• Please review our updated Terms and Rules here
  • From now on we will require that a prefix is set for any items in the sales area. We have created regions and locations for this. We also require that you select a delivery option before posting your listing. This will hopefully help us streamline the things that get listed for sales here and help local people better advertise their items, especially for local only sales. New sales rules are also coming, so stay tuned.

Wanted: bus correction kit for AT&T 6300 (Olivetti M24)

digger

Experienced Member
Joined
Jul 9, 2008
Messages
395
Location
Amsterdam, the Netherlands
Hello everyone,

I'm looking for a so-called "bus correction kit", an after-market hardware hack that fixes an incompatibility in the AT&T 6300 and Olivetti M24 that occurs when using EGA or VGA cards. This product was sold for $45 in the late eighties by W. Allen Associates in Cupertino, Ca. (Can't find any information about this company on the web, only some old references on Usenet.) It consists of a card and a separate disabler chip for the on-board video board.

I'm also looking for an AT&T Bridge Adapter Board (AT&T part #405067281) or an AT&T disabler chip (part #405098625).

If anybody here either has these items or can tell me where to procure them, I would be VERY grateful. :)

Thank you.
 
I think it is much easier to find an AT&T PC 6300 WGS (a 6300 with the endian-bug fixed). Without even trying very hard I found two in the last year, whereas I have NEVER found the hardware endian swap board.
 
So you have been looking for it, then?

Do you have any idea what became of the guy who used to make them? Bill Allen, and his company "W. Allen Associates" in Cupertino, Ca?

Probably long defunct. Who knows if he even lives in Cupertino anymore? If he still lives there today, he probably uses a Mac now. :mrgreen:

I know I could look for a 6300 WGS instead, but my particular itch is to get that one M24, the one I played and tinkered with when I was younger, to finally work with EGA games and support high resolution color graphics.

From what I have understood so far, the 6300/M24 could still take an EGA (compatible) card without the bus correction kit (as long as one manages to disable the indigenous graphics board), and most EGA applications and games would then still work, as long as they'd been programmed neatly. Some stubborn games would cause problems, though.

Any leads are welcome. And user AT&T6300 apparently has (or had) a 6300 model for sale, which contains the kit. As a matter of fact, I found out about the very existance of such a kit through his offering post.

Any leads are welcome.
 
From what I have understood so far, the 6300/M24 could still take an EGA (compatible) card without the bus correction kit (as long as one manages to disable the indigenous graphics board), and most EGA applications and games would then still work, as long as they'd been programmed neatly. Some stubborn games would cause problems, though.

From my experiences with a 6300 WGS this weekend (rescued from work!), this is correct. Half the vga programs (has the AT&T paradise VGA board installed which supports 400-line modes, and a fixed-frequency 480-line monitor) have a good palette, and the other half have screwy palettes and I have to run those in EGA mode.
 
It appears I also still need the 1.43 EPROM images

It appears I also still need the 1.43 EPROM images

Trixter, I hope you can help me with something else as well:

The good news: Dad and I found out that the video board in our M24 is a revision P8, which can be easily disabled through a jumper on the board (jumper W1, for those who wish to know).

The bad news: to my surprise and disappointment, the ROM BIOS revision on our M24 is 1.36, which I found odd, since we had the machine upgraded with a 3.5" 720KB floppy drive. I had assumed that upgrade would always require a ROM BIOS 1.43 upgrade. Apparently not. The 720KB 3.5" drive has always worked and continues to work fine, even with 1.36.

However, I'd still like to upgrade the ROM BIOS to 1.43. I will need someone in this forum to be so kind to dump the contents of *both* 64kilobit EPROMS that together make up ROM BIOS 1.43. Could you perhaps help me with this, Trixter? I'm going to open up a separate thread with this request, since this is separate from my search for the bus correction kit.
 
Okay, here are some photos of the bus correction device.

I've tinkered with the lighting a bit to improve the views. As you can see, this isn't a big deal--just a bipolar PROM and a cable that plugs into a bus converter IC socket. There are some other jumpered places on the bus converter board (see the white wires), but I don't know if they have anything to do with the little correction board.

Since I'm not using another display card, can anyone tell me what the IC on the bus converter board was replaced by the ribbon cable? I'll try to get some time and read the PROM out this weekend.
 
I design custom circuit boards. Cloning this would be a piece of cake, and you could get it fabricated for $20-30 at BatchPCB. Or I could put one in whenever I have some room on a full panel and send it to you for free.

The back side is easy. The front will need some more pictures or probing with a continuity tester.
 
Cool, I'll be following this topic closely, I've been looking for a bus corrector for my 6300 as well.

