DMM counts

[borgunit]

Hi,
While maybe simple, I do not understand completely the impact of a DMM rated number of counts. I am trying to choose between two multimeters. One has 4000 digital counts and the other 6000. What would be the real accuracy difference between these two?

"Everybody is ignorant, only on different subjects." — Will Rogers

 

[VEBill]

Not necessarily accuracy, but resolution.

The maximum count is the maximum display reading. Being able to use a lower scale (because of the higher limit of the display), means that you might be able to get better resolution in certain ranges.

It's funny because when DMMs first came out they had displays with normal digits (0-9). Then someone invented the leading "1" 'half digit', based on the overflow bit. Now we're up to 4000 and 6000. Pretty soon they'll reinvent the leading full (0-9) digit. ;-)

 

[borgunit]

Let me see if I am getting this. Say, I am measuring AC volts and the meter will say 120.20, with a 4000 count DMM, it would be able to show down to .03 (120/4000) while the 6000 count DMM could show down to .02 (120/6000). Changes in voltage would only be reflected when the meter detects its "count" accuracy (.03 or .02 volt change).

"Everybody is ignorant, only on different subjects." — Will Rogers

 

[VEBill]

That's not my understanding.

I understood that 'counts' are (to the user) simply the display range. Of course there is circuitry behind the display but we can ignore that for now.

For your 120.20 Vac example, both meters would display "120.2" - four digits used optimally on both meters.

For a 5.4321 Vdc input, the 4000 count meter would display " 5.43" - four digits but the first one can't handle the '5' and thus is forced to be blank. The 6000 count meter would display "5.432".

*** Standby for confirmation or rebuttal from others because my understanding on this topic is based on assumptions and I may be incorrect. ***

 

[IRstuff]

Note that accuracy is usually given as a percentage of full scale/percentage of reading or both, plus a count error.

TTFN

 

[logbook]

It is only in the transition region between two ranges that you get a significant difference.

Suppose one meter is 4.000V full scale and the other is 6.000V full scale. At 5V the first meter would need to be run on the 40V scale so you are running at about 10% of full scale. The second meter would be running at 83% of FS. A DVM is always more accurate when running at full range (or full scale, depending on where it was calibrated). One issue is loss of resolution, the other is loss of accuracy. This range issue may be important if you are working with 5V power supplies.

On really expensive long scale DVMs the distinction is critical. These meters have an input impedance of >10G ohms on the 10V range and below. With >50% over-range capability you get this high input impedance up to at least 15V, covering measurements in ordinary opamp circuitry. The accuracy is also degraded on the 100V range due to the accuracy of the resistive input attenuator.