The Elliott 803 behaves itself, the IRIS RDI fault returns and more gallery news
Post this page to popular social media
Another busy week and weekend at the museum...
Elliott 803 store fault 'goes away'
With all the problems experienced by Peter O last week he arrived on Saturday ready to do some further investigations. He went to the paper tape output bin where he had left his test tapes and discovered it was empty. It appears an over enthusiastic 'tidy up' had occurred during the week and the tapes had been removed. Fortunately they were found shortly afterwards and returned to the bin to be rewound. With scope and meter he tested the backplane where the 'suspect' board was located last week and all appeared normal. Further tests then showed the problem no longer existed... it had gone away, very strange. With a working system Peter then carried on with some of the tasks he started last week
One of tasks was splitting up the 2 large Algol compiler tapes into 9 smaller tapes so any one tape getting ripped did not mean punching a whole new large tape. This is where he hit his first snag. In the past the tape punch (connected to a PC as he who will remain nameless has not yet fixed the Creed teleprinter on the 803) has occasionally mis-punched the odd character, which happened when the first of the 9 tapes was produced. On further investigation a small software 'buglet' was discovered and the hardware flow control was enabled on the PC serial port resulting in further tapes being punched error free. All 9 tapes were loaded successfully into the 803 and a test program loaded, compiled and ran successfully. The system remained running with both stores for the rest of the day.
The IRIS air traffic control system suffers a setback
After working without problems all of last weekend, the problem of the RDI (Radar Data interface) returned causing the BEG PDP 11/84 (responsible for processing the radar data) to get stuck in a reboot cycle. So it looks like the problem is getting worse and likely to be a hardware fault with one of the boards. More detailed investigations will be needed to try and discover what boards are used for the RDI processing and then what board may be causing the problem. We suspect it is one of the custom built boards so not something that is readily available to swap. For now, operation of IRIS is going to be intermittent.
The operation to copy the main disk in the Walton replay unit was successfull. Investigation of the disk image showed the system was not using any of the recognsed file systems for storing data so may well be proprietory like the rest of the software.
Progress is made on the Electronic Office & Packet Switching galleries
The Electrician was in during the week to refit the electrics in the Electronic Office and to re-route some of the wiring in several other rooms to free up some circuit breakers in the distribution panel. This took several days to complete with the final bits due to be completed the following week. Now the wiring has been sorted out we can set about refitting the room proper. Lots of work ahead in the next few weeks.
The packet switching gallery is starting to take shape now. With all the walls painted, the floor screeded and carpet fitted, the display fitters were in to make a start on fitting out the displays. Some of the equipment demonstrations have been completed by Tony F and progress has been made with the audio/visual aspects of the display.
ICL operator console diagrams are printed
Having found the punch card 35mm slides of the operator console circuit diagrams last week a method of getting a printed copy needed to be found. After several suggestions for equipent that might do the job (which were also expensive), Tony F offered to try and scan them using his flatbed scanner. After a few attempts and adjustments he was able to get resonable quality images and produce several A3 size readable prints from them. Pete H and team now have a fighting chance of finding the problems with the console and getting one working. Hopefully it won't be much longer before he gets a chance to boot up the system proper.
Harwell Dekatron Computer restoration progress
Eddie continues to work on repairing and cleaning the relays. Several of the patch panel sockets appear to be damaged so replacements are being sought - markings on then suggest they were Radio Spares components so they may be readily available. The rust from the chromed areas of the rectifier unit has been removed using a 'home brew' electrolytic chemical bath but the steel mesh on the front still needs more work. It has been decided to put the tape readers and pasper tape editing 'suite' on wooded tables instead of the old ironwork they arrived on, so 2 old tables have been found and re-varnished by Tony.
John P and Peter M have given Tony several very large and a few small boxes full of valves that have been collected by the Colossus team over the years. Tony has already found several useful valves and some very rare, apparently unused GC10A Dekatron tubes (the type that the system was originally built with). He is still a long way from finding all that is needed and several adverts have been posted world-wide by valve enthusiasts so hopefully that will yield some more. Tony has also made a start on building a Dekatron spinner devices which will have 2 purposes; one to show off how the Dekatron actually works (appears to spin) and the other as a means of testing them. He has also made a start on a Dekatron reconditioning unit which can put several thousand volts through the tube which may be able to restore some of the Dekatrons to working order again.