Since I'm not using another display card, can anyone tell me what the IC on the bus converter board was replaced by the ribbon cable?

It's another bipolar PROM, an 82S123AN on mine. I dunno how to go about telling you the contents of it, though, because I haven't the slightest clue how to read one of those things...
 
I suspected as much. I'll buzz the PCB out and then read the PROM on this thing. I strongly suspect that a GAL will probably be better than another PROM in the adapter--it's getting harder to find blank PROMs and equipment to burn them is getting thin on the ground.

I have no idea what the jumper block does on this--perhaps un-does the mod? Well, we'll find out...

After we get a prototype, I'd like to borrow someone's original 82S123 so I can duplicate it.
 
You'd need a programmer of some sort. A simple microcontroller (like an Arduino) can do it. You could also rig something to read it through a parallel port.

My gut instinct they're just using the old PROM's socket as a convenient place to jumper across the board. The new PROM probably contains the exact same data.

If that's the case, you could actually do the same thing by carefully soldering a couple wires onto your motherboard.
 
You'd need a programmer of some sort. A simple microcontroller (like an Arduino) can do it.

Are there any plans for Arduino-based bipolar PROM programmers? I wasn't aware of any. OTOH, a few minutes with my PAL dumper will get the right result, with no programming at all--just a matter of moving some jumpers.
 
it's getting harder to find blank PROMs and equipment to burn them is getting thin on the ground.

You can easily replace it with a modern EEPROM. Just find something that's level-compatible.

I have no idea what the jumper block does on this--perhaps un-does the mod? Well, we'll find out...

It's switching A8 between VCC and GND. Basically it's swapping the PROM between two banks - probably one that's the original data, and the other one is something custom. My original thought about it being just a replacement is probably wrong in light of that jumper.

Are there any plans for Arduino-based bipolar PROM programmers?

Probably. You'd just need a way to level-shift it up for HV programming, right? But I'm not aware of a specific project to do it.

I was just talking about dumping it. There have to be a dozen implementations of ROM dumpers out there.
 
You can easily replace it with a modern EEPROM. Just find something that's level-compatible.

Uh, yeah--seen any 512x8 45nsec TTL-compatible parallel EEPROMs lately?

My guess is that a lazy engineer simply used a bipolar PROM to implement some combinatorial logic. Back in 1983, when the M24 was designed, PALs were fairly new--and if you look at the old Olivetti schematics, you'll note that HALs are specified where they're used--meaning the factory mask-programmed parts were originally used. I was guilty of the same sin back then.

My guess is that a GAL will accomplish the same thing. Either a 16V8 or a 22V10 should be more than adequate.

Depending on your level of kludge-tolerance, programming bipolar PROMs can be anywhere from trivial to sublime. The big problem is that programming currents are very high , but basically, you select a cell and blast the outputs with +12v pulses, one by one. If you goof, you toss the part in the trash and start over. This can get to be pretty costly very quickly. GALs are nice because they're erasable and very fast and most modern programmers will deal with them.
 
I took a quick look... TTL parallel EEPROM isn't dead yet, but the speed grade kills it. It's a dying breed, I guess.

It'd be easy enough use a current EPROM, like http://www.atmel.com/dyn/resources/prod_documents/doc0014.pdf . $1 each at Mouser. But if you want to reverse engineer the thing and reimplement it in a GAL, be my guest. :)

I'm very kludge-tolerant, but I was just talking about dumping it, which an Arduino can do with ease.
 
No, I was originally searching for EEPROM... parallel, TTL-friendly EEPROM still exists (for now), it's just 45ns that's discontinued.

Which is why I suggested the EPROM. My other hobby wrecks $10-30 parts, so watching a half dozen $1 chips go up in smoke is nothing. :)

Still, the GAL hack is neat. I originally thought you were going to reverse-engineer the state table, but if those things have enough flexibility and flash that you can just copy the whole table in, heck, that's an easy way to get it done.
 
Okay, here's the rest of it. The 82S147 is connected to the 82S123 just the way you might expect--A0 to A0, A1 to A1, and so forth. No tricks.

Attached is a dump of the PROM contents--note that half the PROM is given over to a copyright notice! I've never seen that on a small bipolar PROM--usually space is too dear.

