Twisting two wires VS paralleling wires for lowest inductance

Thread Starter

sasaudio

Joined Oct 18, 2009
5
I know this is probably a difficult subject, but I hope for some help.

I don't have equipment, no technical knowledge anymore after severe traffic accident some years ago. I will attempt to be clear, and step by step.

I see that solid 18 gauge, 5 foot length "X" of copper straight wire has inductance of ~ 2410 nh. If I add another 5 foot length of 18 gauge solid wire "Y" for return path, we have an additional inductance of 2410 nh.

Next, I can go two ways. The first is to twist the two wires, "X" and "Y" together to reduce the inductance. I know insulation thickness, turns per inch will play a role in inductance cancellation due to centers distance from each other, but I don't have an insulation thickness value yet.

Next, another way is to keep the two wires, "X" and "Y" very separate and simply parallel additional 18 gauge wires to "X" and to "Y", to reduce the inductance of "X" and of "Y". Say 5 or 10 wires in parallel. I figured that 10 wires in parallel for each of "X" and "Y" would reduce the inductance to approximately 241 nh times 2 ("X" and "Y").

My question is can twisting the single "X" and "Y" wires together really reduce the inductance as much as paralleling 5 or 10 straight "X" wires and paralleling 5 or 10 straight "Y" wires? (I would also separate the "X" parallel wires and "Y:" as well.)

I appreciate your help and time very much.

steve sammet
 

Thread Starter

sasaudio

Joined Oct 18, 2009
5

Danko

Joined Nov 22, 2017
990
"Speaker Cable should always be twisted. Witnessed a System with Parallel wire (Zip Cord) in a home where the receiver and all electronics were unplugged and removed or un-wired from the speaker wires and yet the speakers were clear as a bell still working. While the volume was very low, the home owner was distressed as they could always hear it. Why? Well because the home was wired with UN-Twisted speaker wire and the AM radio station in town was being received by the long wire antenna (Speaker Wire) and transduced in the speakers. So yes ALL Speaker/Audio wire should be Twisted. "
https://electronics.stackexchange.com/questions/43246/is-unshielded-untwisted-pair-cabling-fine-for-audio
I am sorry, calculator in previous link is not working.
Here is working one: http://www.multek.se/index.php?page=twisted-pair-impedance
 
Last edited:

ebp

Joined Feb 8, 2018
2,332
In order for paralleled conductors to reduce inductance the individual conductors would have to be all be separated from each other by distances sufficient to prevent their magnetic fields from interacting.
 

dendad

Joined Feb 20, 2016
3,602
In order for paralleled conductors to reduce inductance the individual conductors would have to be all be separated from each other by distances sufficient to prevent their magnetic fields from interacting.
I thought the idea was to have the magnetic fields coupled so they cancel out, therefore reducing the inductance. If there is no magnetic field interaction the inductance will not be changed.
 

ebp

Joined Feb 8, 2018
2,332
"I thought the idea was to have the magnetic fields coupled ..."

Sorry, yes, if you are talking about conductors carrying current in opposite directions then they should be as close together as possible so the magnetic fields cancel and the inductance is reduced. I was referring to the proposed case of multiple paralleled conductors carrying current in the same direction, in which case you want no magnetic interaction if you are seeking to reduce the overall inductance.
 

Thread Starter

sasaudio

Joined Oct 18, 2009
5
"I thought the idea was to have the magnetic fields coupled ..."

Sorry, yes, if you are talking about conductors carrying current in opposite directions then they should be as close together as possible so the magnetic fields cancel and the inductance is reduced. I was referring to the proposed case of multiple paralleled conductors carrying current in the same direction, in which case you want no magnetic interaction if you are seeking to reduce the overall inductance.
I think you are hitting on what I am talking about. (I am not worried about RFI etc.) Let's use "X" for one speaker wire, and "Y" for the other speaker wire.

If one twists "X" and "Y" wires together, we obtain only partial cancellation of inductance since we cannot have the centers of the two wires in the same spot relative to each other. We can only place the wires next to each other with some sort of distance in between. The calculators take this into account.

Now take the straight wire inductance of "X" wire and "Y" wire and parallel each wire with 9 other wires. So we have 10 "X" wires and 10 "Y" wires.

