Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Guy Holt]

Here’s puzzle.  The resistance of 4/O copper cable is .0504 Ωs/1000ft, or .0000504 Ωs/ft. Yet, when I measure the resistance of 100’ 4/O feeder cable it ranges from .2 to .6 Ωs.  What accounts for such high resistance when 100’ of 4/O should have a resistance of only .00504 Ωs? I know it is not the connection of the copper cable to the Camlok fittings that adds so much resistance because I have taken readings through the rubber jacket of the cable before the fittings.  And, I know that it has nothing to do with the 4/O feeder consisting of over two thousand individual strands of copper rather than a single strand because the resistance tables for SC cable is only slightly higher than THHN cable.  So what accounts for the higher resistance of 4/0 feeder?

[Kevin Graf]

It's not easy to measure small fractions of an Ohm. What type of meter are you using?
Actually you need a four terminal Ohm meter to do the job. Or two DMM and a current source.

* * * * * * * *
For a DIY measurement method, scroll well down the page.
http://www.beta-a2.com/EE-photos.html

[John Roberts {JR}]

Here’s puzzle.  The resistance of 4/O copper cable is .0504 Ωs/1000ft, or .0000504 Ωs/ft. Yet, when I measure the resistance of 100’ 4/O feeder cable it ranges from .2 to .6 Ωs.  What accounts for such high resistance when 100’ of 4/O should have a resistance of only .00504 Ωs? I know it is not the connection of the copper cable to the Camlok fittings that adds so much resistance because I have taken readings through the rubber jacket of the cable before the fittings.  And, I know that it has nothing to do with the 4/O feeder consisting of over two thousand individual strands of copper rather than a single strand because the resistance tables for SC cable is only slightly higher than THHN cable.  So what accounts for the higher resistance of 4/0 feeder?

Unless you are using Kelvin (4 wire) meter connections you are measuring the resistance of your VOM test leads.

JR

[Stephen Swaffer]

Or, using equipment Guy likely has on hand, put a heavy resistive load on the cable-say 4000-5000 watts of incandescent lighting(or more), do an accurate amp reading and check the voltage drop in the cable?

[Scott Holtzman]

Or, using equipment Guy likely has on hand, put a heavy resistive load on the cable-say 4000-5000 watts of incandescent lighting(or more), do an accurate amp reading and check the voltage drop in the cable?

This ^^^

It is also how a current shunt works.  A 1 ohm resistor in line with the positive lead provides a reference.  The voltage drop across the resistor is proportional to current draw.

[Guy Holt]

Or, using equipment Guy likely has on hand, put a heavy resistive load on the cable-say 4000-5000 watts of incandescent lighting(or more), do an accurate amp reading and check the voltage drop in the cable?

Thanks for the insight.  A modified 4-wire Kelvin test with two meters might work in my situation, but first I should probably explain the situation that lead me to measuring the resistance of 4/O feeder. I was working as set genny op on a movie this last fall and encountered an unusual electrical issue.

On multiple locations we encountered a fairly large disparity in the voltages of our three phase distribution system, leading to repeated equipment failures, and finally the breakdown of our  generator.  I was eventually able to isolate the problem to a bad piece of 4/O in our neutral run, but not until several 18kw lamps failed, and our 1400A plant blew up.  According to the tables, a 100’ stick of 4/O should have a resistance under .01 Ωs, this particular stick had an impedance of 110Ω, creating a resistive neutral that, like an open neutral, can turn a three phase distro into a voltage divider circuit where the voltage of each phase floats in an inverse relation to its load. I wrote up my experience for Protocol (use this dropbox link to the article if you are interested: https://www.dropbox.com/sh/x4q0t8c4e54jx6b/AADQmTPAOaI2zv0bGbV46nsNa?dl=0)

Later, it occurred to me that, even before energizing a set, a Rigging Gaffer can identify a resistive neutral by measuring the resistance between neutral and ground at the end of the run. Given that the resistance of a 100’ stick of 4/O should be under 01Ωs,  a reading of more than 4-6 Ω is an indication of a pinch point somewhere in this test circuit (created by bonding ground and neutral in the generator) that will cause trouble when the system is energized.

