surprise STF refurb

Problems with your machine in general.
User avatar
stween
Posts: 237
Joined: Sat Sep 08, 2018 3:10 pm
Location: Brooklyn & Edinburgh

Re: surprise STF refurb

Post by stween »

stween wrote: Wed May 11, 2022 3:15 amI don't totally trust the current behaviour until I replace the RP1+RP2 and RP3+RP4 pull-ups again, this time for something higher than 2.2kohm each since they read as wired in parallel (it's this motherboard layout).

Presumably putting 4.7k in each will give about 2.35k on those lines, which sounds like a more typical setup on other motherboard layouts.
Alright! I think I have a good outcome. rubber_jonnie was right, the DMA chip the board arrived with seems to have been the root.

Final configuration on the pull-ups: RPs 1-6 are each 4.7k, RPs 7-9 are each 2.2k. RP5 and RP6 are how they arrived, I simply didn't change those ones.

That means the resistance measured between the CPU and the appropriate resistor networks looks like this:
  • CPU and (RP1 and 2) is 2.3k
  • CPU and (RP3 and 4) is 2.3k
  • CPU and RP5 is 4.7k
  • CPU and RP6 is 4.7k
  • CPU and RP7 is 2.2k
  • CPU and RP8 is 2.2k
  • CPU and RP9 is 2.2k
I also took out the DMA fix because I didn't do a great job on it.

This machine has now run for hours on the gembench tests in loop, and run over 50 cycles on the diagnostic cartridge disk exerciser.

For a moment I thought I'd left a bad solder joint somewhere: initially I got a low number of disk read errors on the disk exerciser, and I was able to trigger an "E9 Bad instruction fetch" or "EB Bus Error" by tapping the board around the ROMs, where RPs 1-4 are located. But I haven't been able to trigger it in the last day, so now I think I just had some cruft on the desk surface or a grubby contact on the cartridge port. Last suspicion in particular because I've not been able to trigger any similar errors when running without the cartridge.

So I'm taking this as good. I have a surprisingly good early-stage STF here from 1985. The case is a little grubby and one of the keyboard LEDs is broken, but otherwise it's in solid shape. A success in the end!

I appreciate the nudges on the thread, I would probably have given up completely without the pointers on where to look closer. And I learned a lot about how this thing is wired!
Post Reply

Return to “HARDWARE ISSUES”