ICL 2966 restoration during 2012
Post this page to popular social media
Each TNMOC project has either a working group or project team assigned to do the work. Working groups are either managed in association with the CCS (Computer Conservation Society) or solely within the Museum.
Below are updates on the ICL 2966 restoration project during 2012.
29/01/2012 update from Delwyn Holroyd
We had a series of power supply problems in the main cabinets over the last few months of 2011. We are now finally back in the position of having working spares for the main types of supply, with a couple of modules still awaiting attention.
The new multi-purpose peripheral interface PCB is now complete. Testing was well under way with the working EDS80 drive and the scratch pack we had already imaged when we experienced another head contact. This is thought to have been due to the room filling with dense smoke coming from the power supply on another machine (which shall remain nameless!) - there is a helpful illustration in the drive manual showing the size of a typical smoke particle relative to the disk-head gap (huge). It is just unbelievably bad luck for this to have occurred on one of the few days we have run the EDS80 with heads loaded, although thankfully no data has been lost. The drive and pack still need to be assessed for damage and cleaned, but we are hopeful it can be recovered. This episode has made it abundantly clear that we will ultimately need to build drive emulators if we wish to run any of the major operating systems for the machine on a regular basis, all of which require disks.
Work is almost complete on the firmware to support the new interface PCB in it's role as a peripheral interface. The aim is to allow peripherals such as the line printer, card reader and ultimately tape drives to be regularly demonstrated without needing to run up the full machine. We have successfully printed to the line printer using the interface, and booted the machine from it, thus proving operation as both an X (controller) and Y (peripheral) end.
We hit an unexpected problem interfacing the board to the DCU. It uses modern surface-mount RS-485 transmitters to drive the DCU, which are in theory electrically compatible. However the load presented by the DCU receivers was such that the fancy protection circuitry in the transmitter chips interpreted it as a fault on the cable and promptly shut down! This was eventually solved with pull-ups to get around the high capacitance of the DCU receivers.
The card reader store boards and all our stock of spares were tested and found to have multiple RAM chip faults. Replacement RAM (new old stock) has been procured, and a simple tester has been constructed to enable soak testing of the replacements before fitting to boards. We now have three working store boards for the card readers, and work is just commencing on the mechanism control boards.