How does a digital multimeter measure conductance?

[fourthindiana]

Preface to thread: Back in October I created a thread titled "How does a digital ammeter work?" That thread was about how digital ammeters measure current. The purpose of this thread is to learn how digital multimeters measure conductance, not voltage. This is a separate issue for 2 reasons: #1 I am asking about digital multimeters, not digital ammeters and 2# I'm asking about how digital multimeters measure conductance, not voltage. The two different questions could very well have two totally different answers. This makes it worthy of its own separate thread.
______________________________________________________________________

There seems to be a contradiction between what various PF members have told me about digital multimeters. On post #21 of my thread "How does a digital ammeters work?", PF member daven responded to me. I will put what davenn wrote in green text. Davenn wrote the following: I should have really added this in the previous post, just to make it really clear

ALL meters …. current, voltage, resistance, analog or digital, measure a voltage which is then directly displayed or digitised then displayed"

In post #27 of the "How does a digital ammeter work? thread, berkeman also responded. I will put what berkeman wrote in orange text: "It's a simple issue. Yes, DMMs use ADCs to convert the voltages into digital numbers. There are a number of different ADC types, and possibly all are used in different versions of DMMs. You can read more about them at Wikipedia, using these search terms"

Berkeman said that DMMs use ADCs to convert the VOLTAGES into digital numbers. Berkeman did not say anything about how DMMs measure current.

However, in my thread "Ohm reading when there is a short in the condenser motor", I asked jim hardy the following question: "When one lead of a multimeter is attached to a circuit and the other lead of the multimeter goes to ground, is it the case that a multimeter measures conductance in a circuit by applying voltage from its internal battery to a circuit and then the multimeter measures the current flow that leaves the multimeter?"Jim hardy told me that the answer to my question is yes. But I think that maybe jim hardy just meant that when one lead of a multimeter is attached to a circuit and the other lead of the multimeter goes to ground, a multimeter measures conductance in a circuit by applying voltage to its internal battery to a circuit and then the multimeter measures SOMETHING (perhaps voltage) that is used to calculate conductance.

When one lead of a multimeter is attached to a circuit and the other lead of the multimeter goes to ground, does the multimeter measure conductance in a circuit by applying voltage from its internal battery to a circuit and then the multimeter measure the current flow that leaves the multimeter? If so, how does the multimeter measure the current flow that leaves the multimeter? Does the multimeter measure the current flow by charging a capacitor and discharge the capacitor and counting how long it takes (Dual Slope ADC)? How does it work exactly?I have a Fluke 116 multimeter. When my Fluke 116 multimeter is set to read resistance, and when my multimeter measures conductance by applying voltage from its internal battery to a circuit, how does my Fluke 116 multimeter measure the current flow that leaves the multimeter? Does a Fluke 116 multimeter measure conductance by Successive Approximation ADC, Flash ADC, or Dual Slope ADC?

 

[CWatters]

Have you looked at a block diagram of a typical DMM...

http://electricalacademia.com/instr...r-working-principle/attachment/block-diagram/

In that diagram all the boxes in yellow convert the input (be it a voltage DC or AC or current) to a DC voltage which is digitised by an ADC on the right hand side.

 

[fourthindiana]

Have you looked at a block diagram of a typical DMM...

http://electricalacademia.com/instr...r-working-principle/attachment/block-diagram/

I looked at the block diagram, but I don't see how the block diagram answers the question.

 

[CWatters]

Conductance and resistance are measured in much the same way. A constant current (see block diagram) is sent through the component under test. This causes a voltage drop across the component which is scaled (by the buffer amplifier) and sent to the ADC.

 

[CWatters]

Forgot to add.. Usually dmm have range settings. When you change the range I'm not sure if the current changes or if the gain of the buffer amplifier changes. I suspect its the current but haven't checked.

Oh and some meters are auto ranging so this will be done automatically on those.

 

[fourthindiana]

Conductance and resistance are measured in much the same way. A constant current (see block diagram) is sent through the component under test. This causes a voltage drop across the component which is scaled (by the buffer amplifier) and sent to the ADC.

Is what you're saying not different than the question that I asked jim hardy?

Here is the question I asked jim hardy: "When one lead of a multimeter is attached to a circuit and the other lead of the multimeter goes to ground, is it the case that a multimeter measures conductance in a circuit by applying voltage from its internal battery to a circuit and then the multimeter measures the current flow that leaves the multimeter?"

Jim hardy said that the answer to my question is yes.

My interpretation of what Cwatters is saying: When one lead of a multimeter is attached to a circuit and the other lead of the multimeter goes to ground, the multimeter measures conductance in a circuit by applying voltage from its internal battery to a circuit and then the multimeter measures the voltage drop.

Please reconcile this.

 

[CWatters]

Is what you're saying not different than the question that I asked jim hardy?

Here is the question I asked jim hardy: "When one lead of a multimeter is attached to a circuit and the other lead of the multimeter goes to ground, is it the case that a multimeter measures conductance in a circuit by applying voltage from its internal battery to a circuit and then the multimeter measures the current flow that leaves the multimeter?"

Jim hardy said that the answer to my question is yes.

My interpretation of what Cwatters is saying: When one lead of a multimeter is attached to a circuit and the other lead of the multimeter goes to ground, the multimeter measures conductance in a circuit by applying voltage from its internal battery to a circuit and then the multimeter measures the voltage drop.

Please reconcile this.

I can't find a block or circuit diagram for the Fluke 116 so I can't be sure if it measures resistance by..

a) applying a fixed voltage and measuring the current, or by
b) applying a fixed current and measuring the voltage.

I suspect its b) but can't be certain. Jim suggests its a) he might be right

I don't think the Fluke 116 can display conductance. The spec I found suggests it just display resistance.

I don't believe connecting one probe to ground and one to the circuit is significant to your question. A DMM will try to measure the resistance between any two points you connect it to. If you believe its significant can you tell us why?

 

[CWatters]

My interpretation of what Cwatters is saying: When one lead of a multimeter is attached to a circuit and the other lead of the multimeter goes to ground, the multimeter measures conductance in a circuit by applying voltage from its internal battery to a circuit and then the multimeter measures the voltage drop.

No I said it applies a constant current (which ultimately does come from the battery) and measures the voltage across the external circuit.

 

[fourthindiana]

I can't find a block or circuit diagram for the Fluke 116 so I can't be sure if it measures resistance by..

a) applying a fixed voltage and measuring the current, or by
b) applying a fixed current and measuring the voltage.

I suspect its b) but can't be certain. Jim suggests its a) he might be right

I don't think the Fluke 116 can display conductance. The spec I found suggests it just display resistance.

The Fluke 116 only displays resistance, but the Fluke 116 uses conductance to calculate the resistance.

I don't believe connecting one probe to ground and one to the circuit is significant to your question. A DMM will try to measure the resistance between any two points you connect it to. If you believe its significant can you tell us why?

I don't necessarily think that connecting one probe to ground and one to the circuit is significant. I only wrote told everyone that one probe is connected to ground and one to the circuit just in case it is significant. I thought you all should have all the details.

 

[jim hardy]

The analog meter we studied in that shorted condenser motor thread indeed measures current , that's what D'Arsonval meters do. They're an electromagnet.

Digital meters measure voltage.
When selected to Current scale, the selector switch connects a low ohm resistor between the meter's input terminals.
It also tells the DMM to read and display volts on its lowest scale, typically 200 millivolts .

Now if that Resistor is one ohm, than a current of 200 milliamps will produce 200 millivolts. So you'd label that range switch position "200 ma"
Next, if that Resistor is ten ohms, than a current of just 20 milliamps will produce 200 millivolts. So you'd label that range switch position "20 ma".
Next, if that Resistor is one hundred ohms, than a current of just 2 milliamps will produce 200 millivolts. So you'd label that range switch position "2 ma".
And so on.
The Range switch tells the DMM where to put the decimal in the display to read the correct amps, milliamps or microamps in the display . That's because the display can only go , in my example, to 200 counts. Some meters have way more digits.

