Beige-o-Vision

03/03/2024

Here's a longer version of the video The National Museum of Computing posted this week. That's an HP 9100A from the museum's reserve collection.

The HP 9100A is an early programmable electronic calculator. It's probably the first programmable scientific calculator. It was marketed from 1968 to 1974. Put it another way it's a desktop computer with a specialised interface. Each key represents a function in ROM. The ROM is a variant on rope memory. It has a total of 19 registers, but it can store up to 14 program instructions in each of 14 of those registers. Five registers are data only. It understands numbers as floating point BCD. Three registers X, Y and Z are shown on the green phosphor CRT. The CRT is a vector display.

This example had been in an unknown state. So my fellow volunteers and I worked on it over a couple of weekends. What I'm doing here is entering and testing some functions based on the manual. Then I (badly) wrote an implementation of a Fibonacci sequence calculator. That's the program you see it run. Later in the day I entered the whole diagnostic program by hand and let that run for a while. All the logic and memory works. We still need to locate some magnetic cards to really test out the card reader.

09/12/2023

Our project backlog just got deeper to the point of being bottomless.

Behold the new precious.

What you're looking at is an HP 2100. The chassis is an 'S' variant. Apparently that only indicates a set of options. This one has a completely different set of options from the original S spec. In fact it was originally setup with an IO Expander chassis. We didn't end up buying that. As the system currently sits it has 24KW of 16bit core memory. 8KW of that is marked as suspect.

The lump at the bottom is its hard drive. That's a 7900A. It has one fixed 2.5 MB platter in the bottom, and space for a single platter disk pack (another 2.5 MB) at the top. Believe it or not, the hard drive was serviced no longer ago than 2018.

We also have a paper tape reader, though it's not shown. All of this is still setup for the US. So it will have to wait for conversion. However, prior to purchase, it all powered up even if it was 'untested'. Who knows. Maybe it'll just work.

Nahhhh... that would be silly.

30/11/2023

Stopped into https://museum.syssrc.com and Bloop Museum yesterday. Highly recommended!

11/11/2023

And todays clean up and eval job is this Wang 520. Check out the rope memory! Despite the display being on, I'm afraid it's not really operational. Needs some work. Cassette door is off as the case is loose. I'm afraid the printer cover and function key strip are missing.

07/10/2023

So this arrived.
May I present our rather grimy Olivetti P6066.
We're still learning about it, since this was kinda a fluke purchase. It's had a bit of a clean and we took a bunch of pictures for research. If you have more information on these, please do get in touch.
Meanwhile, pictures!

07/10/2023

Another day of working through storage to do a basic tidy on a recent acquisition. This is another rarity. It's a Creed Envoy. Think of it as Britain's answer to the Teletype ASR35. It's an 8bit teleprinter with a punch and reader. Initially it looked pretty rough. But I've taken it down to the chassis, and it looks like it'll clean up. A paint job would help. Inside it's mostly aluminium. There's some surface rust on the steel which I think will clean up. Pictures below.

11/12/2022

Work continues on the Elliott 920M project. This was always going to be an ongoing background project. The key first step is to get a power supply built.

A working draft for the design is now up on the GitHub page.

https://beige-o-vision.github.io/Elliott-920M-Project/power.html

20/11/2022

Can you tell what it is yet? One of the winter projects is to build a power supply for an Elliott 920M this needs multiple voltages in sequence. And because it’s old and possibly grumpy those power supplies need to be monitored and controlled for failure modes so here we an with an analogue mux and a bunch of voltage dividers. Needs that last voltage integrated and some crowbars. Sequencing will be a bought in module, and then there are Hall effect current sensors… Much to do.

20/03/2022

Another rarity and treat from the The National Museum of Computing reserve collection in stores. This is a BC(S) / BCL Molecular 18. They were made for about 20 years. This is a Mark V. It’s a waist high general purpose minicomputer made in the UK. It was typically usually used in sales processing, stock management, etc.

02/03/2022

Some Elliott stuff -- the 920M we’ve been playing with is closely related to the electronic kitchen counter… otherwise known as the 903. Typically the guts wouldn’t be on show as they are here. It’s just be a Formica table top, with peripherals on it, or on some table near by. This is from the The National Museum of Computing

26/02/2022

It’s a volunteering day at The National Museum of Computing. Just getting ready to put this Superbrain back in storage. A new exhibit will be taking its place in the Temporary Exhibition Gallery.

