Jneutron, inductance of twisted vs parallel cables. [Archive] - Audio & Video Forums

PDA

View Full Version : Jneutron, inductance of twisted vs parallel cables.



Tony_Montana
05-19-2004, 03:08 PM
This is the follow up to this thread from CA:

http://www.audioasylum.com/audio/cables/messages/92977.html

I mentioned that if we twist two single wires together, it will have lower inductance than if we left it as parallel configuration (like zip cord).

Your response was:"The inductance will be dependent on the spacing between the wires, and if they are already touching, or spaced with something, twisting them will not alter the spacing."Jn

Did a little bit of research on this subject, and your statement might not be true. I ran cross an article (sorry I lost the link, but I will keep looking) mentioning that twisted wires will have lower inductance than parallel cables (given the same spacing between two conductors).

The article mentioned that the close magnetic coupling between the wires due to twisting [them] will reduces the inductance of the cable, and that this is why twisted pairs have lower inductance than parallel wires.

I see if I can find that link :)

skeptic
05-19-2004, 04:08 PM
Without doing any calculations or cracking open any of my old books, not only doesn't this make sense, but because if you twist them, the number of feet of actual wire for a given footage of running distance would increase, therefore if anything, the inductance per unit length would increase. The tighter the twist, the greater the increase. BTW, by the same logic, so would the capacitance and the resistance. On the other hand, induced hum and noise which for a loudspeaker circuit is usually far below audibility anyway, would decrease. So by twisting the wire, you give up something to get nothing.....unless your intent all along was to create an LCR filter.

Tony_Montana
05-19-2004, 08:16 PM
Without doing any calculations or cracking open any of my old books, not only doesn't this make sense, but because if you twist them, the number of feet of actual wire for a given footage of running distance would increase, therefore if anything, the inductance [and resistance] per unit length would increase.

Hey Skep

You make good argument (especially about the resistance) which can be over come by using larger gauge wire. But for inductance (as you mentioned), we may have double edge sword. So now the question becomes would twisting a pair of wires would reduce inductance to over come excessive inductance produced by extra length of wire due to twisting?

Dam it Skep, why did have to make things complicated :D

mtrycraft
05-19-2004, 08:55 PM
Hey Skep

Dam it Skep, why did have to make things complicated :D

Because it is not a simple one :D
By the way, haven't seen him posting in a while.

skeptic
05-20-2004, 02:44 AM
With a carefully matched load impedence of 75 or 300 ohms fed by an antenna designed to capture it as its voltage source, you might get 50, 100, or even 200 microvolts induced rf in a circuit deliberately designed for that purpose. The power gain between that front end tuner circuit and a loudspeaker is in the millions, billions, even trillions. In a loudspeaker circuit where the source impedence (the power output stage) is nearly zero, the gain is zero, and the load is 8 ohms, there will be NO induced audible noise. If you want to hear how much noise is induced in that circuit by rf, turn your amplifier off and put your ear up to your loudspeaker. If rf noise is being induced by inductive coupling into your sound system and detected so that it is audible, it is happening at a much earlier gain stage. Like the high gain phono input, mic, or tape head (not tape amp) circuit. That's the place to look for hum and noise and the circuit that needs the most rf shielding. Also be sure your tuner is turned off when not in use so that its audio output signal is not leaking past the function selector switch on your preamp. How do I know? For years I lived on the infield of WTFM which was six blocks away. You could get it from the phono stage, channel six on your television set, and some people even got it through the mercury fillings in their teeth. But it never came in as the result of the speaker wire alone.

jneutron
05-20-2004, 05:30 AM
Your response was:"The inductance will be dependent on the spacing between the wires, and if they are already touching, or spaced with something, twisting them will not alter the spacing."Jn

Did a little bit of research on this subject, and your statement might not be true. I ran cross an article (sorry I lost the link, but I will keep looking) mentioning that twisted wires will have lower inductance than parallel cables (given the same spacing between two conductors).

The article mentioned that the close magnetic coupling between the wires due to twisting [them] will reduces the inductance of the cable, and that this is why twisted pairs have lower inductance than parallel wires.

