pdp-15 XVM (1)

Serial Number ????

The pdp-15 was an 18-bit machine built in the early 1970s as a successor to the pdp-9 and pdp-7, themselves direct descendents, in the 18-bit line, of the pdp-1, DECs first system. They were typically used in data acquisition and analysis, engineering, early graphics systems, and experiment and process control. Towards the end of their life cycle, they were repackaged as 'XVM' systems, with a much modified and larger memory subsystem, and a pdp-11 front end. These were DECs last 18-bit systems, and were substantially replaced by the 16-bit pdp-11.

The historic importance of DEC 18-bit systems cannot be underestimated: the pdp-1 got DEC into the computer business, and was a highly significant machine in the development of hacker culture, at MIT in particular - the first computer game, Spacewar, was written on the pdp-1. The followup pdp-7, also 18-bit, was the machine on which Unix was developed and first ran.

This particular system, the first pdp-15 I collected, had been used at the Ministry of Defence missile test range at Aberporth, west Wales, until around 1990. It was used for the collection and analysis of radar returns from missile test firings, as far as I can gather (a certain amount of secrecy surrounded it!). It's one of the few machines I actually paid money for - it was much easier, from a paperwork point of view, to pay the MoD UKP100 (US$150) for the machine as 'scrap' than it was to have it donated free of charge!

When I got hold of it, in 1992, it had been rather roughly dismantled, with some cables cut, and the pdp-11/10 front end had been removed for use as spares, together with some other pdp-11 compatible parts, such as power supplies and regulators. What I got was a complete but partially dismantled CPU/memory box (centre in picture), an FPU & TC59 (magtape control) (left), and an RP15 disk controller (right). I also got some very poor condition disk drives (RPR02) and tape decks (TU45 - 9-track & TU10 - 7-track) - but these last parts were lost along with the I/O cab for my KL10 when a storage unit I used was unexpectedly cleared. I don't count it a great loss as they had deteriorated significantly before I got them, and weren't good for much. That said, where will I ever find some more?!

Close-up of the RP15 - disk controller for the RPR02

The lower half of the front of this cab is taken up with the TC59 magtape control, the rear is the FP15 floating point unit. Immediately below the console is the BA11 which houses the XVM memory system.

Closeup of the TC59 - module side. The pairs of heavy black cables are the standard cables used for options on the pdp-15.

Closeup of the FP15.

The memory expansion box. Base memory for the pdp-15 in 'XVM' configuration lives in the 'XVM' box - a BA11 (11/35, 11/04, 11/34 use the same box) which slides in below the pdp-15 CPU. This contains the XVM hardware and the first two stacks of core. Additional core lives in an expansion box inside the rear door.

The current status of the system is 'incomplete, not really working' - I've rebuilt it to the point where I can power it on, but I can't really do anything with the console - I think it's because I don't have all the memory back in yet, and the remainder isn't correctly configured for the reduced amount. The console itself gives trouble, the lights keep going out! Something to do with the power to it... hard to troubleshoot as it runs off a multiplexed bus with rotary switches to select which data/register/bus it's looking at, and I'm not sure that's working. A very odd machine, more buses than Stagecoach!

I have a complete set of manuals and prints for this system, but no software - coming from a defence research establishment, it was a case of 'you MUST be joking!' when I innocently asked if there were any magtapes or disk packs I could have...!. The front end pdp-11/05 is missing, and I'd probably like to put the system back to a -pre XVM pdp-15 configuration, if I could find the original pdp-15 memory and reverse the vast changes list to the wire-wrap backplane... I could easily get a pdp-11/10 and put a front-end back in, but I think there are some hard-to-find bits missing from the -15 to -11 interface.

There are very very few pdp-15 systems still in existence... I know of four:
1. This one!
2. My other one!
3. There was one in the offsite storage facility of a Scottish museum - it was part of a collection of medical equipment, hooked-up to a scanner.
4. Rob Garners machine in the USA.
5. Max Burnet has one in Australia.
6. Errrrrr I'm sure there are others, but that's all I know of at present. Any other pdp-15 owners, please get in touch!

http://simh.trailing-edge.com/docs/advmonsys.pdf Bob Supnik has written a fascinating document on restoring pdp-15 software (pdf format)
http://simh.trailing-edge.com/docs/architecture18b.pdf Bob has also written a document on the evolution of 18-bit systems (pdf format)
http://www.unusual.on.ca/unusual/pdp15.jpg Kevin Stumpf had one; it's now in the capable hands of Robert Garner.
http://www.acms.org.au/burnet-max.htm Max Burnet has one in Australia.