This is a no-brainer for a GAL16V8. I can work the logic equations if you'd like--but I'll leave off the copyright notice, if that's okay.
-------------
Oh, and the copyright notice? Well, look who was treasurer of the AT&T PC 6300 North Bay Users' Group
 

Attachments

  • 82S147.ZIP
    266 bytes · Views: 7
Last edited:
Code:
0000000: f3f3 d3d2 d292 d3db d8dc 9c4c 4c4c 4c4c  ...........LLLLL
0000010: f3f3 d1d0 d494 d444 4444 4444 4444 4444  .......DDDDDDDDD
0000020: f3f3 d3d2 d292 d3db d8dc 9c4c 4c4c 4c4c  ...........LLLLL
0000030: f3f3 d1d0 d494 d444 4444 4444 4444 4444  .......DDDDDDDDD
0000040: f3f3 d3d2 d292 d3db d8dc 9c4c 4c4c 4c4c  ...........LLLLL
0000050: f3f3 d1d0 d494 d444 4444 4444 4444 4444  .......DDDDDDDDD
0000060: f3f3 dbdc 9cd8 d9db d3d2 d292 4242 4242  ............BBBB
0000070: f3f3 d1d0 d494 d444 4444 4444 4444 4444  .......DDDDDDDDD
0000080: 2843 2920 436f 7079 7269 6768 7420 3139  (C) Copyright 19
0000090: 3837 2c20 3139 3838 2020 5269 636b 2046  87, 1988  Rick F
00000a0: 6c65 7463 6865 7220 2020 2041 6c6c 2052  letcher    All R
00000b0: 6967 6874 7320 5265 7365 7276 6564 2e20  ights Reserved.
00000c0: 0000 0000 0000 0000 0000 0000 0000 0000  ................
00000d0: 0000 0000 0000 0000 0000 0000 0000 0000  ................
00000e0: 0000 0000 0000 0000 0000 0000 0000 0000  ................
00000f0: 0000 0000 0000 0000 0000 0000 0000 0000  ................
0000100: f3f3 f3d3 d2d2 92d2 d3db d9d8 dc9c dc4c  ...............L
0000110: f3f3 f3d1 d0d4 94d4 4444 4444 4444 4444  ........DDDDDDDD
0000120: f3f3 f3d3 d2d2 92d2 d3db d9d8 dc9c dc4c  ...............L
0000130: f3f3 f3d1 d0d4 94d4 4444 4444 4444 4444  ........DDDDDDDD
0000140: f3f3 f3d3 d2d2 92d2 d3db d9d8 dc9c dc4c  ...............L
0000150: f3f3 f3d1 d0d4 94d4 4444 4444 4444 4444  ........DDDDDDDD
0000160: f3f3 f3db d8dc 9cdc d8d9 d3d2 d292 d242  ...............B
0000170: f3f3 f3d1 d0d4 94d4 4444 4444 4444 4444  ........DDDDDDDD
0000180: 2843 2920 436f 7079 7269 6768 7420 3139  (C) Copyright 19
0000190: 3837 2c20 3139 3838 2020 5269 636b 2046  87, 1988  Rick F
00001a0: 6c65 7463 6865 7220 2020 2041 6c6c 2052  letcher    All R
00001b0: 6967 6874 7320 5265 7365 7276 6564 2e20  ights Reserved.
00001c0: 0000 0000 0000 0000 0000 0000 0000 0000  ................
00001d0: 0000 0000 0000 0000 0000 0000 0000 0000  ................
00001e0: 0000 0000 0000 0000 0000 0000 0000 0000  ................
00001f0: 0000 0000 0000 0000 0000 0000 0000 0000  ................

Actually, I'm not surprised by the text at all. Note how A7 is hardwired to GND... Half that chip will never get used. I suspect they picked the chip the same way as the Atmel PROM I grabbed above: technology marched on, and bigger ROMs became cheap.

Looks like you're right about it being a logic hack. I can't see the exact logic at first glance, but there are clearly two related patterns there selected by the jumper. I suspect it all makes sense if you sort it to the right endianness. Good call. :)

Yeah, you're it for the GAL. I've never used one. I'm just a hack hobbyist.

I'll draw up a schematic tomorrow if I get a chance.
 
Looking at the schematic of the bus converter, it also appears that CE/ is always enabled. And there are only 7 unique bits in the output (2 bits match one another).

Update

Try as I might, one of the output equations involves 9 terms and I can't simplify it. So it'll have to be either a GAL22V10 or we can go back to the bipolar EPROM. I've got a small pile of TPB18S22s (256x8 ) which ought to do the trick. The GALs seem to be a bit easier to get than the bipolar PROMs--but 5V logic seems to be dying a pretty rapid death in terms of new parts.
 
Last edited:
conv.jpg

Unfortunately I didn't see your message that the 16V8 wouldn't cut it before I did this. Still, let me know if this looks right and it'll be a simple matter to swap in one of the other chips.

edit: fixed a bug.
 
Last edited:
Back
Top