The straight wire inductance of "X" is reduced by 1/10th, and the same with "Y". Can we obtain the same low inductance by twisting two wires at audio frequencies?

--------
New.
Something pointed out to me is that the calculators for parallel and twisted wires indicate the farther apart I place "X" and "Y" wires from each other, the higher the inductance.

I set the distance between the 1 inch length "X" and "Y" wires at 999999 inches and the inductance sky rockets beyond belief. The magnetic field cannot possibly cause such inductance with the wires so separated.

As the distance between the parallel wires increases, should they not eventually mimmick straight wires, at audio frequencies? If so, then the maximum inductance at maximum wire separation would be the straight wire inductance of "X" and "Y" wire.

Now as a single straight wire, a 1 inch length is only some 19 nh/inch. So if I take two 5 foot lenghs, connect them together and form a 3.8 foot loop, how can the calculator figure 76 nh/inch?

I get a funny feeling frequency is involved with the parallel and twisted wire calculators.

thanks again for your input
steve
 
Last edited:

dendad

Joined Feb 20, 2016
3,602
Really, as you say you are dealing with audio frequencies, those values are almost irrelevant.
The main thing to worry about for speaker cables is the conductor resistance.
I think the losses you will measure due to the reactive components of the cable will be almost immeasurable and definitely not noticeable.
Here is a handy reactance table.
http://www.rfcafe.com/references/electrical/images/Frequency-Reactance-Nomograph.zip
 

Danko

Joined Nov 22, 2017
990
Really, as you say you are dealing with audio frequencies, those values are almost irrelevant.
The main thing to worry about for speaker cables is the conductor resistance.
I think the losses you will measure due to the reactive components of the cable will be almost immeasurable and definitely not noticeable.
Audioguru concern about harmonics up to 300kHz. They believe they feel them.
It is why they afraid even smallest inductance of speaker wires.
 

dendad

Joined Feb 20, 2016
3,602
Audioguru concern about harmonics up to 300kHz. They believe they feel them.
It is why they afraid even smallest inductance of speaker wires.
Even at 300kHz there will be minimal problems. And I am one of those who are very skeptical of sensing that frequency range. Along with the need for low oxygen cables and all the other ingenious marketing hype stuff.
 

Danko

Joined Nov 22, 2017
990
Even at 300kHz there will be minimal problems. And I am one of those who are very skeptical of sensing that frequency range. Along with the need for low oxygen cables and all the other ingenious marketing hype stuff.
Absolutely agree with you. But they are kind of religious and nobody can dissuade them.
 

Thread Starter

sasaudio

Joined Oct 18, 2009
5
http://diyaudioprojects.com/Power/Low-Inductance-DIY-Speaker-Cables/
Low-Inductance DIY Hi-Fi Speaker Cable Construction.
Average results for the 13 foot length:

  • Resistance (R) = 0.0125 Ohms
  • Inductance (L) = 0.600 uH
  • Capacitance (C) = 0.850 nF
First, any alterations will have to do with frequency response (FR) changes. In otherwards tonal balance changes, not spl changes per se.
I am not sure how you came to those figures, Danko. I used the calculators for twist/parallel, and single wire, but as explained below, there is a divergence between the calculators.

----- Quote.
" Here is an example of twisted.

Wire size 18 gauge, dia is 1.023 mm
Separation between centers of two wires: 3 mm
Insulation between wires: ~1mm

Inductance: 18 nh/inch, 1080 nh/5 feet, 2160 nh/10 feet

4 mm separation Inductance is 21 nh/inch, 1260 nh/5 feet, 2520 nh/10 feet

What about the heavy two wire cables, 6, or even 10 mm between centers? 6 mm is 3000 nh (3 uh) for 10 feet, 10 mm is 3624 nh for 10 feet.

However, for 10 parallel wires the inductance is only 482 nh for 10 feet. "
----- End of quote.

As one can see, your figures are quite different from mine.

2520 nh (2.52 uh) at 20khz gives a reactance of 0,317 ohms reactance. In my above example, for 10 parallel wires, the inductive reactance at 20khz is only 0,06 ohms reactance.

Let's calculate using 9 gauge wire with descent insulation. Inductance is going to be around 3500 nh for 10 feet. At 20khz the reactance is approximately 0,44 ohms.

