Bench DMM with multiple scales

[engineer_in_shorts]

Hi,

I have been recently looking at bench DMMs and came across a 7.5 digit one with multiple scales (full ranges), i.e 100mV and 300mV, then 1V and 3V, 10V and 30V and so on.

I think this is to prevent the issue of just going over range in one scale and then going up *10 scale, along with the associated accuracy loss and so on.

Is this an advantage? or just a bad idea?  Why do I not see this more often is I guess it is just a case of more switched in resistive dividers/gain resistors on the amplifier stages.
Thanks

[Simon123]

Multimeter ic usualy has some max voltage, to display max value.
If you would put 100V directly of its input it would maybe destroy device or give overlod signal.
Thats just a way, to attenuate the signal.

[engineer_in_shorts]

Multimeter ic usualy has some max voltage, to display max value.
If you would put 100V directly of its input it would maybe destroy device or give overlod signal.
Thats just a way, to attenuate the signal.

Yes, I understand that thank you, all multimeter have attenuation/gain stage in oprder to match the range full scale to the A-to-D scale.  I am talking about bench DMMs with dot matrix display (so the display is not limited to a number of digits).  Perhaps I need to clarify:

Advantage/Disadvantage of multiple counts within a specific decade as opposed to a fixed number of counts.

For example handheld multimeters tend to be fixed on 20000 counts.  Therefore range maximums are 199.99mV, 1.9999V, 19.999V, 199.99V and so on.
I have seen a DMM with both 100000 count and 300000 count ranges, so a measured voltage will (more likely) fall within a better range?

 

[c4757p]

Lower count ranges read faster at the cost of a digit. Useful for data acquisition, etc.

[engineer_in_shorts]

Lower count ranges read faster at the cost of a digit. Useful for data acquisition, etc.

I think these ranges are at the full resolution, but you have a good point - internally it could be averaging on the 30000 count to achieve the extra half digit.

[c4757p]

Remember that when it says "count", it means that literally. It's actually counting the time to charge a capacitor. The 100k count range simply stops the count when it hits 100k.

*Actual 300k count meters don't use the simple, naive dual-slope technique, they use a modified version that is more complicated, so it's not quite as simple as stopping the count early, but the principle is very much the same. The integration cycle is just cut short.

[wiss]

I guess you could use multiple references, 1V for the 1V range and 3V for the 3V range, but still just use 200 kcounts for both, this would give higher resolution than a 2V range when measuring up to 1V for ex...
In the Keithley 192 the .2 V range is partially implemented by dropping the reference voltage by a factor of 10.