We have an existing 3 phase 2400V delta secondary ungrounded distribution system that we are converting to high resistance grounded with the addition of a zigzag grounding transformer. The only required wiring interconnect from the transformer to the system disconnect (breaker) consists of the equipment bonding jumper and CT circuit to the relay controlling the breaker. The current design has a copper conductor bonding jumper and CT circuit (single unshielded conductors) routed in one RMC (conduit will be bonded to bonding jumper both ends). My fear is that any ground fault current in the equipment bonding jumper (effectively the conductor and conduit in parallel) could/will induce errors in CT circuit. Shouldn't the CT circuit be routed in its own conduit? Any help appreciated.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Equipment Bonding Jumper in Same Conduit as Instrumentation Circuit
- Thread starter bozo
- Start date
- Status
- Not open for further replies.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
130329-0832 EDT
To greatly reduce the induced error voltage in the CT circuit use a twisted wire pair for each CT. Note: that an electrostatic shield around the twisted pair will have virtually no effect on a magnetic field at power line frequencies. So whether the twisted wire pair is shielded or not wont matter.
Why does a twisted pair help? Each twist can be considered a one turn coil. The magnetically induced voltage in each twist is 180 degrees out of phase with its adjacent coil. If the magnetic field over many coils is approximately the same, and each coil has about the same cross-sectional area, then the induced voltage in each coil cancels the voltage in the adjacent coil, and the net error voltage from the twisted cable is very small.
To see the effectiveness of a twisted pair you can run an experiment. Short one end of a twisted wire pair and put an microvoltmeter at the other end, may require some high frequency filtering to the meter. See what the meter reads.
.
To greatly reduce the induced error voltage in the CT circuit use a twisted wire pair for each CT. Note: that an electrostatic shield around the twisted pair will have virtually no effect on a magnetic field at power line frequencies. So whether the twisted wire pair is shielded or not wont matter.
Why does a twisted pair help? Each twist can be considered a one turn coil. The magnetically induced voltage in each twist is 180 degrees out of phase with its adjacent coil. If the magnetic field over many coils is approximately the same, and each coil has about the same cross-sectional area, then the induced voltage in each coil cancels the voltage in the adjacent coil, and the net error voltage from the twisted cable is very small.
To see the effectiveness of a twisted pair you can run an experiment. Short one end of a twisted wire pair and put an microvoltmeter at the other end, may require some high frequency filtering to the meter. See what the meter reads.
.
Last edited:
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
130320-0906 EDT
With about 500 ft of Belden 8205 I am in the noise level of my meter, 10 microvolts.
Edit time timed out.
.
With about 500 ft of Belden 8205 I am in the noise level of my meter, 10 microvolts.
Edit time timed out.
.
- Location
- Placerville, CA, USA
- Occupation
- Retired PV System Designer
130329-0832 EDTEach twist can be considered a one turn coil.
Not trying to be picky, but if you want them to cancel out, you need to say that each half-twist can be considered a one turn coil (made up of two wires segments.)
If the wires are so short that you only have room for one full twist, that will still help.
The pitch of the twist is important in relation to the overall length of the wires and how close the noise and signal wires will be. Twisting both noise and signal wire pairs individually helps even more.
If the EBJ is not carrying current, it should not cause any problems. A neutral, on the other hand, might.
Much thanks to gar and GoldDigger. While it might take me awhile (possibly forever) to comprehend the theory involved, it appears that you did confirm that my fear was valid. Since my original post, the decision has been made to put the EBJ in it's own conduit. Thanks again.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
130322-0917 EDT
bozo:
I like seeing the current transformer outputs in a separate conduit, but my description of the use of a twisted pair to minimize magnetically induced errors is not the reason. Rather my description is why you could run the CT wires, if twisted, with your other wires and probably see no adverse effects.
A different situation --- When you want to reduce crosstalk between adjacent twisted pairs, such as in a CAT-5 cable, then adjacent pairs need to have different twist pitches. Strip back some CAT-5 and see how the four pitches differ. There are four separate twisted pairs in a CAT-5 cable.
