80-Bus News

I recently discovered a hardware problem on my system that could cause a lot of difficulty in some circumstances. It seems that the fault could be fairly common, as I found out that another Nascom 2 user local to me had also had problems.

The power-on-reset (POR) on my system had never worked reliably, from the beginning, but I had not bothered to look for the cause, as it did not trouble me too much and it is not easy to get at my Nascom 2 card. I simply used the manual reset. This continued even after I multi-mapped my system as described in earlier issues of this newsletter. I had adopted the habit of always switching to Nas-Sys (i.e. standard Nascom), before switching off, so that the next time the machine was switched on, resetting would not cause a CP/M boot, with SIMON looking for non-existent disks, thus avoiding ‘Head Banging’.

My normal procedure was to switch on, and press reset TWICE(!!) in quick succession. Nas-Sys would announce itself, and I could proceed to use whatever system I desired, using the System Switch. (It was necessary to use 2 resets to get Nas-Sys booted correctly.)

A few weeks ago though, I started to interface a Gemini GM822 RTC module, and I had trouble with Clock Data corruption. In the search for the cause, I decided that it was time to get the POR working, since I did not know what the Z80 was doing between the manual Resets, and this could be contributing to the data corruption. It did not take long to get the POR working, once I had got the N2 card out of its case. In order to test it in this position though, I had to run the card on its own, with POR set to address 0000H, and with Nas-Sys in its normal socket. The POR appeared to work correctly. However, when the system was reassembled in its case, with the POR set to operate at 2000H/​4000H (depending on the DIL switch LSW1 and my system switch), there were still some very peculiar results.

With Nas-Sys selected (reset jump to 4000H), a manual reset was STILL needed after the POR. However, with CP/M selected (reset jump to 2000H), the POR appeared to work correctly. In order to see what was happening under Nas-Sys, I connected up a TV to the modulated O/P of the N2, and my video monitor to the IVC card O/P. On running up the system, I discovered that the POR loaded SIMON and looked for a disk, irrespective of the system switch position. I still had to make the second, manual reset to get Nas-Sys loaded correctly. Without a logic analyser or some other suitable testgear, it is impossible to be sure where the POR was going, but it seems likely that it was going to 0000H, and then the Z80 was charging up through RAM until it found some program to execute. In my case this was the CP/M boot ROM at 2000H on the paged in EPROM card. (Which would be found irrespective of the position of the system switch.)

Having found out what was happening, the next question was why? The trouble was almost certainly due to some malfunction of the reset-jump-multiplexer. I spent some time poking about with a ’scope and logic probe, and tried changing chips, but all with no result. A colleague who runs identical hardware had no problems on his system. He loaned me his EPROM card to try, but the results were just the same, so the trouble seemed to be on the CPU card. After a lot more thinking and poking about I decided to see if anyone knew the answer, so I rang Nascom and a couple of other people, but the fault seemed unknown to then, so apart from a couple of suggestions, they could not help.

I had been offered the loan of a CPU card by another friend, which I had not previously taken up because it meant a ten mile round trip, but it seemed the next logical step, so I borrowed the card, and tried it. It ran perfectly. I took it out, and decided to check the components of the POR circuit. They were identical to mine EXCEPT for the capacitor C3. On Phil’s card this was 100uf instead of the 68uf supplied and as shown in the circuits and component lists. I rang Phil. Yes, he had had trouble with reset at some time in the past and had probably altered C3, but he could not remember the details. I changed C3 to