Multi meter and clamp meter accuracy ±(% of reading + number of LSD)

[dylanmitchell]

Looking at multi meter and clamp meter specs I see ±(% of reading + number of LSD). I understand +/- percent but don't understand LSD or least significant digit. Please explain how this works. For example:

Fluke 117 AC volt
 ±([% of reading] + [counts]): 0.5% + 2
vs
Fluke 87 AC volt
±(0.7% + 2) True RMS

Amprobe 320 vs Fluke 325

AMP320 AC amps
±1.8% + 5LSD (50 to 60 Hz)
vs
Fluke325 AC amps
1.5% ± 5 digits







±(% of reading + number of LSD)

Where:

Reading = the true value of the signal that the DMM measures

LSD = least significant digit

Where:

Reading = the true value of the signal that the DMM measures

LSD = least significant digit

[bdunham7]

Suppose your meter is reading 10.00 volts (or amps) and has a specified uncertainty of 1% + 5 counts (or LSD--same thing as counts).  This means that for the 10.00 indication, the real value might be as low as 9.85 or as much as 10.15.  That's 0.1 volts for the percentage and 0.05 for the counts.  What you don't often see is the confidence interval that goes with that specified uncertainty. 95% is common, so for a 95% confidence interval in this case, that means that there is a 95% probability that the real value is between 9.85 and 10.15.

B/t/w, don't make the mistake of comparing meters from different manufacturers solely on the their specified uncertainties, especially when the numbers are close.  These numbers come from the manufacturer, not any independent testing service.

[Doctorandus_P]

If you've got enough LSD, all those numbers don't mean a thing anymore.

But they move a bit and have nice colors.

[bdunham7]

If you've got enough LSD, all those numbers don't mean a thing anymore.

Or their meanings become extraordinarily clear.

[Doctorandus_P]

That is probably an illusion.