The only way I can think of modifying the 4-wire Kelvin test to detect a resistive neutral with the tools that I would have on hand is to power a 5 or 10k lamp with a 2-wire circuit (dropping the EGC) consisting of the distro neutral and ground wires.  That is I would repatch the feeders at the generator so that the neutral feeder becomes the hot and the ground feeder becomes the neutral return. Of course there could be no other loads on the distro. I could then calculate the resistance of the entire circuit (neutral wire + ground wire + lamp) by reading the current drawn by the load and the voltage drop at the generator,  and then subtract  the resistance of the lamp alone (calculated by reading the voltage drop across the lamp alone and dividing that by the same current in the circuit), to arrive at the resistance of the feeder wire alone. What this technique will not tell me, of course, is whether the high resistance is in the ground cable or in  the neutral cable.  Either is a problem from a safety standpoint and warrants closer inspection. 

Why a high resistance in a piece of neutral cable is bad is evident from my experience on set last fall.  Why high resistance in a piece of ground cable is bad is less evident.  The purpose of grounding is to prevent electrical shock.  How do you prevent a shock? By removing the source of voltage before it can cause harm. The purpose of the EGC is to be of such a low impedance that it creates a ground fault current of sufficient amperage to trip a breaker and shut the electricity off before it can do damage. In other words we want the resistance of the EGC to be as low a possible, so that it creates as much fault current as possible, to trip the breaker and shut off power as soon as possible.  If a piece of ground cable has high enough resistance, the fault current it generates may not be sufficient to trip the over-current protection, assuring that damage will occur.

Under the circumstances do you think this modified 4-wire Kelvin test is the best means of measuring the resistance of the ground and neutral feeders in a distro, or can someone think of something that would be simpler.

[Art Welter]

I was eventually able to isolate the problem to a bad piece of 4/O in our neutral run, but not until several 18kw lamps failed, and our 1400A plant blew up.  According to the tables, a 100’ stick of 4/O should have a resistance under .01 Ωs, this particular stick had an impedance of 110Ω, creating a resistive neutral that, like an open neutral, can turn a three phase distro into a voltage divider circuit where the voltage of each phase floats in an inverse relation to its load. I wrote up my experience for Protocol (use this dropbox link to the article if you are interested: https://www.dropbox.com/sh/x4q0t8c4e54jx6b/AADQmTPAOaI2zv0bGbV46nsNa?dl=0)

Guy,
Interesting article!
Curious as to what could cause that much resistance without a rather gross physical appearance.
Did you photograph the bad chunk of wire (or Camlok connection) that was responsible for an impedance of 110Ω?
Do you think the resistive point would have caused enough heat to be detected with a infrared thermometer?

Thanks,
Art

[Milt Hathaway]

can someone think of something that would be simpler.

Among many other tests, this does impedance measurements of hot, neutral, and ground legs: http://www.extech.com/display/?id=14409

No probes, it uses any IEC terminated (load side) power cable that is appropriate to your power system. Measures a single phase only, but I haven't found that to be too big of a problem.

The one I have has saved my equipment and crew more than once.

[John Roberts {JR}]

Thanks for the insight.  A modified 4-wire Kelvin test with two meters might work in my situation, but first I should probably explain the situation that lead me to measuring the resistance of 4/O feeder. I was working as set genny op on a movie this last fall and encountered an unusual electrical issue.