I see if I can find that link :)

Cool. It wouldn't be the first time I was wrong.....

Hope you can find the link..it would be interesting reading..bet the article is incorrect, though..hee hee...

Short explanation:

Given real life situ's, as in you don't double the amount of wire used when you twist them, the dipole field produced by the wire does not get any bigger or smaller. Since the inductance of the pair is the result of that dipole field, if it does not change, neither does the inductance. (inductance is the relation between energy stored, and current).

If you REALLY twist them, the cross sectional angle of the wire will begin to get nuts, and there will be a solenoidal component of the field as a result of the angle of the wire. But, keep in mind, it's mate is in the other direction, and the result is local cancellation of the solenoid. And, at the same time, the dipole field normal to the center of axis is lowering at the same time because of that angle.

So, I stand by my assertion..

OH...almost forgot...I tested this using #24, #18, and #16zip, as well as a #10 twisted pair. The inductance did not change regardless of the twist pitch I used.

So, if you twisted the wire to make it shorter, then, yes, it would make the inductance per useable length lower. My assertion stands only for reasonable twists..

Cheers, John

Tony_Montana
05-20-2004, 08:26 PM
By the way, haven't seen him posting in a while.

Don't worry Mtry, we knew Skep is going to be back shortly. It is like feeding a stray cat <img src="http://www.fansofrealitytv.com/forums/images/smilies/catsmiley.gif">. Once you feed them, they will never leave :D



In a loudspeaker circuit where the source impedence (the power output stage) is nearly zero, the gain is zero, and the load is 8 ohms, there will be NO induced audible noise.

Are you sure Skep? Check out this post where the poster complained of Hum (caused by nearby power cord) coming thru speaker wires :)

http://forums.audioreview.com/showthread.php?t=3569



OH...almost forgot...I tested this using #24, #18, and #16zip, as well as a #10 twisted pair. The inductance did not change regardless of the twist pitch I used.

Well, if you test it and measured it, then there are no room for arguments.

By the way, I found the link by Dr. Thompson. 2/3 down, he still claims that twisted-pairs have lower inductance than parallel wires :)

http://members.aol.com/marctt/CV/Abstracts/inductance.htm

skeptic
05-21-2004, 02:10 AM
As implausible as it sounds, I will concede that it is conceivable that in this rare and easily correctable case, it could happen. For 60 hz induced hum from 120 volts. I was referring in my posting to rf noise, the most common concern of induced noise among inexperienced audiophiles. You can see that if the electrical field of 120 volts at 60 hz in close proximity over much of the length of speaker wire can induce a barely audible hum which is eliminated when the source is moved just a few inches, the possibility that this could occur from rf fields with a field strength that can induce barely a few dozen or a few hundred microvolts is non existant. If induced noise were anything but a rare problem with speaker wire, even common wire would be sold shielded and would add to the cost only minimally to say, 16 or 12 gage zip cord.

jneutron
05-21-2004, 05:07 AM
Are you sure Skep? Check out this post where the poster complained of Hum (caused by nearby power cord) coming thru speaker wires :)

There was also that effect reported over at AA Tech.

My suspicion is that there are heavy neutral currents flowing through his building as a result of the distributed neutral earth grounding on the telephone poles in the neighborhood, and the primary to secondary strap on all the line transformers, causing spurious distribution currents to flow. Most likely through the plumbing system. And that neutral current is causing the 60 hz field which is being picked up by the speaker run. The guy I discussed it with noted a radiator parallel to the speaker run, but he did not follow through on finding what was going on.


By the way, I found the link by Dr. Thompson. 2/3 down, he still claims that twisted-pairs have lower inductance than parallel wires :)
http://members.aol.com/marctt/CV/Abstracts/inductance.htm

Thanks..I just read it..Here's the relevant text:

(quote)""For l = 0.5 meter and a wire-wire spacing d = 1 cm, results are: Lo = 0.485 microH for 14 gauge; Lo = 0.53 microH for 16 gauge and Lo = 0.58 microH for 18 gauge. Therefore, for the parallel-wire line with closely-spaced conductors, the inductance is approximately 0.5 microHenries per meter of total wire length.

