I've just got to a major milestone on a 6502 based
prototype (ie it actually
runs my code :) and I've got to say I felt I'd actually achieved something
Well done!
special after debugging with nowt but a meter and
led+resistor probe(s).
Being so restricted in the tools I had really made me *think* about what to
test and the end result is that I know exactly how the thing works.
This is true, and it's one reason for knowing how to use simple test gear
effectively. I have debugged a lot of things that way, watching the
brightness ot the LED to see if siagnal is stuck high, 50-50 (possibly a
clock waveform) or something else :-)
On the other hand, I once spend a sunday afternoon tracing a non-fault in
the memory arbitration circuitry of a Whitechapel MG1, using a simple
logic probe. I nearly went sane (!). The next day I sent off almost all
the money I had and bought a LogicDart....
The reason I san non-fault is that the thing reported a multi-bit DRAM
failure, and it appeared the CPU wasn't finding any RAM at all. That's
why I thoguht it was a control/arbitration problem, not a memory chip
failure. In fact, there was no fault at all. The boot ROM needed more RAM
than was on the mainboard -- IIRC it needed a pair of memory expansion
boards too. Nowehere was this mentioned...
Sorry Tony, but I was really glad I was working on a
breadboard!
Why? I would have thought having known-good connections and low-impedance
power and ground connections (neither of which tends to be the case on a
breadboard) would be very helpful in only having to find the _real_ faults.