Horun wrote on 2020-03-29, 00:30:
[...]
Nice ! Got to love EISA 😁
Love it - but only the way the mother of an absolutely demon spawn child loves it unconditionally... whoever thought this was easier than messing around with jumpers was either doing drugs or naive to the point of stupidity 😮
Getting the system up and running was easy, but then I wanted to add two non-EISA cards - an 8b ISA I/O card for serial and parallel ports and an SB16 as even though I have low opinion of the beasts, there aren't that many cards with EISA cfg, and I seem to have lost my PAS16 somehow ( 😦 - must be in the house somewhere, probably even in the same room. But not findable at present). Nothing too complicated, but some resources needed reserving. Then the fun started. The SCSI + FDD adapter defaulted to 330, and seeing as Ethernet was already occupying 300 (and could only do 330 otherwise). I needed to move it for MIDI. No problem. Four other options. None of which gave conflicts in the config tool. So reboot and - hangs on SCSI. Of course to change settings you need to boot the system to run cfg.exe. And that is on the SCSI HDD that can't boot. So removed SCSI, added an ISA multi I/O card, booted from floppy on that. Ran cfg.exe from floppy. Added SCSI again. Ran cfg.exe from HDD to ensure that the config would be in sync. Tried the next address. Hung again. Repeated everything again. Eventually found that 334 would work. In the end you need all the knowledge you would need for non-PnP ISA, just with an added layer of complexity and an added dependancy before you can fix things that go wrong. Lovely.
Of course, after that I got about half an hour of a correctly working system before SCSI started playing up. Determined that it wasn't the drive or adapter, nor the cable, and it would work with a plethora of adapters and a 68p cable with built-in active termination. That screams "termination issues", but of course my spare 50p terminator is also AWOL. Fortunately someone's selling them locally for a good price so ordered two.
Ugh ! Hate KB errors/issues. Maybe a partly dry cap and it could end up working OK after having power on for a while. Have seen some odd behavior on old boards stored for a while that smooths out after powered up for a hour or overnite. I have a KB error on soc5 that I can not figure out. No visible corrosion, all traces good, volts good, diodes good. Hoping it is not the via 82c416 chip........
Hope so too.
Figured mine out. Turns out there were two separate issues causing all three keyboards I tried to fail. Firstly the board delivers very little current over its DIN port. That's insufficient for a power-hungry Model M and throws off my KVM switch too. So I connected an other AT/DIN keyboard directly. No power issues there, but it had a stuck keyswitch. Fixed that and at least one worry less.
Also a more recent retro experience: picked up a set of Harman Kardon Soundstick II speakers. I was impressed by the sound my old IBM Aptiva speakers from 1996 produced, but these are in a whole different league. Have been listening to classical music and prog rock/metal continuously and it never sounded so good 😀