During the week we had lots of discussions about the 32kB (Yes 32kB, not 32MB or 32GB) Omnibus RAM board that Steve Lafferty prototyped and Vince Slyngstad made into a real PCB. Last week we found that the board was sensitive to the backplane slot that it was in and was not 100% reliable. The conclusion was that one of the decoder chips was too fast and was treating glitches as real signals. We swapped a 74ABT540 IC for a 74LS540 that is about 10x slower. The RAM ran diagnostics for 90 minutes without a single error. Today we added a second serial port to the PDP-8/E so we can boot the OS/8 operating system from a emulated disk and start debugging the 8 diskette controller. We need to make a serial cable this week so we can continue debugging next week.
Posted on: Sun, 18 May 2014 00:08:24 +0000