We can hone the guesstimate further by applying a little physical reasoning. For the parallel-wire case with close wire-to-wire spacing, the close magnetic coupling between the wires reduces the inductance of the loop (this is why twisted-pairs have lower inductance than parallel wires). "" (end of quote)

Note that he calculated the inductance for "closely spaced conductors", and he defined closely spaced as "1 cm". then, he describes twisted pairs, which are not 1 cm spaced, but touching each other..so he is not speaking about the same spacing conditions..

Cheers, John

skeptic
05-21-2004, 09:13 AM
How does the worst case inductance of speaker cable compare to the inductance in most loudspeakers? How about including the huge series chokes used for low pass filters in the woofer circuit? How does it compare to the inductance in the output transformer of a vacuum tube amplifier? Speaker wire inductance by comparison probably doesn't add up to a hill of beans. That is unless somebody deliberately wanted to build an LCR filter in the shape of a speaker wires and figured out how. Sooner or later somebody will and somebody else will tell you how great it sounds.

jneutron
05-21-2004, 09:43 AM
How does the worst case inductance of speaker cable compare to the inductance in most loudspeakers?.

Which driver?


How about including the huge series chokes used for low pass filters in the woofer circuit?.

I believe that only one of the drivers in a 2 way system has a huge inductor in series with it. And to the best of my recollection, lateralization is not huge for the bass.


How does it compare to the inductance in the output transformer of a vacuum tube amplifier?

The inductance if there is no feedback, or open loop?


Speaker wire inductance by comparison probably doesn't add up to a hill of beans.

Probably? you seemed so sure for a while there.. :-)

Cheers, John

Monstrous Mike
05-21-2004, 10:20 AM
There are different types of EMI. There is radiated EMI and conducted EMI. Radiated EMI needs antennae to transmit or recieve signals. When we talk about wires and cables being antennae we generally mean dipole omni-directional antennae with tuning at a quarter wavelength.

The frequencies we are talking about which is 60 Hz for AC power and 20-20000 Hz for audio signals would require quarter wavelength dipole antennae of proposterous lengths to receive conducted EMI.

Maybe I'm missing something here but could somebody explain how 60 Hz could be radiated from a cable inside somebody's house? Does anybody own a speaker wire that is 5 miles long so it can receive signals in the 20 kHz region?

jneutron
05-21-2004, 11:11 AM
There are different types of EMI. There is radiated EMI and conducted EMI. Radiated EMI needs antennae to transmit or recieve signals. When we talk about wires and cables being antennae we generally mean dipole omni-directional antennae with tuning at a quarter wavelength.

The frequencies we are talking about which is 60 Hz for AC power and 20-20000 Hz for audio signals would require quarter wavelength dipole antennae of proposterous lengths to receive conducted EMI.

Maybe I'm missing something here but could somebody explain how 60 Hz could be radiated from a cable inside somebody's house? Does anybody own a speaker wire that is 5 miles long so it can receive signals in the 20 kHz region?

You are referring to the efficient transmission of electrical energy by way of electromagnetic waves. Antennae theory 101.

We are not. We are discussing the coupling of two parallel circuits through mutual inductance. Since this coupling relies only on the capture of magnetic flux within a loop of conductor, the length of the conductor does not have to be on the order of a wavelength to do so. Transmission efficiently into free space certainly does require adequate antennae lengths as you state... mutual inductance does not.

Cable tray systems are a good example of what we speak of..We are not allowed (at work)to split a three phase delivery system into different cable trays, instead, we have to make sure each cable tray has a balanced current condition, as the external fields possible when the phases are placed into different trays will produce tray heating...by keeping three phases together, tray dissipation is limited.

Another good example is the 220 panel in your house..I had a cap go bad on an AC unit..the compressor drew locked rotor current every time it tried to start...That current surge coupled to the front hinged panel of the service, and caused buzzing..the hot wires broadcast a magnetic field, but was not transmitting energy into space in any appreciable amount because of length, yet it put energy into the front panel.

