Digital Multimeters for Dummies

[Martin Czech]

> Subject:  Digital Multimeter Basics / Questions

I'm in the same situation right now.
 
> The Count:  The higher the count, the better because this is the number
> of samples used to calculate the measurement.  True or false?

The data sheets spec sample time, usually about 0.4 sec, the period is usall
an integer of the mains frequency, thus the ADC has a zero there.
If this the case I don't worry about sampling time.

The other thing is number of digits. Today 20000 or even 50000 digits
can be afforded with resonable price. About 200 DM.

> 
> True RMS:  True root mean square.  If it doesn’t say true RMS, it is
> making measurements how?  Is this cheap alternative measurement
> system really that awful?

If you have a scope, you don't need true RMS. I think that RMS is more
important for power measurement ect., ac measurements are very inaccurate
with most affordable gear, I'd say you need a good scope anyway, in the
long run, so ac capabilitys are not important. I shure wouldn't pay for
true RMS.

> 
> Percent Accuracy:  What is completely unacceptable for high quality
> amateur work?

Well, look under number of digits. With 20000 you can read 20000 or 2000,
before switching into a new range. Therefore the maximum error at full
scale should not exceed 0.1%. Otherwise 20000 digits are ridicolous.

> 
> Certificate of Calibration:  Do I need this if I am not designing circuits
> for the military or a medical life-saving equipment?  Are multimeters
> so poor that one should pay extra to guarantee that the multimeter
> does what the maker claims... and then take the maker’s word on it?

Well, the problem is that the multimeter ADC chips are quite accurate.
It is up to the manufacturer to make a good board design, *and* use
a high quality resistive divider for the ranging. And this is the 
problem. Some manufacturers try to fool people with 50000 digits,
but the divider network is crap, only 0.5% or so. I'm really thinking about
buying such a desktop model, and then simply trim out the error
by hand (adding good trim pots).
OTOH there's reaoable quality for hand held stuff.
> 

> Autoranging vs. Manual Setting:  I assume that setting the initial 
> comparison, either automatically or manually, ensures a closer
> measurement and reduces error.  But, what is the advantage
> of one versus the other... or is it the same difference as automatic
> versus manual transmission in cars?

On one hand autoranging could be nice, if only one hand is free to measure.
OTOH I don't like automata to try to think for me. Because what they think
is garbage. If you have a noisy signal, ar slowly moving signal, you
perhaps get no reading at all. Autorange is fine, but you should be able to 
switch it off. If not, get a normal.


> Frequency:  Is this really a useful function on a handheld multimeter
> or I am a better off matching pitch using my Boss guitar tuner?  Or,
> is this function useful in a completely different way?

Most multimeters have it. They have a quartz clock inside,
so the reading should be quite accurate. Only usefull, if you
get enough digits, however. If the gate time is 1s and you measure
10 Hz, you can read 9 10 or 11 Hz. 10% error. But this function is included
in most models. Don't think about it. 

> 3 1/2 Digit:  Does the half digit at the end mean that measurements
> are rounded to one decimal place?  I saw a Fluke with 3 3/4 digits.
> What does the 3/4 digit mean?

No. A digit means 0..9. so three  digits mean 000 ... 999.
3 1/2 digits mean 0000 ... 1999. 3 3/4  could mean 0000 4999 or so.
This is advetising garbage. The data sheet should spec the number of
digits, e.g. 1000 for the first example 000 ... 999. 


> Two general questions...
> 
> Multimeters, even those with the same features (accuracy, functions, 
> etc...), seem to range wildly in price.  Are you paying for a real
> difference in quality or are you paying extra for a brand name
> that doesn’t really mean anything?

Ok, a Fluke desktop will cost allways double than a chinese thing.
It is important to have a data sheet. Write down your needs,
and make a cross if it is fullfilled with this instrument...


> 
> Is it possible for a low cost multimeter to really be good at a large
> number of functions, from DC to AC, to transistor testing, to
> resistance, capacitance, and frequency?  Or, is it the case with
> multimeters that the do-it-all unit is a “jack of all trades and 
> master of none?”
> 

Yes.
As I said, you can get 
0.1% + 2 or 3 digit error with 200mV to 200V range dc
0.2% + 5      digit error for 100 Ohm to 10 Meg Ohm
1% + 3        digit error for 100pF to 1uF Capacitance

for about 200 DM (handheld model, all from my head)

This is all you need. 

The rest of functions is nice, but not required.


m.c.