About a week ago someone asked about comparing the accuracy of clamp ammeters vs. in-line measurement. Having started my own comparison I thought that I'd expand on my testing. I compared the accuracy of Mastech MS2010B/32-800, Uni-T UT210E, Fluke 337 and Tenma 72-555 clamp ammeters/current sample to a Agilent 34401A, Mastech MS8218, and Uni-T UT61E multi-meters. The Fluke 337 is made more for higher currents than were involved in my analysis and the Tenma unit is an economy 1 mV/A sample.
The resistance, DC voltage and DC current tests with measurements under one amp were fairly easy to make. I used multiple voltage standards with mostly 0.01% accuracy resistors to serve as voltage dividers and for current limiting. For myself, the accuracy of an ammeter measuring a constant current source is meaningless for 95% of my current measurement needs. I'm much more often concerned about what, for at least the short term, is a fixed voltage which is feeding a fixed load. So, all of my current tests involve using an accurate regulated voltage with the current determined by a mostly 0.01% resistor. Therefore, any shunt resistor in the ammeter circuit will act as a series resistance and it will reduce the accuracy of the reading. That's what frequently occurs when I use my meter! What's the purpose of eliminating the effect of that shunt resistance during an meter evaluation unless you only plan on making constant current measurements?
Attached is my meter comparison report. I tried to list all of the details including the cumulative accuracy of all of the equipment involved in each test. I abbreviate some things in a way which makes sense to me, but may leave others wondering at first. I left off some redundant tests (e.g. I have multiple 0.01% resistors of many values) and the capacitance measurement accuracy tests to keep the report to two pages.
To verify if the meter meets its specifications for current measurement, after showing the accuracy of the meter's displayed value, I take the effect of the meter's shunt resistance into account and also show its "corrected accuracy." Ideally an ammeter should work accurately regardless of how you use it. The test procedure can seriously effect the actual accuracy of the reading. For example, during the 5 mA test using the Agilent 34401A meter I used an accurate 5 volt source with the current being limited by a 1,000 ohm 0.01% resistor. Due to the meter's 5 ohm shunt resistance the meter's displayed value accuracy was about 0.5% low. Not having a precise 25 volt source, for the 25 mA test I used an accurate 2.5 volt source being limited by a 100 ohm 0.01% resistor. Now that 34401A's same shunt resistance was causing the displayed reading to be about 5% low! However, both of those readings had a corrected accuracy of 0.015% or under. When you see similar jumps in the accuracy of the displayed value it's due to similar changes during testing.
Because of errors like those reported above it's possible for the low price UT210E to have better display accuracy than the Agilent 34401A. But, measuring DC current with a clamp meter accurately isn't a simple task. Because of that the UT210E doesn't have stable zeroing in the DC current mode. So, you have to live with that problem when you are using it. It's easier and less expensive to make an accurate in-line ammeter, but the clamp version has its obvious advantages at times. There are many other and more expensive clamp ammeters available.
AC and higher current tests are harder to accurately make. However, I think that I'm at least meeting my shown testing accuracy. The fact that most of the corrected accuracies meet the manufacturer's specifications somewhat supports that.
Home
Help
Search
Login
Register
