As mentioned above, it kind of depends on what the meter is expected to be measuring.
The 1, 2 and 5 used by Tek works well for scopes because you are looking at a graticule and, at least in the past, always having to do a conversion in your head for the voltage or time represented by the graticule marks. Multiplying by 1, 2 or 5 is pretty simple.? I don't think this reasoning fits analog meters. Tek did make a Type 130 L-C meter and the scales are 1 and 3 lined up at the far right.
Looking at my bench, I have a Heathkit M-25 and an HP 410C, they both use the 1.5 and 5 scales. These are general purpose bench VOMs so not focused on any specific type of measurements. The scales make sense for general purposes. The HP has the 1.5 and 5 lined up and the Heathkit has the 1.5 and 5 offset. I think this is just due to the divider design chosen but could have been the lined up if more precision resistors were used.? Also note the center scale zero on the heathkit. This allowed the user to move the needle to the center for zero and read + and - voltages without reversing leads. Most analog meters cannot do this. The real big meter and the ability to read + and - volts on the heathkit make it more useful on the bench.
???
My Simpson 260 VOM has 1, 2.5 and 5 scales. I think these made sense for the era they were produced, mostly tubes.
My Triplet 630 VOM has a general purpose set of scales; 1.2, 1.5, 3 and 6. Judging by the scales, it is not focused on the tube era. Note filaments were typically 6.3VAC and 12.6VAC which would be off scale on the closest range, but putting them mid scale might be a good compromise. This would have been better if the meter was a little extended to 13V and 7V full scale for tube era measurements on the 12 and 6 scales.
I have several HP AC/DB meters; 3400A, 400A, 400F and 412A AC millivoltmeter. The scales are setup for AC and DB measurements. These meters are 1 and 3.16 full scale. These ranges make sense for AC or DB measurements. In this case you are often thinking about values in dB. In that case you want to the meter to increment scales in 10db step for simplicity in calculations.? My HP 410A has 1 and 3 (though lined up for 1 and 3.16 like other AC meters) scales. While this is a bench VTVM, it was focused on both audio (down to 20hz) and high frequency AC (up to 700Mhz) as a main selling point.
I also have a GW Instek 4278 dual needle/channel AC millivoltmeter. It think it has very useful meter scales. 1.1 and 3.5.? The problem with the 1 and 3.16 scales is that there is no room above to get a measurement that is just beyond the 1 or 3.16. Having a little more scale when you are adjusting something around these numbers means you don't need to move up to the next scale, putting the needle back to the left side of the meter and in a harder to read section of the scale. Also having two channels with dual needles is really useful in stereo design/repair and when designing amplifier stages. I wish the meter was about an inch to two inches wider, like my Heathkit, for a little easier to estimate between the minor divisions.?
Hopefully this shows some options and reasons for picking certain scales.?
If I was designing a new "precision" bench analog meter, I would also consider adding range expansion. For example if the indication was near the center of the meter such as 5 on a 10 scale for example, I would like the option to amplify and shift the value to expand the full scale of the needle to indicate something like 2.5V and 7.5V, or 4V and 6V full scale. Could also be a percentage of the indicated value based on where the needle is sitting when the expansion is engaged.? I have seen some meters like this in the past for specialized instrumentation. If the meter was microprocessor driven, this would be an easy option to implement, verses the ways it was done in the past using offset voltages.
