# Naming 6000 counts multimeter in digit format

A 4-digit multimeter is also called 9999 counts; and it is called 4-digit because all four digits can display numbers from 0 to 9.

A 3½ digit multimeter is called 1999 counts; and it is called 3½ digit because three digits can display numbers from 0 to 9 and the most significant digit can display only 0 or 1.

This multimeter on the other hand is described in its specs as 6000 counts. I guess this means the most significant digit can display a number from zero to 6 or to -6 (?). But in this case we cannot call it 4-digit because its most significant digit cannot display numbers 7, 8 or 9. We cannot call it 3½ digit because its most significant digit can display numbers 2, 3, 4, 5 and 6. Does that mean this multimeter cannot be named as "... digits"?

• If we used a proper logarithmic scale, a 1999 count meter would be 3.3 digits, and 5999 count meter would be 3.85 digits. Commented Mar 25, 2020 at 16:25

It’s a silly approximation to computing base-10 logarithms anyway, maybe from the times when having logarithms on a calculator was uncommon, the log tables were constantly misplaced and the sliding rule was borrowed by the kid for school that day.

A 1 isn’t a “half” of a decade (1 digit=1 decade). A 5 is a half-digit and half-decade.

A multimeter with 1,999 counts has 3.3 digits. One with 9,999 counts has 4.0 digits. One with 6,000 counts has 3.8 digits.

Since 3.3 digits is routinely “rounded up” to 3.5 digits, then 3.8 digits may as well be called 4 digits, or “4 digits but not quite”.

There’s also a historically-relevant convention that the “counts” and by extension “digits” refer to the unipolar measurement range. That’s because the earliest implementations of digital multimeters, and many still today, don’t include the sign within the counter circuit proper. It’s determined separately, not from the counter output. Thus, in the most literal sense, counts exclude the sign.

That’s why we use counts when we want to be clear, because the “digits” have been ruined by the expectation that they are rounded to a single binary digit past the decimal point. Or, as some people seem to think, by the expectation that it’s a mixed-base number: decimal in the integer part, binary in the fractional part. Yeah, it makes as little sense as it sounds. Logarithms are not time. No need for splitting bases. Someone tried to be too clever in marketingspeak at one point and messed it up for all forever after…

The "3½ digit" term is very informal and doesn't have a precise definition. I've seen "3¾ digit" used to refer to 6000-count meters, but I've also seen it for 4000-count meters. Generally it's preferable to refer to it as simply a "6000-count meter". "3½ digit meter" is fine to refer to a 2000-count meter, but "3¾ digit meter" is too ambiguous.

Incidentally, I would generally call a 3½ digit meter a 2000-count meter, since it can display anything between 0 and 1999 (a total of 2000 different numbers, if you include zero). Similarly, a 6000 count meter can display between 0 and 5999; they don't generally display 6000 as far as I know.

• But look at this first page it can display number 6 alliedelec.com/m/d/1ff95544fa5f6490684ba59eb0f7fd7d.pdf How about -6? Can it display that as well? Commented Mar 25, 2020 at 16:26
• Probably! like I said, this is just as far as I know. That said, the fluke meter shown there probably has the display photoshopped on; they always do that for promotional photographs. Determining whether it can actually display 6000 or not would be more effort than it's really worth, honestly. Commented Mar 25, 2020 at 16:28
• My geetar amp goes up to 11 on the overdrive control. Commented Mar 25, 2020 at 16:49
• @user1245 Of more significance is its accuracy (best case:0.5%). The least-significant-digit is mostly useless for many readings...so it might be viewed as a 600-count meter. Having many digits is good where you compare one reading with another soon after. Commented Mar 25, 2020 at 16:50
• @glen_geek There's a difference between accuracy and precision. The fluke 114 that user1245 linked doesn't have a precision rating, but it's probably a bit better than its accuracy. Of course, making a handheld meter too precise/accurate is a bit of a waste of effort, given that they're generally used with unshielded leads in very noisy environments!... come to think of it, I don't think I've ever seen shielded multimeter probes. Maybe that wouldn't be as useful as I'm thinking. Commented Mar 25, 2020 at 16:54