Progress on 2966, IRIS's RDI fault returns, chatting up 'The old lady' and a chill in the air
Post this page to popular social media
Can't think of anything interesting to put here this time so onto....
Undocumented error 39 and hard error 48 faults on the DCU
With a working operators console, Pete H, Delwyn and Nick are now concentrating on getting the SCU/DCU (Storage Control Unit / Device Control Unit) cabinet working. This cabinet controls the main computer memory store and interfaces to the disk drives, printer and card readers so is critical to the overall operation. On Saturday the team were complimented by another visit by Alistair from Fujitsu so Pete was confident progress could be made.
The SCU/DCU has been powering up OK for the past few months but it needs the Ops console to initiate its self-test sequences. This was done and the first of many error codes appeared on the 2 digit display on the DCU and thus started a repeat cycle of checking the codes in the manuals, changing various boards and restarting the system. The situation was not helped by the DCU stopping showing several codes not actually listed in the manuals. Throughout the day undocumented code 39 and documented code 48 appeared frequently and the number of suspect boards began to increase. With the frequent power on/off it was no surprise when one of the 150A 5V PSUs died and the noise from the main blower fan meant it would fail soon so both were replaced. The system has 2 other DCU cabinets so these were being used to source any spares needed (the are not actually needed to get the system operational). They were however down to the last working PSU so our resident PSU expert will have to try and get the faulty ones working again, this time helped with some actual circuit diagrams from the ICL card archive.
So in the end a frustrating day for all as they were not able to get the SCU/DCU passed its self-tests and it was left in the error 48 condition. Pete H will look at one of the other 2 DCU cabinets later in the week to see if they could be powered up and tried next weekend.
IRIS crashed and our hopes are dashed
With IRIS working when switched on during the week, Peter V thought he had found the cause of the RDI fault. However within an hour of switching on on Saturday the fault returned again causing the RBEG to continuously reboot. It is clear this is going to be a long investigation as there are a total of 11 boards involved in the interface (4 in the BEG, 5 in the RBEG), and without anything to go on it will be a case of swap a card and see what happens - we don't have the luxury of just swapping all the cards before we don't actually have enough spares to do that so we have to try and identify the board causing the fault. The main drawback with this is the system could run for several hours or days before it fails. At this rate, Peter reckons it will be Christmas before all the boards are changed.
Dekatron spinning and reconditioning 101
Tony F and team are making intentionally slow but steady progress on Harwell computer restoration. We are still very much in the early stages on this project but it is good to see activity and things being done each week. Eddie continues to clean, check and adjust the relays and has also traced the +50v wiring on the back of the relay racks. It is still unknown as the the wire condition until tests are performed so we don't yet know how much of the original wiring can remain.
Tony F has now mounted the 6 input tape readers + controller on a new polished table and is currently re-routing the wiring loom for them. He has also found a more recent diagram of the rectifier unit and identified the failed components found last week to the +370V supply circuit plus a pencilled note identifying a 'dodgy broken plastic contact' which may be related to the failures.
On Saturday, Tony had a visit from a Dekatron enthusiast and now has the necessary information to make a tester/re conditioner unit for checking Dekatrons. He was also told about how the Dekatrons can 'stick' and ways to recover from this situation. Hopefully in time we will know how many of the Dekatrons fitted are still usable.
Who's been chatting up 'The old lady'
We don't want to worry anyone unduly but there have been rumours around the Museum that PeterO has been seen chatting 'nicely' to 'The Old Lady' all weekend and some say she has been responding well to it. Talking to living plants is understandable but to a lump of metal and electricity may be going a little too far Peter, but I guess if it means the 803 does its thing then I guess we can let it pass. Although trying to talk her into running an 8K program in 4K of store may be a bit too optimistic (he forgot to enable the 2nd store!)
Anyway, whatever he did say resulted in the system running without any problems all day Saturday and apart from a small issue when cold on Sunday, all day Sunday as well. He even managed to fix a programming fault in his cogs program which was previously thought to be caused by a compiler or even a hardware fault.
What do you mean the boiler does not work yet!
The weather has been getting a tad cold over the week culminating in a very cold Saturday. All the volunteers were warned to wear extra layers but it was only after we had arrived that we were informed that the new boiler was not working yet (some more 'tests' were still needed) and the old ones no longer worked. So it was a case of getting as much equipment on as possible to try and warm the place up before we opened. Fortunately some foresight by Pete H meant the heat generated by the ICL 2966 could be channelled into the large systems room rather than going out of the window, which together with the Elliott 803 and other energy inefficient old systems meant the museum was soon up to an acceptable temperature. Not quite what the main exhibits were designed for but we are always looking for interesting ways to solve problems.
Other project progress
The 'packet switching' gallery is still being fitted out by BEX and it is looking really good. There will be lots of information on display panels plus some 'animation' and one or two surprises for visitors. It is still some way off being finished though.
The 'Electronic Office' is also progressing well. With the floor being fitted last weekend, the remaining wiring was completed and we have been finalising some of the technical areas of the display. We can't say too much about what will be on display but it will certainly be unique and very visual.