I wrote this page over the winter of 2016-2017 as I wanted to include pictures along with questions for the knowledgable folk at DECUServe and also to give it wider appeal.
Last update 17/1/17

Previous experience

I have had slight brushes with DEC and its equipment over the years. Perhaps under slightly different circumstances I'd have ended up with them in Reading instead of automotive in Witney. If DEC2 popped into existence tomorrow I'd go knocking on their door and question why later.

At University (note I don't use the word "Uni") though most of my proper computer interaction was through IBM 4341 and 3084 we (engineering) had MicroVAXes running some UNIX . A BSD or Ultrix ? The only definite think I remember was the M shell that had history, something that was pretty unusual at the time.

I had a normal 9-5 job at Lucas Body Systems in Witney. The VAX there (an 11 series I guess) had a wodge of audio cards and associated wiring and was used to develop the talking Maestro instrument pack that everybody loved so much. The term "I.T." didn't exist as I picked up the role of "network manager" for the incoming Netware server that replaced it alongside my official title of "design engineer". I allowed and indeed encouraged a rather good collection of games alongside the development tools on the Netware server A Mitac 386DX. As I remember Mitac had a service arrangement with a household vacuum cleaner manufacturer back them. That's more service than you get these days. We kept the DEC band printer and it was some years before we were allowed to buy a laser (only after a colour dotty matrix). The former VAX room now only contained the 386 but a friend found it useful for respraying his Kawasaki GT550's panels.

By the mid 90s I had thrown in my "safe" job (Lucas Witney closed down a couple of years later anyway) and started contracting returning to Cambridge. Slightly by accident (borrowed a VHDL book whilst working on GSM DSP in Melbourn) I got into "modelling" and that included a DEC-supplied one of the StrongARM for which I had slight interaction with Rich Witek.

A couple of decades on the Centre for Computing History, I had meant to visit in Haverhill, moved rather closer. (For our American friends close is something I can pedal to in twenty minutes and getting in the car is nominally a weekly activity.) An unexplained urge to get back into big computer stuff had me visit the Jim Austin Collection east of York. That place is stunning and houses the university's old 3084 with a book value which I'm sure was quite theoretical in this case, of $8m. A quick mention of Roger Broughton's virtual museum and the very fine Hercules System/370+ emulator.


My only previous brush with volunteering was at the Museum of Technology on the river. It's a great place for steam engines and the like but also some electronics particularly PYE stuff and also a working mercury rectifier.

On my first day as a volunteer at the CCH, a privilege which allows me to wander in to the kitchen virtually unnoticed and make a cup of tea, Bryan suggested I take a look at these three VAXen. He can be seen running the PDPs here.

I have since gathered certain people amuse themselves seeing how "new" volunteers cope with exploding power supplies as it isn't uncommon to get smoke firing things up after decades. In reality I am of of those silicon to soldering iron to switching power supplies to functional programming types and was rather disappointed that all three machines powered up uneventfully. I had just spent time reading the schematics of PDP (not VAX) power supplies too.

Three machines

At the same time as familiarising myself with VAX hardware, the MMJ connector and boot switches and commands I have started on VMS, tinkering with OpenVMS on SIMH and DECUServe. There's a VMS book on the shelf next to the BBC Basic books and of course heaps online. The anal part of me notes useful kernel features present in modern BSDs including Darwin and MS Windows but NOT penguin-based kernels.

The VAXen live in a cosy currently cool corner of the museum next to arcade machines and the PDPs and checking the room is free I wheel them into the 80s classroom to work on them and also grab the amber screen VT220 'cos that's far more fun than shoving a USB serial port into a modern computer isn't it. (The building next door has recently been acquired so this area is becoming less cramped.) Best not mention how long it took to work out how to remove the front panels of these boxes.

Working downwards from eager to maybe dead. For the second machine in particular cards have been removed, contact cleaner applied and various connectors and devices re-seated. Also run with cards removed. The first two machines are Q-Bus I think.

VAX server 4000 ?

The model number is from memory rather than a picture. I'd describe this machine as eager. Unfortunately the disk bays are empty but it does have two working ethernet interfaces and I can network boot OpenBSD off either (NetBSD spewing messages about an interrupt). Not terribly exciting.

Though the connectors made me think SCSI looking up the part number on the termiators says DSSI. Which explains why the mini D-type genders aren't as expected.

VAX server 4500

Its single disk didn't spin let alone seek before we opened and glared at it. Now it tries though there's still no OS and it (the disk) does sound just a little scary. What is the Port A Port B disk interface in the second picture ?

The main problem with this machine is unreliability in during countdown tests. As often as not it locks up at test 17.

As far as I can tell this is sensitive to temperature (better when cold) and also whether the front panel that covers the ribbons between CPU and two memory boards, is present. Yes these have been subjected to contact cleaner.

VAX station 3520

Has an RGB output but the presence of the RGB card surely wouldn't stop something coming out on the MMJ console would it ?

This is the machine that has had the least amount of attention. Does the diagnostic LED count down ? Is there one ? Can't remember.

Unlike the above two I assume Q-BUS machines this one has what Wikipedia suggests is an M-bus. Is that right ? Here is its storage interface.

Is that to SCSI on the left ? I see a 10Mhz bus adaptor device bottom left. But unfortunately because of the different bus to the other machines I still can't see a way to make use of my pile of fully working* SCSI disks in the 400MB-2GB range.

The machine itself has a couple of Fujitsu and one Micropolis drive. I remember Micropolis from when they made drives of a size nobody else did. Our Netware server started with one.

* One of the Conners I tested spun up, spun down a few seconds later then a controller chip blew its top.

Conclusions so far

Could maybe network boot at least the first machine from another running OpenVMS on SIMH but though this has been fun I have to admit these aren't ever going to be visually interesting except for maybe the third machie if only it weren't so ... dead. Inspired by modern software I'm still learning VMS.

email me
my home page