Elliott 803 gets arthritis, ICL 2966 console is working, IRIS investigations and other news
Post this page to popular social media
In a packed program tonight we have.....
Elliott 803's unexpected finding
When PeterO powered up the machine on Saturday morning all seemed well. He ran a few store tests and no parity errors showed up so he loaded the ALGOL compiler and compiled the large "cogs" programme as a test. It compiled and ran OK so things were looking good. However soon after the museum opened at 1pm the compiler started to behave erratically, pointing to some sort of problem. The fault seems intermittent and may explain some strange behaviour seen with the music programs on the previous Thursday. Then just to add to the problem the parity errors in the 2nd store returned. It was time for more fault finding...
A quick test proved it was not a loose connection from the read amplifier in the 2nd store however, the fault could still be cured by pressing in board 46 so it was time to check the back wiring for the board slot. He identified the pins on the edge connector backplane related to bit 23 (the faulty one) and bit 24 (a good one) and put a scope probe on them. As soon as he had done this the fault went away. Waggling the probe on the pin caused the fault to appear then go away... there must be a bad connection on the pin and closer inspection discovered the wire going through the pin but no sign of any solder... it was a dry joint.... in fact the pin looked like it had never been soldered since it was built in the early 1960s. A quick dab with a soldering iron fixed the problem and the 2nd store ran for the rest of the day without incident.
A few funnies with the ALGOL compile did occur, as did the punch which was punching the occasional wrong character although it was fine when copying tapes, so it may be a problem with a particular sequence of instructions... More investigations will be needed next week but that's two out of two fixes... Nice one Peter.
ICL 2966 console fixed but still work to do
Armed with the circuit diagrams of the console, Pete H and his team set about discovering why the console did not work, the last (we thought) remaining hurdle before getting the system working. It was not long before several components were identified as faulty, suitable replacements found and the final delight of seeing the console system finally boot up and show something readable on the display. At last, an attempt to boot the main processor could be attempted. Unfortunately things did not go as expected with the main processor and memory cabinets showing errors. Still, a major milestone has been passed but the champagne moment appears to be some way off yet. Nice work Pete and Co... lets hope the remaining faults can be identified and fixed soon.
Harwell Dekatron Computer progress
Work continues on the preliminary work to restore the HDC to full working order. Tony F is currently working on the rectifier unit and has found one of the HT chokes is open circuit and pitch has been oozing out. There is also an oily residue on the underside of the chassis suggesting a prolonged overload condition had occurred so a clean and rewiring is needed. Some wire wound resistors have also failed but the cause is likely to be something external to the unit. Testing of the other components has been successful so fixing the unit should be possible.
The 6 paper tape units have been remounted on a newly varnished wooded table and 2 Dekatron 'spinners' have been built to show the original GC10A and newer GC10B Dekatrons in action. Eddie continues to methodically remove, clean and adjust the relay sets.
Johan has made a start on checking the 50V relay power supply. The mains and output cables will require replacement due to perishing and cracking. There are some electrolytic capacitors dated 1947 which will need to be electrically re-formed, or possibly replaced.
On Monday, we were visited by Gurney Thomas and his daughter Caroline, and Dick and Doreen Barnes. Gurney and Dick are two of the original designers of the computer. Gurney (whom neither Kevin nor I had previously met) designed the Dekatron store units having previously worked on Dekatron counting equipment at Harwell.
IRIS Air Traffic Control RDI fault investigations
Peter V spent most of Saturday continuing the investigation on why the RDI (Radar Data Interface) fault had returned and was causing the RBEG to reboot continuously. After many hours of board swapping and head scratching he discovered a small 'link' pcb fitted to one of the cards was protruding out from the board far enough to make contact with the board in the next card slot. The spares he found had a chip based link which was half the hight, which once fitted resulted in the system running for the last hour the museum was open before the system had to be switched off. The system was on throughout Sundays opening and ran without any problems so it looks like the problem may have been fixed, but more time will be needed before it can be confirmed. Nice work Peter.
New gallery progress
Work is continuing at a fast pace in the packet switching gallery with more of the display fitted. Still lots to do but the overall concept which has taken many months to design is finally coming together thanks to a lot of help from volunteers, display fitters and our graphic designer.
The electrical work in the revamped Electronic office is almost completed. The floor has now been laid and lots of items for the room have been sourced from various places.
The 2 day weekend event drew a lot of people to Bletchley Park, and with the help of several volunteers, we were able to open on Saturday and Sunday. It was a very busy time with lots of good feedback from those who looked around the museum.