In some 4 years of testing with consumers, I have been changing the value of a 9k ohm resistor (placed across a full range driver) by some 0,02 ohms, and individuals have perceived a sonic difference, mainly how bright, clarity, and the attack times (rise times).

I am pretty sure the twisted and parallel conductor calculators are meant strictly for RF, usually above a few khz. I get the feeling the calculators are inaccurate at audio frequencies, but not sure. I am also coming to the solid conclusion that using parallel wires instead of twisting two huge wires is advantageous.

The inductance should be small with tight twist, close parallel. As the distance increases, the inductance will rise until the two separated wires act like individual straight wires. Yet the two calculators, twisted/parallel, and single straight wire calculators diverge on this point.

I think another plus that is good engineering is ridding of unnecessary capacitance by paralleling multiple wires yet separating "X" from "Y", even if it appears not to cause any problem. I don't have any RFI etc problems, even though I live within 6 miles from TV transmitter antenna.

Cheers and I continue to appreciate feedback.

steve
 
Last edited:

Danko

Joined Nov 22, 2017
990
I am not sure how you came to those figures, Danko.
These parameters are measured. See PDF in attachment, last page.
I am pretty sure the twisted and parallel conductor calculators are meant strictly for RF, usually above a few khz. I get the feeling the calculators are inaccurate at audio frequencies, but not sure.
Capacitance and inductance of wire lines are independent from frequency.
I am also coming to the solid conclusion that using parallel wires instead of twisting two huge wires is advantageous.
Twisted pair has small inductance L.
Twisted pairs do not interact magnetically.
Two twisted pairs, connected in parallel, have inductance L/2 (Does not matter how close they are, because NO magnetic interaction).
In PDF below four twisted pairs in parallel used (L/4) for one speaker.
It is exactly what you want, but without distance between wires requirement.
 

Attachments

Last edited:

MisterBill2

Joined Jan 23, 2018
7,031
In the area of instrumentation and connecting sensors we twist the wires, BUT NOT TO CHANGE INDUCTANCE!! Using a twisted pair reduces the differential noise pickup from external noise fields. And that does matter. The frequency response is affected by the capacitance between the wires, so that is why the selection of insulation materials matters. In speaker cables the resistance matters because the impedance of the speaker is very low. The entertaining thing is that I have met folks who claim to be able to tell the difference between their $200 speaker cables and my #12 stranded wire speaker cables. But I guess that if they paid as much for the amplifier as I paid for my car they are entitled to believe what they wish.
 

Thread Starter

sasaudio

Joined Oct 18, 2009
5
Sorry for being so late, but I had to leave. I have to wonder what, if any research has been performed to come to the conclusion that inductance and capacitance do not matter, or that total gauge does not matter. Unfortunately, readers of this string may have gotten the wrong advice concerning speaker wires.

For the record, the problem with the above data is that in order to reduce inductance, capacitance rises. In your example, the capacitance varies with dia of the wire and insulation thickness. Capacitance is always stored energy that, in this type of scenario, should be minimized. Capacitance and inductance can cause resonance problems, and can easily alter global negative feedback so as to be detrimental to musical reproduction. Nelson Pass understands this quite well if one studies his work.

Inductance and capacitance always influences the rise time. Multiple, studies, involving multiple universities, medical experts etc, have concluded that 5us (microseconds) alteration in slope is perceived by the ear. Jneutron (FermiLab, Cern, Brookhaven National Laboratory fame, music lover) remembered a descent study (as I have confirmed) that 2us slope (rise time) change was detectable by one's ear..This type of change is well above 20khz (although a musical signal needs to be present since we are altering the signal.)

My inexpensive speaker wires ( parallel 18 gauge door bell wire) "sound" different than a single ~8 gauge wire, or other gauge wire.

cheers
steve
 

crutschow

Joined Mar 14, 2008
25,432
My inexpensive speaker wires ( parallel 18 gauge door bell wire) "sound" different than a single ~8 gauge wire, or other gauge wire.
That's pretty much what a friend noticed. He also claimed solid sounded better than stranded wire.
I will only believe those claims if they are done in a double-blind test where they can rapidly switch back and forth between the two types of wires, and no-one knows which is which. :rolleyes:
 
Top