IBM 5150, Part II: Landmark ROM Spaceknight

[MissArgent]

So, I've got my landmark ROMs in. We've got a 16k area failure, but not one simple enough that a swap of the first 4164 in Bank 0 fixed anything. Address 05139, bit 6 is the first to fail on that test. RAM tests fail, with what seems to be a pattern - many addresses having bits 0, 2, 6 and 7 specifically failing. Slow refresh also fails, but the 8237 passed it's own test. Anyone have any thoughts as to what my next steps should be?

UPDATE: Thread obsolete, seem to have sorted it out myself. Thanks, though!

 

[Trixter]

Wait, what was the solution?

 

[MissArgent]

a 16k failure after all, Bank 0 bit 6. I thought i'd ruled that out since stacking changed nothing, but it turned out I was making some hasty assumptions (not to mention misunderstood how the low RAM works). The machine's up and running perfectly now, at any rate! Ended up giving myself a lot more work than I needed to, but extra soldering practice and a crash course in logic probe operation never hurts. Thanks for your help with stuff, btw.