DMM Test Lead Resistance

[Attorney]

Caveat emptor...

This issue has likely been raised in the past, but thought I would pass along a set of observations made when trying various brands of DMM test leads.  Last week, a set of what looked like really nice test leads showed up.  The outer quality seemed fine, much like the Fluke branded leads.  But upon applying a shorted lead test on my Fluke 45 and Keithley 2100 DMMs, the new leads show 0.31 ohms.  Next, I tried my old Fluke 8060A leads.  They read 0.15 ohms.  Finally, the Unicable leads that came with the Keithley displayed the lowest resistance of 0.08 ohms.  The outer diameter of all leads is the same.  Lead length is approximately the same.  Both DMMs show nearly identical readings (within one count on both DMms).  The tests were repeated multiple times to ensure I had maxium tip conduction.

I have not cracked open the new test leads but my guess is that the wire gauge is quite bit smaller than the other sets.  As a practical matter for most use, I doubt this would be of concern except when taking high current measurements - something I don't often do with a DMM.  In those cases, I use precision 1%, 0.1 ohm series resistors and carefully observe voltage drop across a known value.

Anyone have standard length DMM test leads that measure even higher than mine?

Paul

[alm]

Test lead resistance will also introduce an error when measuring low value resistors, like the 0.1 ohm resistor you mention. You can either compensate for this with the relative/null function or use four wire connections (either using a DMM with support for four wire measurements or using a current source and DMM). The relative option is the fastest, but has the disadvantage that lead resistance will often change slightly depending on contact pressure.

[Co6aka]

FWIW, a while back I made myself a set of low resistance leads with some of that 12-gauge silicone-insulated super-flexible red-n-black interconnect-wire. (Now THAT's a lotta hyphenated-words!) Still use 'em with meters that don't have 4-wire capability, and I haven't yet lighted up my eyeballs, or let the magic smoke out.

PS- You could prolly ream out the backside ( I'm in rare form tonight) of the banana plug housings and get the #8 version of that wire in there. Or if not, then the #10 version.

[Attorney]

Test lead resistance will also introduce an error when measuring low value resistors, like the 0.1 ohm resistor you mention. You can either compensate for this with the relative/null function or use four wire connections (either using a DMM with support for four wire measurements or using a current source and DMM). The relative option is the fastest, but has the disadvantage that lead resistance will often change slightly depending on contact pressure.

When using the 0.1 ohm resistor for high current measurements, I measure voltage drop across it -- not resistance.  Then I compute current.   When the input Z of the measuring device is in the megohm range, next to no current flows in the probe leads.  With no current, the voltage measurement across the precision resistor is highly accurate. If I know the E drop and if I know the exact resistance, I can accurately solve for current. This is how many industrial control circuits are measure current.

I'm interested in your statement about the shunt when taking high current measurements. If the leads introduce resistance, and as current increases, won't the lead resistance begin affecting current measurement accuracy?  I'm referring to a test where the DMM is inserted in series with the circuit for the current measurement.

Paul

[robrenz]

alm was referring to when you are measuring the resistance of the 0.1 ohm resistor not when using the resistor as a shunt to measure current.

[alm]

When using the 0.1 ohm resistor for high current measurements, I measure voltage drop across it -- not resistance.

I wasn't talking about measuring voltage across the resistor or using it to measure current, but measuring its resistance. Or any other resistor where < 1 ohm resolution / accuracy is required.

I'm interested in your statement about the shunt when taking high current measurements. If the leads introduce resistance, and as current increases, won't the lead resistance begin affecting current measurement accuracy?  I'm referring to a test where the DMM is inserted in series with the circuit for the current measurement.

Imagine the 0.1 ohm resistor with the DMM measuring the voltage across it as the DMM in current mode. Then put two 0.1 ohm resistors in series with the shunt. Will this affect the voltage across the shunt? It will increase the burden voltage, which might affect the current through the shunt, but the DMM will still correctly indicate the current through the shunt.

