Groups	Advanced Groups Search    Preferences    Groups Help
	"why" is a very common word and was not included in your search. [details]

>I need to measure a DC voltage of less than 12 Volts at greater accuracy 
>than my meter provides.  I need to measure the voltage at 0.1%, while my 
>meter is only 1.0% accurate.  Is there a trick that I can use to do this?
>
>Thanks
>Jim

While other followup posts suggested averaging readings and/or comparing against
a reference voltage (a transfer standard) I think you might want to 
consider the following: there are two components to equipment accuracy.
The first is bias error which is essentially the difference between the
true value (as measured or produced by a more accurate system) and the
average of your readings. Bias error cannot be determined or corrected for
without access to another reference or measurement system. This is the 
error that calibration facilities can correct for. For example, if you measured
voltage 1000 times and the average value was 11.5 V but the true reference
value is 12.000V, you would have a 0.5V bias error. And you will not know this
by averaging. Precision error is essentially the spread (OK standard deviation
for you statistic folks) of readings around the average value. Precision error
is a function of the measurement equipment, measurement environment (i.e. noise)
etc. and cannot be calibrated out, however, it can be quantified through 
empirical measurements....
So, having said all that, you need a meter that has resolution and accuracy
of 0.1% of 12V (11.889V - 12.012V) which implies 1mv resolution on the appropriate scale (say 20V) or a 4 1/2 digit meter (minimum). That will quickly limit your
selection of meters. Even at that, be careful since accuracy specifications usually
include percent of reading + xx digits and the xx digits can dominate on lower 
ranges. For example, my Fluke 8050A has 1mv resolution on the 20V range and
for DC voltage is specified as 0.03% of reading + 2 digits. The 2 digits creates
a 0.02% error when reading 12V. I have seen meter specifications with great % of
reading accuracy but +20 counts.....
At 0.1% accuracy requirement, I would suggest finding a very good 5 1/2 digit
meter.


  -John