Cheers, John

Monstrous Mike
05-22-2004, 01:16 PM
You are referring to the efficient transmission of electrical energy by way of electromagnetic waves. Antennae theory 101.

We are not. We are discussing the coupling of two parallel circuits through mutual inductance. Since this coupling relies only on the capture of magnetic flux within a loop of conductor, the length of the conductor does not have to be on the order of a wavelength to do so. Transmission efficiently into free space certainly does require adequate antennae lengths as you state... mutual inductance does not.

Cable tray systems are a good example of what we speak of..We are not allowed (at work)to split a three phase delivery system into different cable trays, instead, we have to make sure each cable tray has a balanced current condition, as the external fields possible when the phases are placed into different trays will produce tray heating...by keeping three phases together, tray dissipation is limited.

Another good example is the 220 panel in your house..I had a cap go bad on an AC unit..the compressor drew locked rotor current every time it tried to start...That current surge coupled to the front hinged panel of the service, and caused buzzing..the hot wires broadcast a magnetic field, but was not transmitting energy into space in any appreciable amount because of length, yet it put energy into the front panel.

Cheers, John
I understand your point. Lots of people say speaker wires act as antennas but there is only a certain frequency range they can pick up and it is nowhere close to the audio band.

I did an experment once where I took a 50 foot AC extension cord and duct taped it to my rear speaker wire (about a 30 foot run). I then hooked up the amp power to this extension cord and really cranked the volume. I couldn't here any interference in that speaker with the AC running in parallel to it.

I would suspect the mutual inductance and radiated RFI are insignificant in home audio systems and the main 60 Hz interference problems would be due to ground loops or other <i>conducted</i> noise.

I have also tried speaker wire with RCA plugs as interconnects and even as a digital audio cable and have had no success in discovering or hearing any RFI.

The only real problems I've ever had with RFI in my home have been related to video signals.

Tony_Montana
05-22-2004, 03:40 PM
Note that he calculated the inductance for "closely spaced conductors", and he defined closely spaced as "1 cm". then, he describes twisted pairs, which are not 1 cm spaced, but touching each other..so he is not speaking about the same spacing conditions.

Thanks John. That made it much clearer :)



How does the worst case inductance of speaker cable compare to the inductance in most loudspeakers?

On the short cable run, probably it doesn't matter. But on the longer run (35-40 feet), calculation has shown that we will have 1 dB drop at 20 kHz due to cable inductance. You also have to consider other losses such as cable resistance which will be added to cable inductance (which mean combined losses due to inductance and resistance will be greater than 1 dB).

So it always best to avoid losses whenever possible as to keep combined losses below 1 dB :)



I did an experment once where I took a 50 foot AC extension cord and duct taped it to my rear speaker wire (about a 30 foot run). I then hooked up the amp power to this extension cord and really cranked the volume. I couldn't here any interference in that speaker with the AC running in parallel to it.

Mike, would the result be same if speaker wire were running parallel with power cord that is attached to [high current] appliances that have motor such as airconditioner, heater or heat radiator?

I think I mentioned this before, but I had a hum problem that was coming thru my ICs. And guess where the hum was coming from? It was coming from the TV since IC was running too close to back of TV. Once I reroute the cable, hum was gone. So hum and interference might come thru the system from a place that we might expect the least :)

jneutron
05-24-2004, 07:04 AM
I did an experment once where I took a 50 foot AC extension cord and duct taped it to my rear speaker wire (about a 30 foot run). I then hooked up the amp power to this extension cord and really cranked the volume. I couldn't here any interference in that speaker with the AC running in parallel to it.

Bet ya used an extension that was twisted internally...

And the speaker wire wasn't..so 60 hz coupling would by design be very low.

Try repeating it with romex. It's dipole field doesn't spiral, so it can couple the entire length..

And, try running a kilowatt of lights or heaters in the cord while it's alongside the speaker run...that way, the audio doesn't cover the hum...as the hum would only show up during draw..

Cheers, John