uncertainty estimation for a DMM voltage measurement

My DMM has a accuracy of 0.0030% of reading + 0.0030% of range, when the voltage range is set to 100mV. If a measurement with this DMM gives 0.1mV, then is the true input voltage between 0.1mV - 3.003uV and 0.1mV + 3.003uV? Or is there a different factor than the uncertainty that I have to account for?

Your calculation is correct but the result is not really the uncertainty of your measurement, it's only the maximum instrument error.

If you want to calculate the uncertainty following the ISO standard you need to go a bit further. In this document you have a very good example:

Fluke DMM Uncertainty Calculation

If you follow the example presented with your data you'll get to the uncertainty values you're looking for. Give it a go and post your results, then we can let you know if your answer is correct.

If accuracy is ±0.0030% of a 100mV range, ±0.0030% * 100 mV = ±0.00003 * 100 mV= ± 3 μV

So if your example was 0.1 mV = 100 μV ± 3 μV

• Accuracy can be specified differently as a % of reading with a number of counts.
• It can be defined in terms of each part of error; gain, offset, non-linearity
• normally examples like yours are expressed in parts per million, ppm

Features             34460A  34461A  34465Anew  34470Anew
Digits of resolution 6½      6½      6½         7½
1 year DCV accuracy  75 ppm  35 ppm  30 ppm     16 ppm

• low cost DMM's might be specified in terms of ±% of reading and ±counts

• Could you please elaborate the 1e3 term in "±0.003/100 *1e3 = ± 0.03 μV"? I guess it's 1mV because the left hand side is dimensionless now – Huisman May 27 at 6:21
• unit conversion %mV to μV. software calculators accept shorthand 1e3 for powers of 10 – Tony Stewart Sunnyskyguy EE75 May 27 at 6:24
• I think 1e3 should be 1e-3? And next, please elaborate how "a 100mV range" equals to "1e-3". It should at least have dimensions. – Huisman May 27 at 6:33