[BravoV]

I'm interested in your statement about the shunt when taking high current measurements. If the leads introduce resistance, and as current increases, won't the lead resistance begin affecting current measurement accuracy?  I'm referring to a test where the DMM is inserted in series with the circuit for the current measurement.

Understanding the Kirchhoff law may help you to grip better understanding on why connecting "anything" in series will always have the same exact current passed through anything that is connected in the connection.

Meaning the same exact current will pass through the DMM -> lead -> DUT -> lead and back to DMM again.

[Attorney]

When using the 0.1 ohm resistor for high current measurements, I measure voltage drop across it -- not resistance.

I wasn't talking about measuring voltage across the resistor or using it to measure current, but measuring its resistance. Or any other resistor where < 1 ohm resolution / accuracy is required.

 

But it was in my introduction.  Neither one of us are on the same page.  If you go back and look at my intro, I stated that the 0.1 ohm resistor is used for high-current measurements.  Nothing more.  I'm not looking at measured resistance accuracy when I'm measuring voltage across a precision resistor in order to compute current.  When computing current this way with a Hi-Z input DMM, I don't care about wire lead resistance at all. 

I'm interested in your statement about the shunt when taking high current measurements. If the leads introduce resistance, and as current increases, won't the lead resistance begin affecting current measurement accuracy?  I'm referring to a test where the DMM is inserted in series with the circuit for the current measurement.

Imagine the 0.1 ohm resistor with the DMM measuring the voltage across it as the DMM in current mode. Then put two 0.1 ohm resistors in series with the shunt. Will this affect the voltage across the shunt? It will increase the burden voltage, which might affect the current through the shunt, but the DMM will still correctly indicate the current through the shunt.

Agreed that it's still measuring the accuracy across the shunt.  Let's assume for a moment that our test leads were to deteriorate to say...100 ohms resistance.  Let's run our DMM in series -- and current is 10A DC.  The DMM in current mode is measuring the right current no matter the lead resistance.  However, my point from the beginning (and that is the title of this thread) is that the lead resistance can affect the circuit by reducing the in-circuit current and give an erroneous reading as to the circuit current if the DMM were not there.

Circling back...Because the lead resistance can reduce circuit current, I place a 0.1 ohm resistor in series with the circuit.  This value means the circuit is being perturbed as little as possible for the purpose of the current measurement.  In doing so, I've removed test lead resistance out of the equation.

Hoping to align us on the same page...


Paul 

     

[Attorney]

I'm interested in your statement about the shunt when taking high current measurements. If the leads introduce resistance, and as current increases, won't the lead resistance begin affecting current measurement accuracy?  I'm referring to a test where the DMM is inserted in series with the circuit for the current measurement.

Understanding the Kirchhoff law may help you to grip better understanding on why connecting "anything" in series will always have the same exact current passed through anything that is connected in the connection.

Meaning the same exact current will pass through the DMM -> lead -> DUT -> lead and back to DMM again.

That's right and to my point.  The greater the test lead resistance, the greater the voltage drop in the leads.  Kirchhoff still applies and the leads have now just reduced total circuit current.  The DMM is reading the true current through it,  but it's seeing a value that's affected by lead resistance as current increases.  The leads are affecting circuit current, not just the displayed measurement.

See my post about using the 0.1 ohm resitor to circumvent the current measurement problem.  But keep in mind that the leads are then used across the resistor to measure voltage in order to compute current.

Paul

[BravoV]

Lead resistance is not a significant factor (up to certain limit of course) when the DMM is in voltage mode, because the DMM input impedance is so high (Mega/Giga Ohm) that makes those minuscule resistance at the lead become meaningless.

[Attorney]

Lead resistance is not a significant factor (up to certain limit of course) when the DMM is in voltage mode, because the DMM input impedance is so high (Mega/Giga Ohm) that makes those minuscule resistance at the lead become meaningless.

That's why it's often more accurate to use the DMM in voltage mode across a small resistance to measure current.  That way, the DMM in no way affects circuit current, nor the measurement -- apart from it's own DC voltage resolution/accuracy.

Paul

[alm]

Yes, lead resistance affects the burden voltage and may indirectly affect the current measurement, as I stated in my second post.

You're apparently just looking for confirmation of whatever point you're trying to make, so I'm out of here.

[ModemHead]

Anyone have standard length DMM test leads that measure even higher than mine?

For measuring test lead resistance, I made a tip-shorting connector from two 2mm pin jacks and a short length of 12 gauge copper wire.  Plug the leads into a power supply set for 1A current, and measure the millivolt drop across the plugs.

Most of the leads I have around here are various Fluke and Pomona sets.  The venerable old Fluke TL75s generally measure the lowest at around 60 to 70 milli-ohms.  The "premium" TL71s usually measure the highest, anywhere from 100 to 200 milli-ohms, possibly due to the tip alloy.

Pictured are some modular leads I use a lot which are mish-mash of Fluke leads and Pomona probes.  This pair measures about 90 milli-ohms, despite the extra pair of connections.

None of this makes any real difference to me in actual use.  I generally use patch cords and/or clips for current measurements, not probes.  And 4-wire kelvin clips for low-value resistors.

[mos6502]

My Fluke 77 has a shunt resistance of 5.545 Ohms on the 300mA range and 0.095 Ohms on the 10A range. A 30 mOhm lead resistance isn't going to affect the circuit in any significant way. Really, the only place where you need a four wire setup is when measuring impedance. And, surprise, impedance (LCR) meters do have four wire capability ( at least the good ones).

Incidentally, a really good way to measure lead resistance is an LCR meter with split banana jacks like the DEREE DE-5000. It will do a four wire measurement and show only the true internal wire resistance.

In the pic, you can see a 110cm long Hirschmann lead that is marked as 1 sqmm. The measured resistance is 19mOhms. Using the formula A=rho*l/R, the wire area calculates to 0.973 mm^2 - only a 2.7% error from the expected value, which is an insanely good result considering all the physical variables.

If you measure both the length and the thickness, this is a also a good way to tell if you're actually dealing with copper or some other copper plated metal.

And one more thing, when using an LCR meter to measure resistance, you also eliminate any thermoelectric effects that will occur at crimp/solder joints in the cable.

[Attorney]

My Fluke 77 has a shunt resistance of 5.545 Ohms on the 300mA range and 0.095 Ohms on the 10A range. A 30 mOhm lead resistance isn't going to affect the circuit in any significant way. Really, the only place where you need a four wire setup is when measuring impedance. And, surprise, impedance (LCR) meters do have four wire capability ( at least the good ones).

Incidentally, a really good way to measure lead resistance is an LCR meter with split banana jacks like the DEREE DE-5000. It will do a four wire measurement and show only the true internal wire resistance.

In the pic, you can see a 110cm long Hirschmann lead that is marked as 1 sqmm. The measured resistance is 19mOhms. Using the formula A=rho*l/R, the wire area calculates to 0.973 mm^2 - only a 2.7% error from the expected value, which is an insanely good result considering all the physical variables.

If you measure both the length and the thickness, this is a also a good way to tell if you're actually dealing with copper or some other copper plated metal.

And one more thing, when using an LCR meter to measure resistance, you also eliminate any thermoelectric effects that will occur at crimp/solder joints in the cable.

Thanks for the suggestion and for the reminder that the DE-5000 has split jacks.  Last week, I placed and order and it should soon arrive.  This will be one of my first tests with the device. 

Paul

[robrenz]

I am a big DE-5000 fan but unless you have matching split 4 wire banana leads, the jacks do you no good.  My fluke 2/4 kelvin probes are split bananas to match the 8846A but the split is in the wrong direction to match the DE-5000. So you need to use the alligator clip or tweezer units that plug in to the component test slots to get the 4 wire capability out to the test point. Remember you need to do a open/short cal with whatever you have plugged in for leads. If you want the best accuracy you need to replace the alligator clips with real kelvin clips that keep source and sense separate all the way to the component leads.

[mos6502]

I am a big DE-5000 fan but unless you have matching split banana leads, the jacks do you no good.

Well, in this specific case they do - they're perfect for measuring the resistance of test leads.

But yeah, most of the time, I use the TL21 or TL22 or the leaf terminals.

Although, even if you used normal test leads, at least the split jacks cancel out the connector resistance. I can't believe they're not standard on all multimeters. The price increase would be what, 50 cents?

My fluke 2/4 kelvin probes are split bananas to match the 8846A but the split is in the wrong direction to match the DE-5000. So you need to use the alligator clip or tweezer units that plug in to the component test slots to get the 4 wire capability out to the test point. Remember you need to do a open/short cal with whatever you have plugged in for leads. If you want the best accuracy you need to replace the alligator clips with real kelvin clips that keep source and sense separate all the way to the component leads.

Correct. Before I did the measurement above, I did an open cal and for the short cal I connected the leaf terminals with a short piece of silver plated wire.

Anyway, I ordered a set of these:

http://www.ebay.de/itm/LCR-Meter-Test-Leads-LCR-test-Clip-Terminal-Kelvin-Test-Line-BNC-CABLE-/190916817101?pt=LH_DefaultDomain_0&hash=item2c738770cd

And I'm going to wire them into the TL-21 case, as the short wire croc clips are pretty useless and not really four wire at all.

I have to say that the DE-5000 is easily one of the most if not the most useful tools for me now. I'm learning new things all the time. Like how the skin effect affects resistance - the resistance of the 19 mOhm test lead increases to 58 mOhms at 100kHz. And that's the real part, i.e. ohmic resistance. And I'm still surprised how consistently accurate it is. I wouldn't even be pissed if I had paid $350 for it. At $120, it's unbelievable value.

PS I'd love to know if/where you can buy split 4mm plugs - cable/panel/PCB mount, doesn't matter.

[cs.dk]

I have to say that the DE-5000 is easily one of the most if not the most useful tools for me now.

Sorry for breaking in - But where do you find these meters? Ebay.de shows _one_ hit on this. It was from UK to 365£

[mos6502]

I bought mine from Japan, seller akibashipping:

http://www.ebay.de/itm/DER-EE-DE-5000-High-Accuracy-Handheld-LCR-Meter-with-TL-21-TL-22-JAPAN-NEW-/141104368904?pt=LH_DefaultDomain_0&hash=item20da79cd08

[cs.dk]

Thanks alot.. Don't know why I didn't get it in my search.

EDIT: Ordered this thing now - How do you make 4-wire measurements, i don't get that part?

[mos6502]

It's a known problem. The eBay search is broken. Sometimes things show up on one domain but not the other, even for items that have worldwide shipping. So it's a good idea to repeat your search on other eBay TLDs.

[cs.dk]

With your search-term, it works fine on Ebay.de, .co.uk, .com, etc.. Weird.

I've updated my post above, about 4-wire measurements.

[mos6502]

Thanks alot.. Don't know why I didn't get it in my search.

EDIT: Ordered this thing now - How do you make 4-wire measurements, i don't get that part?

The simplest is to stick your component into the leaf spring terminals. That will do a real 4 wire measurement.

You also get the tweezers and the croc clips. They are 4-wire up to the tips. So the wire resistance and the contact resistance to the meter are eliminated and only the contact resistance between the croc clips/tweezer tips and the component remains.

The connection boxes of the tweezers and the croc clips can be easily opened and you can solder your own leads to them. I posted a link further up to some actual 4 wire tweezers that I plan to replace the croc clips with that will also do a real 4 wire measurement all the way to the DUT.

Generally, 4-wire measurements become important when dealing with low impedances: mOhms, pF, nH.

Check out Dave's teardown, he talks about the connection options:

[cs.dk]

Thank you for your reply 

Seems like a nice meter. I can't find the review though.

[mos6502]

Yeah, strangely, Dave never did a review on it. I did find this post, though:

https://www.eevblog.com/forum/blog/when-is-the-10khz-tonghui-vs-10-khz-agilent-lcr-review-coming/msg73261/#msg73261