Now - to measure Resistance (or Conductance) with a DMM they do something similar.
Accept for the moment that it's not difficult to make circuit that produces a constant current.and we call that (not surprisingly) a "constant current source".
So they just replace that resistor with a constant current source.

Test current is pushed through the unknown resistance by the constant current source.
DMM measures the voltage that it took to do that.
If test current is 1 milliamp then 200 ohms between the meter leads will result in 200 millivolts. So you'd label that position of the range switch "200 ohms" .
If test current is 0.1 milliamp then 2000 ohms between the meter leads will result in 200 millivolts. So you'd label that position of the range switch "2K ohms" .and place a decimal after the first digit.
If test current is 0.01 milliamp then 20,000 ohms between the meter leads will result in 200 millivolts. So you'd label that position of the range switch "20K ohms" .and place a decimal after the second digit.

And so on.

It really is that simple.

You'll probably find your Fluke has diode check scale that works the same way but uses 2000 millivolts not 200 -
that's because 2 volts will force current through a diode so you'll read around 0.6 on a good one.
And 2 volts will force current through a Darlington transistor's two base-emitter junctions. A good one will read about 1.2 volts.

hope this helps.
You ought to practice simple Ohm's Law calculations like this in your head while in the shower or watching TV. You need to make them intuitive.

Good luck in your HVAC endeavours . The world needs practical men who can fix things.

old jim

 

[Tom.G]

I see @jim hardy posted just as I was starting this post. He and I are saying the same thing but in different words. Perhaps between the two attempted explanations there is enough to help with your concerns.

"When one lead of a multimeter is attached to a circuit and the other lead of the multimeter goes to ground, is it the case that a multimeter measures conductance in a circuit by applying voltage from its internal battery to a circuit and then the multimeter measures the current flow that leaves the multimeter?"

Jim hardy said that the answer to my question is yes.

That is true for the ANALOG meters. There is a set of resistors in the meter. The different Resistance ranges select which resistor is in series with your unknown resistor, then the voltage is measured across the internal, known, resistor.

how does my Fluke 116 multimeter measure the current flow that leaves the multimeter?

I looked at the block diagram, but I don't see how the block diagram answers the question.

The current source in the block diagram is the Blue box at the top left.
http://electricalacademia.com/instr...r-working-principle/attachment/block-diagram/
It puts out a known current, set by what Resistance range the meter is set to. The meter then reads the voltage across the unknown resistor and displays this as the measured resistance. If you have, or can borrow, a second meter you can measure the voltage across the resistor while measuring its resistance with your meter. That works best with resistor values from about 100 Ohms thru 47k Ohms.

Hope this helps.

Cheers,
Tom

 

[jim hardy]

If you have, or can borrow, a second meter you can measure the voltage across the resistor while measuring its resistance with your meter.

That's a great learning exercise.

Connect two meters together + to + and - to -..
Place one on ohms and the other on current.
What does each read? Why ?
Then switch the current meter to volts.
What does each read ? Why ?

That'll be interesting - I've never done it with autoranging meters. They might chase one another's tails.

old jim

 

[Merlin3189]

I see a load of responses have appeared since I went to bed, so I've deleted most of my explanation and just leave some thoughts.

Why do you want to know what sort of ADC is used? I too have spent some time searching and I can't find any info from Fluke on this. I think that is because it doesn't matter to me, you, 99.999% of their customers nor probably to most Fluke employees. So long as it performs to their spec, that's what's wanted and no one but the Fluke designers needs to care how they achieve it - and they may prefer to keep it as a trade secret.
Maybe it's just idle curiosity. In that case, tell me when you find out, because I'm idle and curious as well! My guess would have been dual slope integration, because that's what was common back when manufacturers liked to tell us such things. Since yesterday I'd probably revise that to successive approx, since that seems to be the opinion of some websites, which try to explain DMMs nowadays.

I can understand your wanting to know how resistance is measured, because although the voltages and currents are small ( <2.7 V, <350 μA for 116) they could conceivably affect some measurements. Jim's come up with the right idea about that, so you should be able to check this yourself easily.

