Dave Plowman (News) wrote:
In article ,
alan wrote:
On 03/01/2013 11:05, Dave Plowman (News) wrote:
In article ,
alan wrote:
As long as the connecting leads can handle the current and are "low"
resistance, the resistance of the leads makes no difference to the
measurement.
Swapping leads on my various DVMs - using what appears to be decent
quality leads - will give different resistance readings for very low
values. Try it.


It doesn't matter if the measuring leads are 0.01ohms or 1 ohm when
measuring the _Voltage_ across a shunt resistor with a DVM (or
voltmeter) with a high impedance input. There may be 100s of Amps
flowing through the shunt resistor but microAmps in the measuring leads.

But those connecting leads I was referring to will make a difference to
the results. In the same way as test leads do when measuring very low
resistance. As when doing that you're actually measuring voltage drop
across a resistance.

If you're measuring the voltage drop due to 100 amps across a few
milliohms as in the case proposed, then the error due to probe
resistance should be negligible, and well within the tolerances of
anything outside a standards lab. Assuming 2kpv for the meter, (500
microamp/ volt FSD), the meter impedance will be on the order of dozens
of ohms, with a current of less than 100 microamps. If the meter is the
now standard? (Electronic multimeter) 10 Megohm input impedance, then
the current in the leads will be hundredths of a microamp. The
resistance of the leads and their connection to the shunt should be
hundredths of an ohm at most, given clean contacts and a steady hand.
The major errors are going to be bad contact between the probes and the
shunt, thermal noise in the shunt, thermal noise in the leads ands
meter, and induced currents from nearby AC circuitry, especially stuff
like switch mode supplies, which can generate some very strange induced
waveforms.

The biggest error in this measurement will probably come from the change
in resistance of the circuit being measured by adding and removing the
shunt, unless the shunt is permanently in circuit, with a proper meter
connection for it. In this case only, the effect of the meter being
connected or not on the circuit being measured will be minimal. Now, how
accurate is the meter, and will these effects be swamped by meter error?

One way round this problem is to use the supply lead as the shunt
resistor, and calibrate the voltmeter accordingly.

--
Tciao for Now!

John.