITCH SYSTEM 156385' KVM.
The cards sold by Black Box even still have the Cybex silkscreen on them! (as seen in the included pictures, grabbed from ebay)
I've also included the manual for that KVM, along with the pictures of the CPU-card, power-supply, Serial-interface card, and the back-plane.
Plus a listing of all the significant chips on the board. Two 40-pin chips that I can't currently dump.
Possibly the Programmable Watchdog Supervisory E2PROM chip could be dumped, but it's not socketed.
62-pin ISA-style card-edge, four RJ-45 ports, and eight config-switches that I couldn't find mention of in the manual.
Not sure if there's anything I could do with this aside from selling it, untested, as-is, or break it down for parts.
For now, I'll preserve it as-is, maybe compare its pin-out to an ordinary ISA slot later down the line, on the off-chance they just copied that for their back-plane.
(the CPU-card uses a PCI-style interface, as can be seen in the pictures, while the power-supply uses pin-headers, odd mix of connectors/standards)
Have fun out there and see you next time!

 

[GearTechWolf]

Five more ROMs for you all!
First, the two ROMs from my mystery WANG terminal board.
Then the ROM from an HP terminal and one of the two ROMs from a Televideo terminal. (other is blank/possibly fried)
Finally, the BIOS from a 486 motherboard, a rather weathered MB-8433/40/50UUC-A Ver:2 that is missing its keyboard controller.
All the chips aside from the WANG ROMs needed their pins sanded to work properly.
Both the HP terminal and the Televideo terminal are missing critical parts, physically removed by force in the case of the HP, a victim of scrappers ripping the heat-sinks off I suspect. I removed all the electrolytic caps from the Televideo years ago, with the intention of replacing them, but it seems to be missing some inductors as well. (two for the HSYNC, at least)
So all three boards are difficult-to-impossible to repair. Still, at least I salvaged the ROMs! (mostly)
That's all for now. Enjoy!

 

[VileR]

and one of the two ROMs from a Televideo terminal. (other is blank/possibly fried)

Interesting... what Televideo terminal is that? Did it have MS-DOS support? I see that the full Code Page 437 is there, along with the native(?) and ISO charsets. Sub-0x20 graphical symbols and all: