Low R measurement on a DMM

[pope]

Forgive me if this not the correct section for the following question but I'm not sure where else I should post it. Mods, please feel free to move it if needed.

Considering a DMM, which reads a value of 0.1Ω when the probes are sorted. If we do a continuity test on a device and we measure 0.3Ω, can we assume that the actual resistance between the two points is 0.2Ω? i.e 0.3Ω minus the resistance of the probes (0.1Ω)? 

And if so, what's the benefit of a 4W multimeter?

[grumpydoc]

1) yes

2) to avoid the problem encountered in 1)

[Kleinstein]

It is OK to subtract the reading one gets from just shortening the probes. That is about the best one can do with 2 wire ohm mode, though the resitance to subtract may not be that constant - it can vary with the quality of the contact, though one may not see this with 0.1 ohm resolution.

The 4 wire ohm mode avoids the contact and cable resistance problem, with no need to the contact resistance to be stable.

[Fungus]

Considering a DMM, which reads a value of 0.1Ω when the probes are sorted. If we do a continuity test on a device and we measure 0.3Ω, can we assume that the actual resistance between the two points is 0.2Ω? i.e 0.3Ω minus the resistance of the probes (0.1Ω)? 

No, because there will be two extra resistances between the points of the probes and the circuit.

And if so, what's the benefit of a 4W multimeter?

It takes the resistance of the probes completely out of the equation.

nb. You can also do it with two multimeters and a current source, you don't need a special 4W meter.

[grumpydoc]

No, because there will be two extra resistances between the points of the probes and the circuit.

Well, one extra point of contact anyway.

You're right of course, there's an extra unknown (or unknowns) but in practice who would not just subtract the 0.1Ω either mentally or using the meter's "rel" function (if it has one, though all but very low end DMMs do).

[grumpydoc]

It might be worth pointing out that a normal DMM which only offers 2-wire measurement is not a great instrument for measuring 200 milli-ohms anyway, as Fungus hints the errors caused by contact resistance mean your accuracy will be about a single digit (OK, that might suffice for some applications).

As ever, what are you trying to measure, and why?

[pope]

As ever, what are you trying to measure, and why?

Continuity between PE and different points of the enclosure.

[2N3055]

4W measurements (Kelvin type) separate current and sense path.

That means that sense path is high impedance and measures voltage so any resistance (including contact resistance) has little influence.
On the other hand, as current source, 4W meters use constant current source, that will force certain constant current even if wiring and contact resistance in that current contour varies and is not stable. So you have stabilized current and separate voltage path.

Copper wires have quite high thermal coefficient. If you do 2W measurement by shorting wires, then measuring resistor, by the time you measure resistor things change... It is not stable.

4W measurement technique helps with measurements, compensates in changes and differences in cables etc etc... By the time you need to measure something below, say 10Ω, influence from cables is quite visible. Below 1Ω, influence form cables can be greater than DUT..

To put it colloquially, with 2W measurements your meter calibration is valid at multimeter sockets. With 4W calibration is (almost) valid (assuming correct good quality cabling and connection) at point of measurement, I.E at cable probes connected to DUT...
It is not exactly like that (metrology guys will kill me for this oversimplification) but it illustrates the difference.
4W purpose is to try to achieve that.

[JohanH]

Continuity between PE and different points of the enclosure.

It's not so useful to know with high accuracy. If the chassis is mains connected (as you have PE), PE is meant to be able to conduct mains voltage and short circuit current. Appropriate wiring for high current should be used. PE wire must be secured to chassis with a strong screw, preferably using a star washer that bites into the metal.

But if you have any kind of conductivity that your DMM can show (such as under 1 ohm), it will probably be fine. Of course, depending on application, you might want all parts of the chassis to have good connection between them (for shielding, EMI, ESD etc.). But again, you just want good conductivity, but the exact value isn't a big deal.

[pope]

Continuity between PE and different points of the enclosure.

It's not so useful to know with high accuracy. If the chassis is mains connected (as you have PE), PE is meant to be able to conduct mains voltage and short circuit current. Appropriate wiring for high current should be used. PE wire must be secured to chassis with a strong screw, preferably using a star washer that bites into the metal.

But if you have any kind of conductivity that your DMM can show (such as under 1 ohm), it will probably be fine. Of course, depending on application, you might want all parts of the chassis to have good connection between them (for shielding, EMI, ESD etc.). But again, you just want good conductivity, but the exact value isn't a big deal.

Well, usually there are some requirements that have to be met. Maximum of 100mΩ between the PE and any part of the enclosure, if I'm not mistaken.

[nctnico]

As ever, what are you trying to measure, and why?

Continuity between PE and different points of the enclosure.

AFAIK you need a special tester for that which uses high AC currents (up to tens of Amps). At least if you want to adhere to standards.

[JohanH]

You could have a tiny wire or connection somewhere measuring 100 mΩ, but it won't last the short circuit current that PE is designed for. So it depends on your application. If it's for hobby use, use visually good wire connection, trust your DMM and use double fused IEC socket and you are good to go. If this is for some commercial application, it gets complicated...

