Quote:
Originally Posted by Superscope
Most of the better Meters I've seen so far are 1% +- 2Digits.
This means with a resolution of 0.1uA I could be more than 0.5uA out,
or have I missed something?
That is not really accurate enough for the Job!
|
The accuracy figures can be quite misleading, depending on how they are applied. Also the tolerances for A, mA, uA, V, mV, AC, DC etc. can all be different on the same meter!
In your example 1% +/-2 digits, presumably on a 3.5 digit (1999 counts) dmm on V DC range.
The 1% refers to the accuracy of the measured value, so if you were indicating 100V on the meter then the accuracy would be 100V +/- 1V, so displaying 100V you could have 99V to 101V at the probes.
The +/-2 digits is the tolerance of the last significant displayed digit, so its value will depend upon the number of digits on the meter.
So for a 3.5 digit meter the display for 100V input on the 200V range would be 100.0V. The least significant digit is 10ths of a volt (100mV) so +/- 2 digits would give an additional error of +/- 0.2V. The total tolerance will then be 1V + 0.2V or +/- 1.2V
So for a displayed value of 100.0V the final accuracy would be from 98.8V to 101.2V.
If you have a 3.5 digit meter on the 200uA range, measuring 37.5uA, then:-
1% of 37.5uA = +/- 0.37uA
2 counts = +/- 0.2uA
Total tolerance will be +/- 0.57uA (just as you thought 
)
Therefore accuracy when displaying 37.5uA is from 36.9uA to 38.07uA
That would be for a quality meter that has been calibrated and that meets its claimed accuracy, such as HP, Fluke, Keithley etc.
1% accuracy is pretty poor, and some of my vintage 4.5 digit DMM's claim "+/-0.1% +4 counts" (4 counts on 4.5 digit meter is 10x better than 4 counts on 3.5 digit meter)
For example a fluke 8600 DMM (around £40) claims for DC current +/- 0.1% input reading + 0.01% of range (which is 2 counts on a 4.5 digit meter (19999 counts)!) (for voltage it is +/- 0.02V of input + 0.005% of range, which is not bad for a 30 year old design
)
so 37.5uA gives an accuracy of +/- 0.037uA + 0.02uA = +/- 0.039uA or 37.44uA to 37.55uA.
Significantly better than the 1% 3.5 digit meter.
However all this assumes the meter is calibrated to this level!
However if the meter is not calibrated, you can still use it as a "transfare reference" by measuring a known good value on a good AVO, and then adjusting your target AVO's voltage to exactly the same displayed value due to "Precision".