11/02/2022

Making some slow progress on evaluating the Elliott 920M. Completed a first pass on documenting the modules and their locations. This is important since there’s no available schematic, and we have no spares. So having some ability to interface with and troubleshoot the machine will depend on getting to grips with how it all works. Understanding the component groups and trying to map those to known functional units in the machine architecture will help.

Update here:

https://beige-o-vision.github.io/Elliott-920M-Project/modules.html

11/02/2022

Some shots of the guts of a DEC RM03 Disk Pack Drive. It goes with the PDP-11. The team The National Museum of Computing has one working. Now they want to get this one going too.

08/02/2022

Under the cover’s this the PDP-11/70 CPU at The National Museum of Computing. The systems under restoration.

08/02/2022

From the weekend at The National Museum of Computing, another ongoing restoration project. This is a PDP-11/70… and yes it’s working.

07/02/2022

Also from The National Museum of Computing, another Avionics repackage of a British minicomputer - the Argus 400.
https://www.instagram.com/p/CZq3QI-NHrN/?utm_medium=copy_link

07/02/2022

From the weekend The National Museum of Computing https://www.instagram.com/p/CZq11lJtxr4/?utm_medium=copy_link

01/02/2022

Wow. That’s a whole lot of circuit boards. Ignoring the block to the left for the moment, there are 8 bays with 4 columns. Each of those has something like 13 or 15 separate modules. No sockets all wire wrap. And then there’s the that dark block on the left. That’s all core memory - little magnetic donuts strung together to look like a screen door netting then folded up into a brick. The translucent blocks are all the control circuits. That’s just … exotic. And then there are the modules themselves. That’s another story. What you’re looking at are the guts of that Elliott 920M Computer, designed around 1966-67.

31/01/2022

Been doing some initial evaluation of the Elliott. You can see some high res pictures and notes here:
https://beige-o-vision.github.io/Elliott-920M-Project/initial-observations.html

31/01/2022

Meanwhile, hopefully with finish shooting pickups for that last of the HP 9845 series this week.

31/01/2022

Surprise new project thread started up this weekend. We've started a project log for that one -- here: https://beige-o-vision.github.io/Elliott-920M-Project/

This is the Beige-o-Vision project page for the Elliott 920M. It’s one of the 900-series of 18 and 12-bit small computers manufactured from the 1960s by the British computer company Elliott Brothers. The 920M is an 18-bit family member. It’s compatible with other 920 series members and the 903. ...

22/01/2022

https://www.instagram.com/p/CZB-KvjtjuC/?utm_medium=share_sheet

21/01/2022

Meanwhile, from over on Instagram, here’s a detail snapshot for one of The National Museum of Computing’s large systems gallery objects. It’s an ICL disk pack drive from a model 1900-something mainframe. Just kind like the atomic-age aesthetic, don’t you?

21/01/2022

Working on trying to program some 3D models on the HP 9845. Wanted to show off what it can be made to do. But that’s holding up the next video a bit. Maybe reserve that for the future?

10/01/2022

With one CRT down, thankfully we can still complete the next video with the spare. But we're back to have to peer through the screen rot. With no more spares, we won't risk renovating this one.

09/01/2022

Tiny sliver chip from the terminal of the CRT, long hiss of refilling gas. A few frustrated tears shed watching a great deal of time, care and money going right down... the tube. 😪 😒 The gallows-humour word play helps... a little.

09/01/2022

Urgh!!!! Disaster!!! Managed to kill the 9845's tube on reassembly! No big dangerous implosions. But that dispiriting sound of hissing gas as one of the pins in the neck takes a knock and the gas leaks in. I'm grateful we have a spare, but I think we'll have to live with the screen cataracts on that one for a while. It'll be some time before I try all that again!

08/01/2022

Spent a good part of the day today removing the screen rot from the HP 9845 so we can start shooting the next video on using the system. Still need to reassemble everything. Hope it all still works after this amount of fannying about!

01/01/2022

New video is up. We get under the skin on an HP 9845!
https://youtu.be/rtmmghk0Df4

We go under the covers to see how HP assembled its early workstation the 9845B. We’ll be looking at the assemblies and talking about the processor and some ...

27/12/2021

Coming up (hopefully) for New Years 2022… the next YouTube video in our series on the HP 9845! Keep watching this space, and of course… subscribe on YouTube.