But your more general quibble about currents and voltages seems misguided. IMO as a general rule an ADC cmeasures a voltage. I say that because they ultimately have to compare the input with some reference, which will usually be a bandgap, zener or avalanche device.
But for a multimeter, digial or analogue, it's irrelevant. What makes a basic meter into an AVO multimeter is the array of switchable high precision resistors. These have the amazing property of converting currents into voltages and voltages into currents. Whatever range you select, your input is applied to a suitable combination of these, so that a proportional voltage or current, as required by the ADC, is provided.
Whether the ADC is defined by current or voltage makes no difference. In current modes your circuit will see a low resistance and in voltage modes it will see a high resistance, both as defined in the specs. This would be true for any multimeter, however it was implemented.

The Fluke 116 only displays resistance, but the Fluke 116 uses conductance to calculate the resistance.

This is a puzzling comment. Apart from the fact that conductance and resistance are just different numbers for the same thing - like fuel consumption in mpg or l/km - I'd have said that what a meter displays is what it's measuring.
You could argue that it is calculating resistance by measuring voltage and/or current, but for conductance the measurements would be exactly the same ones, only the calculation is slightly different. When you've measured the voltage &/or current you have exactly the information you require. There is no need to calculate resistance and convert that to conductance, nor vice versa: you just do the calculation in the way that gives you the quantity in the form you want to display.

 

[jim hardy]

You could argue that it is calculating resistance by measuring voltage and/or current, but for conductance the measurements would be exactly the same ones, only the calculation is slightly different. When you've measured the voltage &/or current you have exactly the information you require. There is no need to calculate resistance and convert that to conductance, nor vice versa: you just do the calculation in the way that gives you the quantity in the form you want to display.

That is exactly so.
Conductance and Resistance are two ways of looking at the same property - just like "glass half full" or "glass half empty" .
I introduced @fourthindiana to the concept of conductance in an earlier thread.
The reason for that was he'd asked a question about what will an ohmmeter read when there are parallel paths .
I wanted to lead his thinking toward how currents through parallel paths add to one another - you just sum them up..

Planting that germ of thought early in his studies,
and familiarizing him with the reciprocal relation of resistance to conductance,
should make calculation of resistances in parallel almost intuitive for him when his class gets to that.

"Take the reciprocal of the sum of the reciprocals" is as much a mouthful as $R_P = \frac{1}{\frac{1}{R_1} +\frac{1}{R_2} +\frac{1}{R_3} +\frac{1}{R_...}etc}$ is an eyeful.
While it's intuitive to all of us now, think back to when it wasn't.
I remember feeling intimidated by that algebra and wanted to prepare him for it .

That's why i taught him "the ohmmeter measures conductance of the unknown,
in fact it measures the sum of all the conductances in the unknown.".
I could have said "...it measures the parallel combination of all the resistances in the unknown" but i don't think he's calculating parallel resistances yet.
and to a beginner , "sum" is way less intimidating than "reciprocal of sum of reciprocals".

That was my thinking. I could be all wet .

Per the manual, auto-ranging also "sets the Meter’s input impedance to approximately 3 kΩ to reduce the possibility of false readings due to ghost voltages."

now THAT's clever ! We've had threads on 'ghost voltage' before
https://www.physicsforums.com/threads/ghost-voltages.926498/#post-5848151

 

[Windadct]

Just dropping in - but... if you really understand how an analog ( we'll say current movement) is used to measure voltage, or resistance... then it should be pretty straight forward to see how a digital ( voltage sensing) meter can be used to measure current or resistance, etc. ( Granted I did VoTech HS where we actually built all of the standard configurations, before I went to Uni...)

A true current source is not needed if we have a sensing circuit of the output current, or realizing that a digitally controlled current source ALSO would probably have a voltage sensing circuit as part of its control loop (if it is digitally controlled..).

Also - once these elements are digitized with some accuracy, then any mathematical manipulation is trivial.

As for the ADC used, I expected that Fluke is using their Own ASIC _ however, eevblog (Dave) did a teardown on a Fluke 117, and it is using a TI MSP430 family chip... this family has a variety of ADC options, up to 14 bit depth.