.
bozo:
I like seeing the current transformer outputs in a separate conduit, but my description of the use of a twisted pair to minimize magnetically induced errors is not the reason. Rather my description is why you could run the CT wires, if twisted, with your other wires and probably see no adverse effects.
A different situation --- When you want to reduce crosstalk between adjacent twisted pairs, such as in a CAT-5 cable, then adjacent pairs need to have different twist pitches. Strip back some CAT-5 and see how the four pitches differ. There are four separate twisted pairs in a CAT-5 cable.
.
130322-0917 EDT
bozo:
I like seeing the current transformer outputs in a separate conduit, but my description of the use of a twisted pair to minimize magnetically induced errors is not the reason. Rather my description is why you could run the CT wires, if twisted, with your other wires and probably see no adverse effects.
A different situation --- When you want to reduce crosstalk between adjacent twisted pairs, such as in a CAT-5 cable, then adjacent pairs need to have different twist pitches. Strip back some CAT-5 and see how the four pitches differ. There are four separate twisted pairs in a CAT-5 cable.
.
gar,
Thanks for the clarification. In our situation the switch to separate conduits was logistically driven also. It appears that I should utilize a twisted pair for the CT circuit even though it is routed in a dedicated conduit so as to minimize any possible interference from adjacent wiring in the switchgear in which it is terminating. Thanks also for the CAT-5 info. Might be able to impress some friends with that knowledge. Thanks again for the help.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
130326-1349 EDT
bozo:
Yes, run signal wires as twisted pairs. This is why so many Belden cables contain twisted pairs. A single pair or multiple twisted pairs. 8723 is two separately twisted pairs, and each pair is individually shielded, but only one drain wire.
Also a coaxial cable gives you a similar cancellation effect if the magnetic field is fairly uniform from side to side of the coax. Notice how the magnetic field on one side of the center wire induces a voltage in the center wire that is 180 out of phase with that from the opposite side of center.
At one time long ago I wanted to provide a large overload current to some circuit breakers, and I need to record the current. I used about a foot long piece of Nichrome bar stock. Connected the two ends to a scope via a piece of coax. The termination of the coax to the bar was initially such that a moderate loop was produced.
The waveform on the scope was totally incorrect. Why? Because the 10,000 amperes or so produced a sufficient magnetic field to induce a voltage in the one turn loop. This showed up as a sharp initial pulse. e = N * df/dt. The solution was to tape the coax center wire tight to the resistance bar (the current shunt). This made a very small cross-section one turn loop, and essentially eliminated the problem.
An even better solution would be to use a tubular shunt with the voltage sensing lead in the center of the tube.
.
bozo:
Yes, run signal wires as twisted pairs. This is why so many Belden cables contain twisted pairs. A single pair or multiple twisted pairs. 8723 is two separately twisted pairs, and each pair is individually shielded, but only one drain wire.
Also a coaxial cable gives you a similar cancellation effect if the magnetic field is fairly uniform from side to side of the coax. Notice how the magnetic field on one side of the center wire induces a voltage in the center wire that is 180 out of phase with that from the opposite side of center.
At one time long ago I wanted to provide a large overload current to some circuit breakers, and I need to record the current. I used about a foot long piece of Nichrome bar stock. Connected the two ends to a scope via a piece of coax. The termination of the coax to the bar was initially such that a moderate loop was produced.
The waveform on the scope was totally incorrect. Why? Because the 10,000 amperes or so produced a sufficient magnetic field to induce a voltage in the one turn loop. This showed up as a sharp initial pulse. e = N * df/dt. The solution was to tape the coax center wire tight to the resistance bar (the current shunt). This made a very small cross-section one turn loop, and essentially eliminated the problem.
An even better solution would be to use a tubular shunt with the voltage sensing lead in the center of the tube.
.
- Status
- Not open for further replies.