Compounding Errors: Combining Accuracies of Fluke Meter & Clamp

[elimenohpee182]

Hey Guys,
I'm not sure the correct procedure for comounding errors using some test equipment. I'm trying to measure electrical current using a fluke meter and a fluke current clamp. Here are the two devices I'm using specifically:

Fluke Power Analyzer Model 437 Series 2
Fluke i5sPQ3 AC Current Clamp

Here is where it is confusing. The Fluke meter has an accuracy spec as well as the current clamp. For instance, the meter says in the range 0.5 - 200A, the accuracy is 0.5% + 5 counts. The clamp says the accuracy is 1% + 5mA in the range of 10mA - 1A.

What is the appropriate way to combine these two accuracies to establish a single accuracy? I would think to add them, but I've seen some people that seem to take an average, using a RMS-type formula.

Any suggestions?

 

[Baluncore]

Meter ±0.5% ±5 counts.
Clamp ±1% ±5mA, in the range of 10mA to 1A.

Combined ±1.5%, ±5 counts, ±5mA, over the range of 10mA to 1A.

 

[jack action]

Say you have 1.0 A on your meter. The meter accuracy says that the actual current measured is within the largest range between ±0.5% (0.995-1.005 A) and ±5 counts (0.5-1.5 A); 0.5-1.5 A it is.

Then the current sent by the clamp (which is anything between 0.5 A and 1.5 A) could have been within the largest range between ±1% (0.495-1.515 A) and ±5mA (0.495-1.505 A); Hence the actual current of your 1.0 A measurement could be anywhere between 0.495-1.515 A.

Doing the same procedure with a 0.5 A measurement yields:

meter:

• ±0.5%: 0.4975-0.5025 A
• ±5 counts: 0.0-1.0 A (largest)

clamp:

• ±1%: 0.0-1.01 A
• ±5mA: (-0.005)-1.005 A

Possible actual current measured: (-0.005) - 1.01 A

If you used your clamp with the 1-5 A range with a 1% accuracy, then with a 5.0 A measurement, the actual current range is:

meter:

• ±0.5%: 4.975-5.025 A
• ±5 counts: 4.5-5.5 A (largest)

clamp:

• ±1%: 4.455-5.555 A

Basically, the ±5 counts on the 0.1 A resolution of your meter is killing the accuracy of your clamp.

 

[dlgoff]

... the meter says in the range 0.5 - 200A, the accuracy is 0.5% + 5 counts. The clamp says the accuracy is 1% + 5mA in the range of 10mA - 1A.

When using these accuracies to determine an uncertainty of a measurement, only comparison with a standard will provide a way of eliminate or compensate for any calibration errors.

 

[Baluncore]

I note that the “Fluke i5sPQ3” AC Current Clamp appears not to be a current ratio transformer, but is specified to produce a voltage output of 400 mV/A. I expect that requires the “Fluke Model 437 Series 2” Power Analyzer then be used as a volt meter on the range up to 5A * 0.4V = 2V FSD.

But the 437-II meter range is specified as Vrms 1V to 1000V phase to neutral, with resolution 0.01 V ± 0.1% of nominal voltage.
for 0.5A * 0.4V/A = 0.2V, 0.01V is 20:1 = 5%
for 1.0A * 0.4V/A = 0.4V, 0.01V is 40:1 = 2.5%
for 2.0A * 0.4V/A = 0.8V, 0.01V is 80:1 = 1.25%
for 5.0A * 0.4V/A = 2.0V, 0.01V is 200:1 = 0.5%
So the limitation is not always the 1% accuracy of the clamp, but is the resolution of the voltmeter.

To get more accurate results it would seem necessary to use a higher resolution voltmeter, change to a current transformer clamp, or use a voltage amplifier between the clamp and analyser.