[BeBuLamar]

My meter has resolution down to 1 milliohm but I doubt that I can get accurate reading of a 0.2 ohms resistance.

[BillyO]

On some better meters the REL function works in ohms as well.  While this helps take the lead resistance out of the measurement, it does not help with contact resistance to the DUT.  Even with a 4W Kelvin arrangement, you still need to ensure clean, low resistance contact otherwise you will still be making inaccurate measurements.

[alm]

For the normal case of measuring say a low value current shunt, you'd do a four wire measurement by feeding a current, for example from a bench supply, through the resistor and measure the voltage across the resistor at its terminals (separate connections for voltage and current). Then divide the voltage by the current, and you get the resistor. For very low value resistors it helps to use a larger current, like 1A, to increase resolution.

For earthing resistance measurements there will likely be specific standards in your local electrical code. I wouldn't assume a random DIY solution would be sufficient there.

[bdunham7]

Considering a DMM, which reads a value of 0.1Ω when the probes are sorted. If we do a continuity test on a device and we measure 0.3Ω, can we assume that the actual resistance between the two points is 0.2Ω? i.e 0.3Ω minus the resistance of the probes (0.1Ω)? 

And if so, what's the benefit of a 4W multimeter?

If your meter uses a 1mA test current (typical but not universal) then a resistance of 0.3R will result in a voltage of 300µV.  So how well your meter can measure the difference between 0.1R and 0.3R depends on both the ability of the meter to resolve the 200µV difference and on any small contact resistance you introduce in the circuit when you change the position of the leads from shorted together to measuring your resistance.  You would want to do this a few times each way to see how repeatable your measurements are.  If they seem reliably repeatable, it wouldn't be all that unusual for  decent 4.5-digit meter (or better) to be able to resolve those measurements to at least 1 digit. However it is often a tricky and error-prone process.  Using a better meter and 4W kelvin-clip leads you can usually get 2-3 digits accuracy because you eliminate the effects of varying contact resistance and voltage drop in the leads.

[TimFox]

Also, with many DMM test leads, you can see a noticeable shift in resistance depending on the pressure you exert on the probes, which limits your repeatability for 2W ohms.

[DX1]

If you are trying to measure for safety then you need to use high current and measure voltage drop.
A simple way to generate high current is with a toroidal transformer and a single turn (OR small number of turns) adequately rated conductor wire through the core and connected to the measurement points. Ideally with a variac to adapt the primary voltage from 0 to 230V and a clamp meter to measure the current. Measure AC voltage across the measurement points directly (Not using the current injection connections.) AC voltage/AC current = resistance.

Use a hefty transformer to avoid burning it out.
You can get better accuracy using a current shunt adequately rated in the current loop and measure the voltage drop across the shunt.
You need to check the standard you are using to see what current you need and also the allowable resistance.
Be aware that the current can be high enough to melt cable, so ramp up the variac slowly from 0V. Also keep testing time short.

I did suggest using a Robin digital loop tester, but on checking there seems to be too much variation for it to be useful.
I am assuming you know how to do this safely. If not sure don't attempt it.

[2N3055]

For the normal case of measuring say a low value current shunt, you'd do a four wire measurement by feeding a current, for example from a bench supply, through the resistor and measure the voltage across the resistor at its terminals (separate connections for voltage and current). Then divide the voltage by the current, and you get the resistor. For very low value resistors it helps to use a larger current, like 1A, to increase resolution.

For earthing resistance measurements there will likely be specific standards in your local electrical code. I wouldn't assume a random DIY solution would be sufficient there.

DIY absolutely won't. For it to be done right you need to use proper certified earth continuity tester... I believe it is called PAT tester..
Currents used would be up to 25 A..

[DX1]

The OP will be doing precompliance testing, or production testing or simply checking his equipment is OK.
DIY can solve this, see my earlier post. Again this does involve mains wiring and high current. If you do not know how to do this safely don't.

For the normal case of measuring say a low value current shunt, you'd do a four wire measurement by feeding a current, for example from a bench supply, through the resistor and measure the voltage across the resistor at its terminals (separate connections for voltage and current). Then divide the voltage by the current, and you get the resistor. For very low value resistors it helps to use a larger current, like 1A, to increase resolution.

For earthing resistance measurements there will likely be specific standards in your local electrical code. I wouldn't assume a random DIY solution would be sufficient there.

DIY absolutely won't. For it to be done right you need to use proper certified earth continuity tester... I believe it is called PAT tester..
Currents used would be up to 25 A..

[AVGresponding]

As ever, what are you trying to measure, and why?

Continuity between PE and different points of the enclosure.

AFAIK you need a special tester for that which uses high AC currents (up to tens of Amps). At least if you want to adhere to standards.

Don't know how things are on that side of the moat, but here in the UK the PAT regulations have become ISITEE regulations    and high current testing of the PE is no longer required, my guess is due to the overwhelming majority of stuff not even having a PE these days, and also that so many testers are battery powered and just can't deliver the higher currents that mains powered ones used to do.

