• Please review our updated Terms and Rules here

Help with what must be a very rare S-100 machine

S

sambrown

New Member
Joined
May 17, 2010
Messages
3
A new old computer has floated not my shop - a 1982 era-ish Archives machine, built in Davenport Iowa. Fans of Byte magazine might recognize the name from Jerry Pournelle's column, where he constantly praised the Archives keyboard and used it on his principle CPM machine. It is complete with CPM 2.2, all the manuals and software (Wordstar and Datastar) and other than some minor cosmetic issues and a slightly damaged keyboard is in good shape - Except for the real problem - the keyboard connector was cut off of the caable right at the db-25 connector. There are five conductors in the coil cord, four of which appear to have been used in pins 2,3,24, and 25 of the connector. I am pretty confident of one of the connections, but not so sure of the rest. The manual has pinouts for everything except for the keyboard connector.

Does anybody have any information about this really beautiful little computer?
 
Is it specifically the keyboard pinout you're looking for? The pins on the DB25 look like those used for regular RS-232, although if that's what it uses it looks like you're missing the signal ground. If you open the keyboard up and look where each of the wires goes, I'm sure we can figure it out.
 
If it was me wiring a 25D the 2&3 would be transmit & recieve data (no idea which way round though) and 24 & 25 would be power to the keyboard (12v?? + & - ??) but again no idea which way, and is ground on pin 1?. As cosam said you need to open up the keyboard, take some photos, and also trace the connections both ways back to a pin on a chip if possible, that way you/we know which the inputs & outputs are.
 
I'm thinking power on 25 too (unused for RS-232). This would be easy to check at the computer end. I'd have thought 7 for ground though, and 24 might even be a clock. But yeah, it could be anything, so this is all just speculation ;-) Let's see what the guts of this thing look like...
 
I am pretty sure for forensic reasons (evidence of a previous repair attempt) that +12 (red wire) is on pin 24, and logic would suggect that ground is on pin 25, with corresponding red and black wires in the coil cord and that would be easy to check with a voltmeter. If that works out okay I could power the keyboard and see which pin produces serial strings, but thats only half the answer, so as suggested I need to trace it back to the chip somewhere. Would an 8250 be the likely device for early 1980's? There 3 wires left in the coil cord, but one appears to have never been connected (no fatigue failure at the cable ending), which takes us to a 50/50 thing. I'll post some photos after I take it apart tonight. Given the era it seems likely that this was a serial based device.

Also - I have read some horror stories about powering up old systems with dried out electrolytics and old tantalum capacitors. This box is literally filled with electrolytics. Seriously. Never saw anything like it. Big ones, little ones, odd shaped ones.. Should these all be replaced?

cosam said:
I'm thinking power on 25 too (unused for RS-232). This would be easy to check at the computer end. I'd have thought 7 for ground though, and 24 might even be a clock. But yeah, it could be anything, so this is all just speculation ;-) Let's see what the guts of this thing look like...
Click to expand...
 
sambrown said:
I am pretty sure for forensic reasons (evidence of a previous repair attempt) that +12 (red wire) is on pin 24, and logic would suggect that ground is on pin 25, with corresponding red and black wires in the coil cord and that would be easy to check with a voltmeter.
Click to expand...
Yeah, logic can often be subjective ;-) I have come across at least one device in which ground was a red wire and +6V was black! Get a volt meter on there to be sure.

If that works out okay I could power the keyboard and see which pin produces serial strings, but thats only half the answer, so as suggested I need to trace it back to the chip somewhere. Would an 8250 be the likely device for early 1980's? There 3 wires left in the coil cord, but one appears to have never been connected (no fatigue failure at the cable ending), which takes us to a 50/50 thing. I'll post some photos after I take it apart tonight. Given the era it seems likely that this was a serial based device.
Click to expand...
If you have something with which to detect the serial data coming out of the keyboard (scope or logic probe should do it) you might save yourself opening up the keyboard. Otherwise we just need to find the datasheet of whatever makes the keyboard tick and see what's what. It could be RS-232 or something else with separate clock and data lines. What the keyboard connects to at the computer end will also give you a clue, of course. I wouldn't worry about the extra wire; it's probably just a matter of the manufacturer using a standard cord which happened to have more conductors than required.

Also - I have read some horror stories about powering up old systems with dried out electrolytics and old tantalum capacitors. This box is literally filled with electrolytics. Seriously. Never saw anything like it. Big ones, little ones, odd shaped ones.. Should these all be replaced?
Click to expand...
That depends a great deal on who you ask ;-) I don't generally worry about the little 'ns on PCBs but it's worth being careful with the power supply if the machine's been dormant for years or of unknown condition. I usually start with a visual check and probe for shorts before powering up either completely disconnected or with a dummy load, if the PSU needs it. If one cap blows, I'll replace it. If more go, or the ripple is all out of whack, I would consider replacing the whole lot just for peace of mind. If I was working on something rare or with those huge, beer-can-sized electrolytics, I'd be more cautious.
 
You must log in or register to reply here.
Back
Top