> Nor mine. For that matter, Tony won't
like my answers, either :-)
On Mon, 25 Jun 2012, Tony Duell wrote:
Since you posted them publically, I think it is
reasonable for me to
reply to them, whether you like it or not :-).
I like it very much!
I was thinking about you with every sentence of my post. I was hoping
that you would join in, fully knowing that we would not agree on the
"cheap tools" aspect.
AS I think I said, there is no real 'right' or 'wrong' way to look at
this. There are many people here who do design, constrruction, reparis,
etc. And we all do it differnetly. All each of us can say is waht works
for us. Anyone else can read what we're all saying and decide which bits
A few days ago, a friend was repairing a hot-tub (spa)
controller. He was
pretty sure that he had 220V, but the cheap meter said 0. Then the probe
The correct procedure, I beleive, is to measurew a known live point, then
to measure the point you want to work on, then to measuer the live point
again. Only if all 3 readings are as you expect do you start work. Even
that's not infallable, but it's better than assuming the meter is OK.
A firend of mine has a voltmeter for checking 11kV (andn maybe 33kV)
overhead lines. In the carrying case is a little EHT PSU (perahps about
10kV), powered by batteries. The proedure is to measrue that, then the
line, then the test PSU again.
fell off of the end of its wire - so THAT'S why it
read 0. A meter, good
or bad, can not be a substitute for common sense.
That is very true...
You are skipping over something vert important
More than a few things! I only intended it as a framework, to be expanded
into a more thorough introduction to how to get started.
The whole issue of 'ground' or a common voltage reference point is a lot
more complex than it first appears. Don Vonada said 'There is no such
thing as ground' :-)_
To me that statement has 2 meanings.
The first is that you can take any point you like as your reference to
measure votlages with repsect to. In some systems it makes sense to take
different reference points for differnet parts of the circuit, the
obvious common example being an SMPSU where you would take the so-called
0V output terminal as nthe reference for votlages on the output side of
hte isolation barrier but the -ve side of the maisn smoothing capacitor
for votlage aroudn the chopper on the maisn side of hte isolation
barrier. Be careful that in the latter case there is a considerable
voltage between that refernece point and mains earth amd you can't
connect (most) 'scopes directly yo it.
The second meaning is more subtle. It's the fact that any connection --
any bit of wire, PCB track, etc has impedance. At high frequencies the
inductance of a bit of wire becomes imporatnet. So you might have 2
poitns that appear to be connected together by a nice, solid, PCB track
and still find a ripple on one wrt the other. Ths 'groudn bounce' is a
particular provlem with high-speed digital electroncis.
voltmeters in genral, have 2 leads. The poiunt is that it makes no real
sense ot say 'the votlage here is 5V'. What you mean is 'the voltage here
is 5V _with respect to this point which I am calling ground'. Knowing
what point to take as the common 'ground' is often not obvious when you
are startign out.
I did mention that (in the next step), but these instructions need a LOT
Sure, I was just trying to provide some.
My list of simple hand tools for ocmputer repair
is quite long, and to be
fair you problaby don't need all of them. But I find I need screwdrives
from 1mm to about 8mm, Phillips and Pozidriv ones too. And Allen hex
tools, torx drivers, nutdrivers, etc. In some cases it does depend on
what you are sorkign on. If you stick to C64s, you will not need Bristol
Spline tools. If yoyu work on Flexowriters, or even IBM 5155s, you do.
I think that a set of screwdriver bits is a good start there. There are
better screwdrivers than the bit holders, BUT, that could encourage using
the wrong size when a simple bit change (with a larger selection) could
have been done. My father understood the difference between "plain" and
Covnersely, having dedicated screwdrives for the common sixes can
encourage you to just pick up the right tool, whereas picking out the
corret insert bit and putting it in the holder is more work and you mare
feel 'Aarn it, I'll use this little flat balde scrrewdriver jamped across
the slot to get this Phillips screw out'.
I have a mixtrue of dedicated screewdrivs fro common things (flatblade,
Phillips, Pozidriv) and interchangable blades or bits for things like
Allen hex and all the security screws.
Get all of the technical documents for all of the machines that you want
to work on, or are even just curious about
Easier said than done in soem cases
a career of its own!
Im my case, yes :-)
I think one thing that Fred and I agree on (from
reading this ) is that
there is no short-cut to being able to repair computers. You _do_ need
to learn to use tools,. you do need ot learn how things should work, you
do need to learn basic (and more) electronics.
Sure there are 'stock fault lists'.
Thigns that tell you that if it
starts up with this sort of image on the monitor then you need ot change
that chip. 99% of the time they're rignt./ 1% of the time they are not.
And at that point, if all you have is the stock fault lsit, you are stuck.
They're great for "probabilistically" suggesting where to START looking,
Of course. If you know that 99% of the time the fault you're seing is
caused by a particular component, you do start by checking that :-). But
as they don't work _all_ the time, I prefer to teach people the
absolutely genral methods of logically tracing faults. When you can do
that, you know enouyh to find the stock fault lists, and waht to do when
they don't help.
but one still needs to master the tool usage, AND
understand enough theory
to be able to confirm the suggested diagnosis.
Sure. AS I said in my very first mesage, in all fault repair (this
doesn't just apply to computers, or even to electroics, IMHO it also
applies to things like medicine), there are 2 distinct stages :
1) Finding the fualt -- how to make and interpret measurements, etc
2) Putting it right -- using the tools to replace the faulty component.
You will ahve to learn how to do both.
But if you are doing it for yourself, if you want
ot fix that machine no
matter what, then you have to learn the 'real' methods of fault
There is no substitute for learning logical processes
It's a great pity that this sort of thing is rarely taught any more. I
meet very few people who can get a set of facts and reason logically from