Don Bruder wrote:

(snip)

Problem is, I've never managed to wrap my head around the "how-to"

involved in changing from one to the other. Every explanation of the

idea I've encountered leaves me scratching my head.

(snip)

Meters that need only a small voltage to read full scale and also need

only a small current to read full scale are well suited to be made

into a volt-amp meter. The trick is to arrange with external

resistors to have the full scale requirement to be met with the

desired full scale measurement.

Lets say you have a magnetic movement that hits full scale at 1

milliamp of current and has 10 ohms of internal resistance. Ohm's law

relates voltage, resistance and current (Resistance = volts / amps) so

when 1 milliamp is passing through that internal 10 ohms, there will

be a 0.01 volt drop across the meter.

To arrange for this meter to be a volt meter that reads full scale at

20 volts you need to add more resistance in series so that only at

that voltage will the 1 milliamp full scale current pass through the

meter. Since that occurs with 20,000 ohms (ohms law again), total and

the meter already has 10 ohms, you would need to add another 19,990

ohms in series. (Resistor in series simply add up)

If, instead you wanted to make the meter read full scale for 10

amperes, you would need to parallel it with a big, low value

resistance, such that the parallel combination dropped the full scale

voltage of 0.01 volt when 10 amps passed through both the resistor

(mostly through the resistor) and the meter's 10 ohms. 0.01 volt/10

amps = 0.001 ohm. When working with parallel resistors, it is easier

to convert them to conductances (1/resistance), since conductances in

parallel simply add up. Then convert the effective conductance back

to resistance with a 1/conductance calculation.

So 1/10 ohms + 1/shunt resistance (that carries most of the 10 amps) =

1/ total resistance

So the shunt resistance needs to be about... well crap, it is so close

to .001 ohm ohm that the meter's 10 ohms makes almost no difference.

You leave the .001 ohm shunt in series with the current, permanently,

and switch the meter to either be connected across the to be measured

voltage with a 19,990 ohm resistor in series, or connected directly

across the shunt resistor. The shunt will lower the voltage to

whatever load is in series with it by 0.01 volt when 10 amperes is

passing through.

Now back to your meter. If it already has the equivalent of the

19,990 ohm resistor already inside the case, you will not be able to

use it without doing surgery on it, since this resistor cannot be in

series when you make current measurements. You need ot start with a

simple meter movement with both a low current full scale requirement

(so the voltage dropping series resistor will not get so hot) and a

resistance (equivalent ot saying that it has a low full scale voltage

requirement) because that full scale voltage is what the shunt must

drop top measure current.

By the way, the shunt resistor may be able to be a piece or wire that

is already involved in getting current ot the load. If you can pick

two spots on it that have the desired full scale drop you can just

connect two small wires to those points and take them back to the

meter. The difficulty comes in knowing where those two points are.