Recommended logic analyzer/scope tools? (was: RL02 write faults,

Tony Duell ard at p850ug1.demon.co.uk
Sun Jan 8 15:08:14 CST 2006


> 
> > Incidetnally, on most DMMs (including my Fluke), the fuse is for the 
> > current ranges only. The voltage and resistance ranges work fine with the 
> > fuse blown. But of course there are plenty of other faults that can cause 
> > a zero reading with a voltage on the input. Heck, even test leads can go 
> > open-circuit.
> 
> Never assume this is the case - I've got a couple of rat-shack DMMs
> where the fuse is simply in the positive meter lead - pop it and you lose
> all functionality.

Right...

And as I said, even if the fuse wouldn't cause the voltage ranges to 
fail, plenty of other things could, starting with the leads themselves. 
NEVER assume the meter is working if your life could depend on it. Check 
it on a known voltage source of the same sort of voltage.

> 
> I never trust a meter to be working until I see it read something I expect
> (I think this comes from the early days when all I had was a rat-shack
> meter which used to get dirty contacts causing failures from time to time).
> 
> First thing I do when "entering" equipment is measure the power supply.

Yep, I always check the PSU (prefereably  on dummy load) first, too. So 
many problems are traced to PSU problems. I once spent a long time doing 
battle with the Unibus logic in a PDP11/45, the fault turned out to be 
that the PSU was sitting at 4.2V( should have been 5V, of course). The 
darn machine would work most of the time, and then do some really odd things.

> This both tells me that the supply is "reasonable" and that my meter is
> working (granted a double-failure could occur which happens to give me
> a incorrect by reasonable reading, however this is not all that likely so I
> usually don't get a second optinion unless I have cause to suspect the
> PS).
> 
> At the beginning of a session measuring resistance, I touch the probes
> together to be sure "zero is zero" - just habits I have gotten into, but they
> have turned up problems on more than one occation.

Having been brought up on analogue meters where you have to set the zero 
on the resistance range by hand, I do this too. Force of habit. 

Just as, if I'm using the continuity beeper in my DMM, I tap the probes 
together when I've selected it (just to check I've pressed the right 
buttons). And then tap them together every so often thereafter, just to 
make sure it's still working.

> 
> 
> 
> > Something I was taught early on (in a different context, but it applies 
> > to electronics too). 'It is much easier to make measurements than to know 
> > what you are measuring'. With most 'socpes, if you touch the probe on a 
> > point in your circuit and twiddle the knobs you will get _some kind of 
> > trace_. It may well mot be any use in actually tracking down the fault, 
> > though. It may not even be a true representation of the signal at that point.
> 
> Not sure if this is a statement for or against, but my point is that by knowing

It's a releated statement, not directly 'for' you, but on that side of 
things, and I agree with what you're saying

My point is that it's very easy to get readings. They may or may not be 
meaningful, they may or may not help with the problem you'r trying to 
solve. Experience (and knowing what you expect) will help you to take 
meaninful readings, and to interpret them correctly.

If you get a reading that disagrees with what you expect, then there are 
2 posibilities. Either you've not measured what you thought you measured, 
or there is a fault. You then need to investigate which (and not to 
assume the second always!). If the rteading agrees with what you expect 
it's likely everything is OK (yes, it could be a fault _and_ a 
mismeasurement, but that's very unlikely).

> what you are looking for, you have a better idea of "if it looks right" or not.
> Many times I found trouble because a signal was present but didn't match
> what I expected. And lots of times I investigated and had a "eureka" moment

Oh sure. Many times there's _something_ on the testpoint, bnt it's not 
what it should be.

-tony




More information about the cctech mailing list