1. Home
  2. Embedded Programming
  3. Mid '80s uP from Western Electric/AT&T -- WE212

Mid '80s uP from Western Electric/AT&T -- WE212

I have recently revived my interest in microprocessors (goes all the way back to an Altair 8800 kit I purchased and built in '76). Having a general interest in electronics starting in my childhood, but also having a lot of experience in software from my career, I feel this group might be one I'd like to get to know better (I gather there are a few of you out there who are also in their (f-mumble)'s, age-wise :-).

Anyway, I recently acquired an old communications rack produced by AT&T (containing 8 9600 bps DSUs, I believe, just sitting in the warehouse gathering a thick layer of dust), and from the ICs it appears to be of an early-to-mid-80s vintage, prime time for early

8-bit microprocessors. And, to my delight, after examining the boards, I see lots of goodies such as 8255 PIOs and 8253 SIOs (going by memory, might have been 8251s, whatever Intel called them). And, the cards have other goodies such as nice little bat switches and 4-character *smart* LED displays (the 16-segment ones that can display the 64-character upper-case ASCII subset). And there are a typical mixture of 74xx series chips to glue it all together. But, since this is WE gear, a lot of the parts have internal Western Electric part numbers of the style "WEnnn". Some of the parts have both WE and "generic" numbers, which helps a lot, but the 40-pin uP chips (assumed because of their wiring to the 8255s, ROMs, etc) have only the WE part number "WE212". I have found exactly *one* reference to this CPU on Google
formatting link
. I was hoping it was an 8085 or a Z80 or friends based on the "common" I/O chips used, but my preliminary wire tracing from the CPU to other "known" chips on the board doesn't reveal a match (yet).

I'm hoping one of you out there might be familiar with this chip; there are several per DSU, I think 3, nice ceramic ones, and I'd be inclined to "donate" one or two to a good cause, especially for some pinout or other info that might be available (though this would not be a requirement).

Sorry for the long-winded message -- and pointers to better places to ask will be gratefully received. My interest is hobby-related, not commercial, if that matters, and I'm certaintly open to share what I know, for what it's worth :-)

--Steve

Reply to
Steve Bour
Loading thread data ...

I was at Bell Labs in the 80's and WE had an 8 bit uP called the MAC-8. I don't remember if WE212 was the part number for it but if they are MAC-8s it was an interesting chip that was tuned to run C programs. You might find some info on the web by searching for MAC-8. Performance wise it was not that great, though it was comparable to other 8 bit uPs of the day running at similar clock rate (2 MHz I think).

Dave Rooney

Reply to
Dave Rooney

I don't have any facts for you, but comp.os.cpm and alt.folklore.computers are more likely to hit pay dirt IMO. Thus I have crossposted without snipping.

--
Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net)
   Available for consulting/temporary embedded and systems.
     USE worldnet address!
Reply to
CBFalconer

Great lead, Dave! I just now did a quick search and found another reference to the MAC-8 as being manufactured as a 40-pin DIP and running at 3Mhz (unfortunately the person with this information was also looking for a pinout and other info just like me), and some other references about its use with C. At least I have a working example, so I may contine my reverse engineering project in all my spare time :-) I'm rather fond of C as a programming language, so it *would* be kind of cool having one of the first embedded examples of the use of C, straight from the horse's mouth (Bell Labs/WE/AT&T), no less :-)

Thanks again, Dave, and I'll update this thread as further details are discovered...

--Steve

Reply to
Steve Bour

I picked up a WE MAC-8 eval kit on eBay several years ago which included documentation. I'll pull it out of storage and scan it.

Reply to
Al Kossow

Al: thanks a million! Any interest in a spare chip, assuming it all matches up? :-) Group: if/when Al gets a chance to scan this data, I'll post it on an anon FTP site with access info here, for those others interested.

BTW, I have further supported the theory that what I have *is* probably a MAC-8, based on a few resumes I've found of folks having worked on AT&T embedded projects of that time period. I've also discovered that it was one of the first micros to use memory-mapped registers. I'm thinking I might build a front-panel for one to keep my Altair company...who knows, maybe an S-100 MAC-8 card *for* the Altair...hmm...

--Steve

Reply to
Steve Bour

There are articles about these chips ( the 8 bit CPU, a 4 bit single-chip "optimized for C", the 32 bit CPU that was probably identical to National Semiconductor ) in the Bell Labs Technical Journal and the Bell Labs Record. I can look them up and scan them but it would take a while.

MfG JRD

Reply to
Rafael Deliano

formatting link

Reply to
Al Kossow

The 32-bit CPU was probably the CRISP. Hobbit was a bit later. Neither was much like the National 32k series.

Reply to
Al Kossow

CRISP was about 1986 ( there is a 6 page description in the ASPLOS11 proceedings 1987 ). Hobbit ATT92010 in 1992 ; ATT92020 in 1994 ( there are short descriptions in "Byte"-magazin )

"AT&T Bell Labs Record" April 1984 announces start of production of a WE32100: WE32100 32 bit CPU ; 180k transistors ; 1,75um ; 18 MHz By 1985 the chipset included: WE32101 MMU ; WE32101 Clock ; WE32103 DRAM-Controller ; WE32104 DMA-Controller ; WE32105 system interface unit ; WE32106 MAU "math acceleration unit" = FPU ; In the 1988 IC-Master there are several pages covering the softwarecompatible 1um 24 MHz WE32200 CPU WE32201 MMU ; WE32204 DMA ; WE32206 MAU

Earlier chips mentioned: WE32000 16 Bit CPU 1981 ; ca. 100k transistors ; 2,5um ( the WE32001 was used in the AT&T 5620 DMD Display/Terminal about 1984 ) WE8000 developed in 1976 CMOS microprocessor ; 7,5um later 5um ( in 1981 ) ; 10k transistors ; 0,2 MIPS WE4000 "microcomputer" = single chip controller CMOS ; 0,1 MIPS ; 30k transistors ; There is a 4 page description in the Bell System Technical Journal 1979 p. 959: 1024 - 3840 x 4 ROM ; 80 - 192 x 4 RAM ; 43 instructions Operates on 4, 8, 12, 16 bit data

Names changed over time:

1980 it was MAC-8 ; by 1981 BELLMAC-4, BELLMAC-8, BELLMAC-32

MfG JRD

Reply to
Rafael Deliano

The WE32100 family was used in some of the 3B computer line, 3B2, 3B5. During AT&Ts brief venture into the minicomputer/workstation world.

3B1 was 68K based and 3B20 was bit slice, AMD 29xx I think.

Dave Rooney

Reply to
Dave Rooney

Thanks -- this is great! From a quick glance, it does indeed appear the equipment I have uses MAC-8s (for example, address lines I traced appear to match up).

I appreciate the time you took to scan these. I can even read the schematics!

I'll followup to the list as I progress with this project of mine -- and Al, just let me know if there's something I can do for you...

--Steve

Reply to
Steve Bour

Thanks for the offer...it appears a local university here has a good collection of BLTJs in bound volumes so when I get some time I'll visit it and do the research. I appeciate the pointers, and I've found a few of my own on the web, so that'll probably be a good source of information.

Thanks,

--Steve

Reply to
Steve Bour

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.