3-1/2 digits?

[Axtman]

What is meant by 3-1/2 digits when describing a multimeter?  What is a 1/2 digit?

[Falcon69]

it's where you will have 4 digits, but will read 115.6.  that period stays.

[Falcon69]

sorry, i was wrong, take a look at this

http://www.london-electronics.com/digitresolution.php

[PedroDaGr8]

Not true.

It refers to the first digit in say a 4000 count meter. On that first digit it only goes from 0-4 so it only goes halfway up. It's a half digit. Colloquially anything that isn't a full digit 0-9 is called a half digit.

Sent from my EVO using Tapatalk

[Wytnucls]

[codeboy2k]

They've started doing other wonky stuff too, like 5-3/4 digits ... I guess that means it goes to 750000 counts?

I dunno.. why don't they just give the damn counts...

Counts!! I need counts!!! focus dammit.. echen... where are you?

[AG6QR]

The counts should be related to the digits by a logarithmic/exponential relationship.

A 1 digit meter can measure anything from 0 to 9, or 10 ^ 1 counts.

A 2 digit meter can measure anything from 0 to 99, or 10 ^ 2 counts.

A 3 digit meter goes to 10 ^ 3 counts.

A 4 digit meter goes to 10 ^ 4 counts.

You can see the logarithmic trend.  We have the following equations, which should, in theory, be extendable to fractional digits.

counts = 10 ^ digits

digits = log10(counts)

A 3 1/2 digit meter, therefore, ought to have about 10 ^ 3.5 counts.  That's pretty close to 3000 counts.

A 2000 count meter ought to be called a 3.3 digit meter or maybe 3 1/3 digits. 

5 3/4 digits should be 56243 counts, which would be rounded to 50000 or 60000.


I realize that I'm approaching this from an engineering mindset, while many in the marketing department people don't understand logarithms and/or try to play fast and loose with the numbers to push them to whatever they can get away with.  So don't take these too literally.   In practice, you can't trust the fractional number of digits claimed; you've got to ask for the counts.

[Rick Law]

The counts should be related to the digits by a logarithmic/exponential relationship.

A 1 digit meter can measure anything from 0 to 9, or 10 ^ 1 counts.

A 2 digit meter can measure anything from 0 to 99, or 10 ^ 2 counts.

A 3 digit meter goes to 10 ^ 3 counts.

A 4 digit meter goes to 10 ^ 4 counts.

You can see the logarithmic trend.  We have the following equations, which should, in theory, be extendable to fractional digits.

counts = 10 ^ digits

digits = log10(counts)

A 3 1/2 digit meter, therefore, ought to have about 10 ^ 3.5 counts.  That's pretty close to 3000 counts.

A 2000 count meter ought to be called a 3.3 digit meter or maybe 3 1/3 digits. 

5 3/4 digits should be 56243 counts, which would be rounded to 50000 or 60000.


I realize that I'm approaching this from an engineering mindset, while many in the marketing department people don't understand logarithms and/or try to play fast and loose with the numbers to push them to whatever they can get away with.  So don't take these too literally.   In practice, you can't trust the fractional number of digits claimed; you've got to ask for the counts.

It is the marketing kind that changed 1GB to 1 million bytes - just to show a larger number.  Prior to that, it is always 2^10 meaning 1GB=1024MB.  Some even has 1MB=1000KB making the X.xGB or X.xTB very misleading.

I am still confused with disk sizes and rely on raw LBA count or other means.

[PedroDaGr8]

The counts should be related to the digits by a logarithmic/exponential relationship.

A 1 digit meter can measure anything from 0 to 9, or 10 ^ 1 counts.

A 2 digit meter can measure anything from 0 to 99, or 10 ^ 2 counts.

A 3 digit meter goes to 10 ^ 3 counts.

A 4 digit meter goes to 10 ^ 4 counts.

You can see the logarithmic trend.  We have the following equations, which should, in theory, be extendable to fractional digits.

counts = 10 ^ digits

digits = log10(counts)

A 3 1/2 digit meter, therefore, ought to have about 10 ^ 3.5 counts.  That's pretty close to 3000 counts.

A 2000 count meter ought to be called a 3.3 digit meter or maybe 3 1/3 digits. 

5 3/4 digits should be 56243 counts, which would be rounded to 50000 or 60000.


I realize that I'm approaching this from an engineering mindset, while many in the marketing department people don't understand logarithms and/or try to play fast and loose with the numbers to push them to whatever they can get away with.  So don't take these too literally.   In practice, you can't trust the fractional number of digits claimed; you've got to ask for the counts.

It is the marketing kind that changed 1GB to 1 million bytes - just to show a larger number.  Prior to that, it is always 2^10 meaning 1GB=1024MB.  Some even has 1MB=1000KB making the X.xGB or X.xTB very misleading.

I am still confused with disk sizes and rely on raw LBA count or other means.

Personally I like the new definition better. Yes it short changes the consumer, but the fact is, it's correct. I hated that kilo meant one thousand expect with computers where it was 1024. These modern definitions are true to the sense of the SI prefixes.

Sent from my EVO using Tapatalk

[sync]

It is the marketing kind that changed 1GB to 1 million bytes - just to show a larger number.  Prior to that, it is always 2^10 meaning 1GB=1024MB.  Some even has 1MB=1000KB making the X.xGB or X.xTB very misleading.

I am still confused with disk sizes and rely on raw LBA count or other means.

This urban legend will never die. Hard disk sizes are traditionally measured with 1M=10^6 and so on. Way before the PC existed. But the half digit is probably a marketing invention. On early DMMs it was not used, e.g. a 120000 count HP 3450A was named a 5 digit meter.

[bronson]

Yes, "1/2 digit" doesn't really make sense...  It's from the bad old days when the 7 segment display was a big part of a device's cost.  The manufacturers added two vertical bars to the left to display 199 instead of just 99, then claimed that double the range means you received another 1/2 digit.

Counts makes more sense and is easier to compare but it's a bit unwieldy to talk about 2e6 count multimeters.  And what on earth does "display count = 1000" mean for the 6 1/2 digit 34401?

      http://www.newark.com/agilent-technologies/34401a/multimeter-digital-bench-6-1-2/dp/91F3030

So it's hopeless.   

[Monkeh]

It is the marketing kind that changed 1GB to 1 million bytes - just to show a larger number.  Prior to that, it is always 2^10 meaning 1GB=1024MB.  Some even has 1MB=1000KB making the X.xGB or X.xTB very misleading.

I am still confused with disk sizes and rely on raw LBA count or other means.

This urban legend will never die. Hard disk sizes are traditionally measured with 1M=10^6 and so on. Way before the PC existed.

It and the confusion will never die so long as people insist on being stubborn, ignorant, and illogical.

[Rick Law]

It is the marketing kind that changed 1GB to 1 million bytes - just to show a larger number.  Prior to that, it is always 2^10 meaning 1GB=1024MB.  Some even has 1MB=1000KB making the X.xGB or X.xTB very misleading.

I am still confused with disk sizes and rely on raw LBA count or other means.

This urban legend will never die. Hard disk sizes are traditionally measured with 1M=10^6 and so on. Way before the PC existed. But the half digit is probably a marketing invention. On early DMMs it was not used, e.g. a 120000 count HP 3450A was named a 5 digit meter.