> From:    Larry <[log in to unmask]>
> > Digital meters tend to be more precise than analog.
>
> Perhaps I'm just in a nit-picky mood or just want to argue but
> I disagree. Analog meters are just as precise as digital, if not more so.

Hey, no flames.  Analog meters can be just as precise as digital - but I
think you will find that the better analog meters cost more than the
equivalent digital meters - if only because they need bigger (physical size)
precision resistors and other components.

> I learned the electronics trade on the old Simpson 260 analog multimeter.
> I have used both digital and analog and can see where digital has its
> advantages, but in electronics troubleshooting, analog is better because
> you can watch the changing value. With digital, you watch the numbers change
> too fast to follow, or too slow (sample rate) to know what's happening.
>
> Watching the needle move in exact time with the measured voltage is
> a better troubleshooting aid than simply reading an end result, in some
> situations.

As long as you realize that you may be dealing with variations that are too
fast for the meter to follow, or non-sinusoidal wave-shapes (SCR partial
sine-waves, or square waves).  Of course, if you know what you are
dealing with, you can correct accordingly, depending on whether the meter
is peak-reading (like most simple AC meters that rectify the AC to DC), or
true-RMS (Root Mean Square) such as an "iron vane" AC meter.  AC can get
messy - that's why electricians prefer a good digital true-RMS meter.

> Better digital meters have bar-graph indicators which simulate the
> needle movement of analog meters but the slow sample rate still screws
> up the exact measurement.
>
> Digital meters are for electricians and hobbyists. Analog meters are
> for technicians who seek precise measurements.
>
> Standing by for the flames,
> Larry Hooper

To continue in the spirit of friendly debate, I believe the context was
someone inquiring how to measure the +5 VDC and +12 VDC on a power supply.
Let's suppose we have two nearly equivalent meters:

A true RMS digital meter rated at +/- 0.25 % FS, +/- 1 bit, 20 MOhm input
impedance, 15 pF input capacitance.  "3 and a half digits".

A 20,000 Ohm/Volt analog VOM (Volt-Ohmmeter) using the DC Volts scales.
Also rated at +/- 0.25 % FS, and by an incredible coincidence, having the
same scale ranges, on a 4.5 inch scale.

Because we are talking DC, the true-RMS feature is a useless frill.  On
the 10 Volt scale +/- 0.25 % FS (Full Scale) corresponds to +/- 0.025 VDC
The digital meter reads up to 1.999 or 19.99, depending on the scale, so
this is an additional +/-0.01 VDC on the 10 V scale for a total error of
+/- 0.035 V total error.

The analog meter has the same +/- 0.025 VDC component error.  If the scale
has 10 major divisions (roughly 0.7 inch each) and these are subdivided in
10 subdivisions equivalent to 0.1 VDC each.  Estimating the subdivisions to
1/5 of a subdivision is reasonable +/- 0.02 VDC, for a total error of +/-
0.045 VDC.

As you can see, the digital meter is slightly more precise, and easier to
obtain that precision.  Also, because we are measuring a power supply, the
circuit impedance is too low to affect the measurements.

This is not always the case.  Suppose we have two 2.0 KOhm resistors in
series, with 12 VDC across them.  It's obvious that the voltage across each
is, ideally, 6.0 VDC.  At 20,000 Ohms/Volt, on the 10 Volt scale, the analog
meter will have a resistance of 200 KOhms.  When measuring across one
resistor, this is in parallel with the 2.0 KOhm resistor, for an effective
resistance of about 1980 Ohms in series with the other 2.0 KOhm resistor.
This results in a reduction to 5.969 Volts when measured across either
resistor.  The digital meter, with an input impedance of 20 MOhm, results
in a measurement of 5.9997 Volts.

Of course, you could reverse the above calculations with the analog meter
to come up with the voltage across a resistor when the voltage is not
"pulled down" by the voltmeter - but it's a pain.

Boyd Ramsay

[log in to unmask]

         The NOSPIN Group Promotions is now offering the NOSPIN
        File Download CD, Abit's Gentus Linux, Linux Power CD and
      the RedHat Linux CD.  All CDs are always provided at COST!!!
                      http://nospin.com/promotions