[R-390] Re: R390 panel meter resistance

[Roy Morgan]

At 08:16 AM 5/6/03 -0700, W. Li wrote:
>A few days ago, a msg appeared about measuring the internal resistance
>of a R390 panel meter,

That was me.

>Now, if you think about it, this time-honored method results in false
>results, because by adding the shunt resistor across the meter, you
>actually LOWER the total resistance of this simple series circuit....
>and thus INCREASE the total current flow, thus falsely altering the
>deflection of your R390 meter.

Quite true. Lets take an example.

1.2 volt AA cell, normal R-390A Carrier meter.
Quoting "Paul H. Anderson" <pha@pdq.com>:
"R-390/R-391/R-390A carrier meters are 1 MA FS, 17 ohm, as far as I know."

So for one ma current with 1.2 volt supply we need a total circuit 
resistance of:
R = EI = 1.2 x .001 = 1200ohms.
The fixed (or adjustable) resistor then needs to be 1200 - 17 ohms or 1183 
ohms.

As a first approximation, when the 17 ohm meter is paralleled with a 
resistance of equal value the resulting resistance will be 8.5 ohms, and 
the total circuit resistance will be 1183 + 8.5 = 1191.5 ohms.  This will 
in fact increase the deflection of the meter by a factor approximating 1200 
/ 1191.5, or about 1.007.  That is seven tenths of one percent. This is tiny.

I say "first approximation" because the method asks you to set the parallel 
resistor for half scale deflection. Thus the resistance you set will be 
just enough smaller than the resistance of the meter to compensate for the 
tiny increase in circuit current due to the overall reduction of total 
circuit resistance.

>A fresh battery is really a constant
>voltage device, and not a constant current device.

I agree.  The battery is unlikely to change voltage much due to either 
discharge or change in circuit current.  However, the circuit described 
above is approximately a source of constant current, at least to less than 
one percent change.

>One workaround is to keep this series test circuit current constant.
>This method involves a second mA meter. Insert any low mA meter (say a 2
>mA one) into the circuit, then connect your R-390 panel meter IN SERIES
>with it and your battery and pot. Now adjust the pot to get full
>deflection of the R390 meter, and note the reading on your 2 mA meter.
>Remove your R390 meter, and substitute a 10-turn precision 500 or 5000
>ohm pot, and adjust it alone to get the ORIGINAL READING on your 2 mA
>meter. Now the resistance of the 10-turn pot is exactly the same
>resistance as the internal resistance of the R390 meter, since under
>this scheme the test circuit's current was kept constant.

This is a basic method of substitution, and is all very fine, but:

(I am assuming the circuit is similar to above with a 1.2 volt source and a 
limiting resistor.)

The change in circuit current with the R-390 meter replaced with a short 
will be approximately 17/1200 of the original amount.  By shorting the 
R-390 meter, the 2 ma meter's needle will deflect approximately 1.5 percent 
higher.  If the R-390 meter is replaced with approximately twice it's 
internal resistance (34 ohms) the 2 ma meter will deflect about 1.5 percent 
less.  Determining the resistance that causes the 2 ma meter to return to 
it's original deflection will be quite difficult.  Mechanical sticking will 
likely make it impossible.  A digital multimeter with 4 or more digits in 
current mode would make it quite easy to do.

If a higher source voltage is chosen with correspondingly higher series 
resistance, the changes in mechanical deflection of the 2 ma meter as the 
substitute resistor is adjusted will be proportionately less.  A 9 volt 
battery, for instance, and 9 K resistor will cause the change in deflection 
to about one seventh as much, or 0.2 percent.

>I'd like to take credit, but this subject was already discussed in some
>detail in the 73 Test Equipment Library volume 1 (1976) pp33-35.

Here are some factors to consider as you ponder the fine points of the 
method that the 73 Test Equipment Library suggests:

1) Many meters normally used by Hams have higher internal resistance than 
the R-390 series meters.  The 73 Test Equipment Library method was likely 
developed with these higher resistance meters in mind.

2) As the the internal resistance of a meter becomes a higher portion of 
the overall circuit resistance  (as with meters of higher resistance), 
determining the value if the substitute resistor gets easier.

3) Lower supply voltage will have the same effect.  The meter resistance 
and the substitute resistor value, will be greater with respect to the 
overall circuit resistance.

4) As the supply voltage gets higher, the current source more nearly 
approximates a perfect source, and is more immune to small changes in 
circuit resistance.  This will make the 73 Test Equipment Library method 
more difficult to use without digital current meters.

5) Many meters, even good ones, have mechanical hysteresis and resistance 
that would hide changed deflection due to small changes in current.

6) The width of a graduation line in a normal panel meter may be about one 
percent or less of the total scale length.  The width of an R-390 Carrier 
meter graduation is about 1/40th of the scale length or about 2 
percent.  The needle is about the same width in some meters.

7) Most R-390 type meters cannot be set to mechanical zero, thus making it 
difficult to set mid-scale if using the "traditional" method.

Roy

- Roy Morgan, K1LKY since 1959 - Keep 'em Glowing!
7130 Panorama Drive, Derwood MD 20855
Home: 301-330-8828 Work: Voice: 301-975-3254,  Fax: 301-948-6213
roy.morgan@nist.gov --