On multiple locations we encountered a fairly large disparity in the voltages of our three phase distribution system, leading to repeated equipment failures, and finally the breakdown of our  generator.  I was eventually able to isolate the problem to a bad piece of 4/O in our neutral run, but not until several 18kw lamps failed, and our 1400A plant blew up.  According to the tables, a 100’ stick of 4/O should have a resistance under .01 Ωs, this particular stick had an impedance of 110Ω, creating a resistive neutral that, like an open neutral, can turn a three phase distro into a voltage divider circuit where the voltage of each phase floats in an inverse relation to its load. I wrote up my experience for Protocol (use this dropbox link to the article if you are interested: https://www.dropbox.com/sh/x4q0t8c4e54jx6b/AADQmTPAOaI2zv0bGbV46nsNa?dl=0)

Later, it occurred to me that, even before energizing a set, a Rigging Gaffer can identify a resistive neutral by measuring the resistance between neutral and ground at the end of the run. Given that the resistance of a 100’ stick of 4/O should be under 01Ωs,  a reading of more than 4-6 Ω is an indication of a pinch point somewhere in this test circuit (created by bonding ground and neutral in the generator) that will cause trouble when the system is energized.

When you measure resistance between neutral and ground you are also measuring the quality of the ground-neutral bond, but high resistance there is a problem that needs attention whatever the cause.

The only way I can think of modifying the 4-wire Kelvin test to detect a resistive neutral with the tools that I would have on hand is to power a 5 or 10k lamp with a 2-wire circuit (dropping the EGC) consisting of the distro neutral and ground wires.  That is I would repatch the feeders at the generator so that the neutral feeder becomes the hot and the ground feeder becomes the neutral return. Of course there could be no other loads on the distro. I could then calculate the resistance of the entire circuit (neutral wire + ground wire + lamp) by reading the current drawn by the load and the voltage drop at the generator,  and then subtract  the resistance of the lamp alone (calculated by reading the voltage drop across the lamp alone and dividing that by the same current in the circuit), to arrive at the resistance of the feeder wire alone. What this technique will not tell me, of course, is whether the high resistance is in the ground cable or in  the neutral cable.  Either is a problem from a safety standpoint and warrants closer inspection. 

not crazy expensive (<$100).

You could trick a cheap VOM to measure low Z wire resistance by dumping a reference current into it, then measuring voltage... A couple amps of AC current could probably facilitate a usable reading. As long as the current is not flowing in the VOM leads you have replicated a Kelvin connection.

Why a high resistance in a piece of neutral cable is bad is evident from my experience on set last fall.  Why high resistance in a piece of ground cable is bad is less evident.  The purpose of grounding is to prevent electrical shock.  How do you prevent a shock? By removing the source of voltage before it can cause harm. The purpose of the EGC is to be of such a low impedance that it creates a ground fault current of sufficient amperage to trip a breaker and shut the electricity off before it can do damage. In other words we want the resistance of the EGC to be as low a possible, so that it creates as much fault current as possible, to trip the breaker and shut off power as soon as possible.  If a piece of ground cable has high enough resistance, the fault current it generates may not be sufficient to trip the over-current protection, assuring that damage will occur.

Under the circumstances do you think this modified 4-wire Kelvin test is the best means of measuring the resistance of the ground and neutral feeders in a distro, or can someone think of something that would be simpler.

Yup, any open or high Z mains feed can be hazardous. Some more than others.

JR 

[Stephen Swaffer]

So, why complicate the test?  A simple voltage test at the genny compared to the load should give you the voltage drop of the circuit.  Ohm's law will you tell you what it should be.  If its out of spec (1% or ??), then a simple voltage tests at the genny from hot to ground and ground to neutral (should obviously be zero!) compared to voltage tests at the load end from ground to hot and ground to neutral should tell you were the problem is.  Disconnecting the feeder from genny and loads and using a megger/insulation tester to verify isolation of all wires should verify no shorts exist.

I have seen aluminum 1000 MCM so corroded internally that it was extremely high resistance.  I have seen #12 SO and 4/0 weld cable that flexed so much the internal strands were all broken-but the outside looked fine.  The break could be located fairly easily by pulling the wire-if all the copper is broke, the insulation is easy to pull apart.

[ProSoundWeb Community]