Wanted: another 12C for reverse-engineering (two-chip, non-diaper)

[brouhaha]

I'm looking for another 12C to purchase for reverse-engineering to support making Nonpareil's Voyager hardware simulation more accurate. Mark Shin was recently kind enough to gave me a 3210B serial prefix unit for that purpose. That's one of the SST units (described in Chuck McCord's presentations at the 2022 HHC and HPCC conferences), and although I'll attempt to find the test mode that allows access to the bus, I'm not there yet.

In the mean time, I'd like to get a two-chip 12C, but the single-rigid-PCB model rather than the original version with the polyimide flex circuit for the chips and LCD. The early version had the electronics wrapped in a black plastic sandwich ESD shield, sometimes called the black diaper. While that version is electrically fine for what I need to do, it's much harder to solder wires to the appropriate signals, and I'm not sure that I can cut the ISA line between the Nut and the R2D2, which I plan to do to inject instructions. (A microcontroller will control a 74HC4066 transmission gate to disconnect the ISA line when injecting instructions at addresses the R2D2 would respond to.)

I think the single-rigid-PCB non-SST model was made from 1982 through at least 1988.

I need a unit where the keyboard works fairly reliably. I bought one in the right serial range on eBay, but the keyboard is very unreliable, which appears to be at least in part due to many of the heat stakes popping off the board. In general it's hard to find what I'm looking for on eBay because most 12C sales on eBay are either the much newer versions using one or two CR2032 coin cells, or the photos don't clearly show the serial number.

Aside from the keyboard and display working correctly, and passing the self tests, it's OK if the unit is somewhat beat up. I don't want to spend a fortune to buy a pristine unit.

So if you have such a unit you'd be willing to sell me, please DM me with your asking price.

(Other Voyagers models of that same construction would be welcome, but those are rare and valuable, so quite possibly out of my price range.)

Thanks!
Eric S.

[Mark H. Shin]

Quote: the right serial range

If you would save me the trouble of researching the proper serial range, I have two USA models:

2540A33867
2908A08457

Either of which I am willing to donate to the cause...

Edit: After reviewing Chuck McCord's Powerpoint slide, it appears S/N 2540A33867 (1985 week 40) would fit the bill, if I'm [not] mistaken...

[mfleming]

Older 12C's were selling at throwaway prices a few years ago. What serial number range (and country) is appropriate?

[brouhaha]

(11-16-2022 12:24 PM)Mark H. Shin Wrote:  If you would save me the trouble of researching the proper serial range, I have two USA models:
Edit: After reviewing Chuck McCord's Powerpoint slide, it appears S/N 2540A33867 (1985 week 40) would fit the bill, if I'm mistaken...

I think you're correct. 2540 should be the two-chip, single-PCB non-diaper model. I beleive that model first appeared some time in late 1982 or early 1983, but I'm not completely ceertain.

It's very unclear when the switch to SST actually happened in production, but it seems reasonably likely that a 2908 (early 1989) might be an SST.

Unfortunately when I was dumping Voyager ROMs years ago I didn't keep good records of the ones I opened, but it was a fairly small number of samples.

Thank you!

[Mark H. Shin]

Just out of curiosity , I went ahead and opened it up...



U.S.A. S/N: 2540A33857
1LH1-0304 ROM
1LM2-0001 CPU

It's on its way...

[brouhaha]

(11-16-2022 12:27 PM)mfleming Wrote:  Older 12C's were selling at throwaway prices a few years ago. What serial number range (and country) is appropriate?

I don't know about "throwaway prices", but there certainly are many on eBay that are fairly inexpensive. It's just hard to tell which ones might be the construction I need. I bought three for about $20 each, and while one turned out to be the right model, it also has a very flaky keyboard.

The serial range (year/week) that has the single-PCB two-chip construction might vary by manufacturing site, but I don't know that for sure. It's likely that USA production switched over before other sites. My general guess is that units made between late 1982 and 1988 (serial prefixes 22xx through 28xx) may be that construction.

Since Mark has generously offered to send me yet another 12C, this time S/N 2540, I can't really claim that I need any more units, though I also wouldn't turn them down if offered since sometimes in the experiments it has been helpful to have extra units.

At some time in the future, after I've made progress with the SST 12C, I'll want to obtain an SST 15C. The 15C is different than the other Voyagers in using (originally) two R2D2 chips. That meant that all of the 15Cs had three chips, until the SST was used, and then the 15C had two chips. I'll eventually want to experiment with one of those.

[brouhaha]

(11-16-2022 05:15 PM)Mark H. Shin Wrote:  Just out of curiosity , I went ahead and opened it up...

Thanks for sharing the photo. That's definitely the right thing.

I believe the 1LM2 CPU and 1LH1 R2D2 to be cost reductions from the 1LF5 CPU and 1LE2 R2D2 in the earliest Voyager calculators. This was probably migration from an older to a newer CMOS process with finer geometry. I haven't seen any official information, though. The progression seems to have been:

1LF5 CPU, 1LE2 R2D2
1LM2 CPU, 1LH1 R2D2
1LQ9 SST
1RR2 SST
2AF1 SST

The 15C always has an additional R2D2 in a square QFP with the LCD driver not bonded out. I've only seen 1LH1-0302 for that, but it is possible that there may have been a version with the 1LE2 (suffix unknown).

I think the 1RR2 or 2AF1 were only used in the 12C, after the other Voyager models were discontinued.