11 Oct
2007
11 Oct
'07
4:05 p.m.
No, I'm
working under the assumption that hardware works logically, whe=
n
it fails to work there's a good reason for
it, and that before you can
claim to have repaired it you have to have found that reason.
Heh, I remember troubleshooting a PC that would work with a CD ROM=20
disconnected, but not with it connected - even if you left the IDE cable =
off=20
(actually it's long enough ago that it might have been a Mitsumi interfac=
e)
Guess what it was? The 'scope told me, the +12 rail was all over the pla=
ce. =20
Once the second CD ROM was connected up, it swung wildly from 9v to 15v. =
=20 How am I
inconveniencing myself? What do you think I want to do that
could be done more easily by having a modern PC? (And remember I think
_you're_ inconveniencing yourself by not having electronic test gear an=
d
learning how to use it)
You've said yourself that you don't have (easy) access to resources like=20
bitsavers.org, because some of the formats used require fairly modern PCs=
. =20
I'd find that pretty inconvenient. ARGH! When are you going to understand that I
don't replace parts witho=
ut
knowing what's failed and why!. I simply
don't. I never will...
I must admit, I don't just replace bits without at least trying to guess =
why=20
they might have failed. Obviously if the cause of the failure is obvious=
,=20
it's not worth spending time on (most recent example being (sorry folks) =
one=20
of my cars, where it was "Let's see, why might these ball joints have 6mm=
of=20
play? Maybe because of 19 years of British roads?"). It's a lot less inconvenient to me to stick
to hardware I know, that I
understand, and that I can keep going to avoid any such problems.
Not keen on learning new stuff, then? something
worth at best hundreds and probably in fact worth perhaps
the cost of a pint or a train ticket is not sensible pragmatism, it's
You also have to realise that not everybody knows people who give away
old PC hardawre. I don't, for example. not commendable attention to detail, it's bl**dy
daft. And to say "I
won't run it if I can't repair it" is dafter still!
I disgaree, but there you are...
I've repaired computer boards (the memory board for my PDP11, for one) an=
d=20
I've repaired "sealed" ECUs in cars (the suspension ECU in Citro=C3=ABn X=
Ms, many=20
times). I *replaced* a ball joint, because it was 30 quid and I've got n=
o=20
chance of making one, especially for that kind of money. I *replaced* a=20
OK, although I drove the local Citroan parts centre nuts back when we had
the BX. I rebuilt the steering rack (the procedure is in the workshop
manual), and ordered all sorts of little bits -- damper yoke, shims,
clips, etc. They had never had to supply those before -- apparently their
workshop just replaced the entire steerign rack assembly.
serial controller card, 'cos it was a fiver and
again I doubt I could bui=
ld=20
one for that.
No, but you could ggt the 1488/1489 for less than that, couldn't you? 90%
of RS232 problems are the driver/receiver chip. Althought I had a fault
in an HP9816 where not onlyhad both 1488s and both 1489s failed, but so
had the 8250 serial chip [1]. No idea how....
]1] Yes, I know it's a 68000 system. And I know that's not the normal
serial chip in a 68K machine. But that's what HP used.
-tony