[Gyro]

These days, it's permissible to use proper electrical insulation / continuity testers for limited PAT testing, but BS7671 still requires a minimum of 200mA test current for PE - enclosure continuity.

[pope]

The meter I'm using is a Brymen BM869s but as it has mentioned even a 4W meter isn't suitable for this job.

Probably a hi-pot tester is the appropriate instrument for such measurements but unfortunately I can't afford one at the moment (they seem to go for over 1k) and secondly I don't feel very confident to use one as I've never done it before. AFAIK, you have to somehow discharge the DUT after the test and I have no idea how to do this. Also, although I appreciate the DIY recommendation I don't feel competent to to do it myself.

So, I was wondering whether a PAT tester would be sufficient. For some reason it's almost impossible to buy one in EU. It seems to be a UK thing. Searching around I found this one (which will set me back a bit over 500 euro) but I don't know anything about its quality and whether it makes sense to buy one. For example in a youtube video there's the following comment:

"I would love to know how this even qualifies as a PAT tester, it has a FAIL threshold of two Meg on a class one device according to the manual, yet the official threshold is one Meg. On a class two device it has a fail threshold of four Meg, when the the official threshold is two Meg, this all seems so very wrong when other testers are correctly calibrated, and this one seems to pass items other machines would fail."

https://www.firststopsafety.co.uk/product-page/memorypat-tester-with-display-and-memory

[2N3055]

The meter I'm using is a Brymen BM869s but as it has mentioned even a 4W meter isn't suitable for this job.

Probably a hi-pot tester is the appropriate instrument for such measurements but unfortunately I can't afford one at the moment (they seem to go for over 1k) and secondly I don't feel very confident to use one as I've never done it before. AFAIK, you have to somehow discharge the DUT after the test and I have no idea how to do this. Also, although I appreciate the DIY recommendation I don't feel competent to to do it myself.

So, I was wondering whether a PAT tester would be sufficient. For some reason it's almost impossible to buy one in EU. It seems to be a UK thing. Searching around I found this one (which will set me back a bit over 500 euro) but I don't know anything about its quality and whether it makes sense to buy one. For example in a youtube video there's the following comment:

"I would love to know how this even qualifies as a PAT tester, it has a FAIL threshold of two Meg on a class one device according to the manual, yet the official threshold is one Meg. On a class two device it has a fail threshold of four Meg, when the the official threshold is two Meg, this all seems so very wrong when other testers are correctly calibrated, and this one seems to pass items other machines would fail."

https://www.firststopsafety.co.uk/product-page/memorypat-tester-with-display-and-memory

Are you confusing earth continuity tester (that tests that grounding wiring is low enough resistance) with an insulation meter (HI Pot that tests that insulation is proper i.e. very high resistance)?

That tester has both of these functions. It can test insulation leakage from live to ground with 500V applied, and separately with 150mA it can test that housing is connected to earthing on plug with low enough resistance..

As for insulation test, more is better .. few MΩ is already very low resistance...   If instrument threshold is 4MΩ and standard is 2MΩ it means tester is more rigorous than standard, which means it errors on the safe side... Which means it will show failed on some marginal devices that would PASS on strictly calibrated devices.

Problem is that even 4 MΩ is actually very bad insulation on most devices outside industrial environments.. 100W toroidal transformers i tested recently had resistance between primary/secondary in GΩ ranges. One that was outside in damp, salty, environment for 5+years was more than 50 MΩ still...

[DX1]

Please can you explain what you are trying to do?
If low resistance PE to case and do not need to test to specification and have a power supply, Set current limit on power supply to 1A, connect the points you wish to test. Then measure the voltage drop with the Brymen. Thus use Ohms law to calculate resistance. Wire it 4 wire, which means that you connect to the measuring points, not to the power supply leads or clips. You can measure the power supply current as well to improve accuracy.

The meter I'm using is a Brymen BM869s but as it has mentioned even a 4W meter isn't suitable for this job.

Probably a hi-pot tester is the appropriate instrument for such measurements but unfortunately I can't afford one at the moment (they seem to go for over 1k) and secondly I don't feel very confident to use one as I've never done it before. AFAIK, you have to somehow discharge the DUT after the test and I have no idea how to do this. Also, although I appreciate the DIY recommendation I don't feel competent to to do it myself.

So, I was wondering whether a PAT tester would be sufficient. For some reason it's almost impossible to buy one in EU. It seems to be a UK thing. Searching around I found this one (which will set me back a bit over 500 euro) but I don't know anything about its quality and whether it makes sense to buy one. For example in a youtube video there's the following comment:

"I would love to know how this even qualifies as a PAT tester, it has a FAIL threshold of two Meg on a class one device according to the manual, yet the official threshold is one Meg. On a class two device it has a fail threshold of four Meg, when the the official threshold is two Meg, this all seems so very wrong when other testers are correctly calibrated, and this one seems to pass items other machines would fail."

https://www.firststopsafety.co.uk/product-page/memorypat-tester-with-display-and-memory