Dead AIM 65??

[Tony Duell]

> 
> 
> Anyway, back to the original subject.
>  There is no magic bullet to bring these older machines
> back to life. Some of the things that can help
> are, not in any particular order:
> 
> 1. Volt meter
> 2. Logic Probe
> 3. Set of schematics
> 4. Oscilloscope
> 5. Soldering iron w/ solder
> 6. Parts to swap
> 7. assembler/disassembler
> 8. EPROM programmer
> 9. Source listing
> 10. Needle nose pliers
> 11. solder sucker and/or solder wick

I would disagree with (6). My metric for how well a classic computer 
repair has gone is how few components I replaced before I found the real 
cause. My aim, of course, is to make measurements, think about them, then 
replace just one part (or a small set of parts, all shown to be 
defective) and get the machine to come up first time. Alas I am not that 
good.



> 
>  Also, one needs clear logical thinking ( not all that
> common in the world today ).

I would totally agree with that. You can do a lot with minimal test 
equipment if you really understnad what you are doing, and can think 
about the symptoms. 

Occasionally. I find myself saying 'If only I had a 64 channel logic 
analuser I'd get this darn computer working'. What this generally means 
is that I don;'t really understnad what I am doing. And after a lot more 
thought, I realise I could have found the fault with nothing more than a 
VOM and logic probe.

I've recently been chatting to a chap in Germany about repairing an 
HP9810. I've suggested measurements to try (he knows what he's doing, 
he's jsut never worked on a bit-serial processor built from TTL before). 
I explain what he's looking for, what his results mean, etc. On the 
second chip replacement, the machine came up (and I think the first one 
he replaced was defective too). That, IMHO, is real debugging...

>  The debugging procedure is in specific order:
> 1. gather symptoms and observations.
> 2. Form hypothesis of possible cause.
> 3. Test to see if hypothesis is correct.
> 4. If problem not found, include results into observations
>    of step one and repeat.

Yes, the soldering iron is often the last thing you use when repairing a 
classic computer. Your first steps _must_ be to make measurements and 
think about them. 

And please don't get me started on the ridiculous idea of replacing 
random parts until the machine seems to work...

-tony