Author Topic: Paralleling speaker wires and advantages.  (Read 2777 times)

Offline steve

  • Audiologist
  • *
  • Posts: 1238
Paralleling speaker wires and advantages.
« on: December 31, 2016, 01:24:59 PM »
I posted a reply on Audio Circle and my forum, and thought the information might be useful here.

just was wondering if anyone has ever taken two pairs of speakers cables and "piggybacked" them from single binding posts on amp output to single pair of binding posts on the back of a speaker, basically running 2 pairs of speaker cables in "parallel" from single binding (L & R) post amp output to single binding (L & R) posts speaker input? Why, you ask? Is because I am really liking the sound of two different sets of speaker cables and would like to combine them at the same time to see what improvement/difference happens...

Thanx, Jeff "MOONCRIKIT"

Not surprising to me, Jeff, as the inductance is significantly reduced. I am currently using (5) 18 gauge wires in parallel. Two are solid wires and three are Jenalabs 18 gauge wires. I think you have done a nice job with components and setup.

The advantages are clear in my setup. I was using 3 Jenalabs 18 gauge wires in parallel, which DC resistance equates to about 13 gauge wire. DC resistance is independent of frequency. I then added two more 18 gauge solid wires to make 5 in parallel. I have now reduced the DC resistance to about the same level as an 11 gauge wire.

However, the inductance of a straight wire is a different story. Inductance is frequency dependent. For comparison, the inductance and impedance, for 5 feet of single wire and parallel wires at 20khz are:
 
     Single                Single          Single         5 parallel
18 gauge wire          13 ga.         11 ga.           18 ga.

   2410 nh              2232 nh       2162 nh         482 nh

  .0325                   .0104           .0066           .0065              DC resistance

  .30 ohms             .28 ohms      .27 ohms       .06 ohms           Inductive reactance

I kept the capacitance to minimal, near zero which is almost always good (never say always). So while the 5 parallel 18 gauge wires equate to an 11 gauge wire in DC resistance, the inductive reactance is only 1/4.5th that of the 11 gauge wire.

(You may have heard the past arguments over the small wire parallel with the large wire improving the highs. The highs were increased because the inductance of the two wires was about 1/2 that of a single wire, thus more highs.)

Now for general consumption.

First, though, all instruments have harmonics. Usually many harmonics, such as 2nd, 3rd, 4th, 5th etc etc. A harmonic is a multiple of the fundamental signal. If a trumpet produces a fundamental of 200hz, the 2nd harmonic is 400hz, the 3rd is 600hz, the 5th harmonic is 1000hz (1khz) etc.

What gives an instrument its distinctive sound is the harmonic structure that instrument creates. A trumpet has a different harmonic structure than a violin. The 2nd harmonics on up are different amplitude and phase with a trumpet than with the violin.
 
The inductive reactance of the speaker wire will alter both the phase and the amplitude because each higher harmonic is at a higher frequency by definition. And harmonic changes will alter our perception of the note/instrument being played. Ever hear an instrument sound tinny, or bass heavy? That is because of fundamental, harmonic reproduction problems.

What is interesting is that Olsen's research demonstrated that the higher the harmonic number of a note, the more sensitive our perception is to that harmonic change. Instruments won't sound natural.

For example, if we alter the 2nd harmonic and 3rd harmonic the same amount, the alteration of the 3rd will be more noticeable than the 2nd. In this case, the 3rd is "weighted" more than 2nd harmonic. The 4th will be more "weighted" than the 3rd harmonic, the 10th more than the 2nd etc.

Back to Jeff. I can understand why you noticed a difference in your stereo system by paralleling speaker wires. In essence, you significantly lowered the inductive reactance of your speaker wires. Interesting to see what the speaker wire capacitance now measures.

Let me add, venue and component quality will also affect masking and thus one's ability to perceive differences. It appears evident to me that Jeff did a nice job with his system and setup.

Cheers and all the best.

Steve






« Last Edit: December 31, 2016, 01:42:38 PM by steve »
Steve Sammet (Owner, Electron Eng, SAS Audio Labs, Ret)
SAS "V" 39pf/m 6N copper ICs,
SAS Test Phono Stage
Acutex 320 STR Mov Iron Cart
SAS 11A Perfect Tube Preamp
SAS 25 W Ref Triode/UL Monoblocks
2 way Floor Standing Test Speakers

Offline steve

  • Audiologist
  • *
  • Posts: 1238
Re: Paralleling speaker wires and advantages.
« Reply #1 on: January 10, 2017, 09:16:08 PM »
In response to questions/comments and posts in which I could have explained and provided more information in my above post. I post portions of those responses here.

Below are great questions/comments that need addressing in order to more completely understand the subject of interconnects (ICs) in their various forms.

Quote
Separating the (+) and (-) conductors will increase the loop inductance.

In general, as we attempt to closely wrap the (+) and (-) conductors to lower the inductance, the capacitance rapidly increases, thus increasing the load on the amp at higher frequencies. Ultimately we wish both very low inductance and very low capacitance.

For example, take any amplifier and check it with a mid/high frequency square wave, and distortion analyzer, and then add output capacitance. The high frequency response falls/ the rise time slows. We also have distortion and possible resonance changes etc.......

Nordost also understands with their ribbon speaker wire. Now run 5', 10', 20' feet of closely wrapped cable and measure the capacitance and what a mid frequency square wave looks like.

I  have thought about running 8 multiple 20 gauge or 12 multiple 22 gauge conductors in parallel for even lower inductance, with still virtually no capacitance. But at 5 foot length, it seems overkill imo.

I will take very low inductance and very low capacitance VS very low inductance and high capacitance every single time.  

Quote
While we may wish for low inductance and low capacitance they are opposite sides of the same coin. If one goes up the other goes down.

The only time we have the teeter/totter effect is when they are in close proximity, such as wrapping together. Farther apart, as my example, we simply have no capacitance and the inductance of a straight wire. I listed some comparisons of straight wires in my initial post, as well as the effects of paralleling 5 conductors.

By the way, we can not perfectly cancel inductance for two important reasons.

1. The + and - return wires would have to occupy the exact same space. A short would occur.
2. Insulation and its thickness separates the wires even further. Otherwise, again, we have a short if they touch.

It is good to parallel several conductors for speaker wire to create very low inductance while making capacitance a non-factor, thus completely eliminating a variable when setting up a stereo.

At 17 feet long, I have seen as high as 15,000 pf (.015uf), which is what the amp output would see. That is very very high. Any amplifier is altered to some extent with added capacitance. I do recognize aesthetic reasons. (By the way, tube amps are even more susceptible.)

Quote
Square waves unless they are low-pass limited are rather deceiving.

Not at all. Square waves are very important in determining phase shift, resonant problems (peaking and/or ringing), more general frequency response variations, channel separation etc. Much can be learned from using low, mid and high frequency rectangular/square waves.

Many times I can correlate a rectangular wave to what I am perceiving. Of course a distortion analyzer is also helpful with sine waves, especially at high frequencies where the capacitance load would phase shift the global negative feedback signal, thus altering the distortion figures.

As I mentioned above, these are great questions/comments that needed addressing in order to more completely understand the subject of interconnects (ICs) in their various forms.

Cheers

Steve
« Last Edit: January 10, 2017, 09:21:35 PM by steve »
Steve Sammet (Owner, Electron Eng, SAS Audio Labs, Ret)
SAS "V" 39pf/m 6N copper ICs,
SAS Test Phono Stage
Acutex 320 STR Mov Iron Cart
SAS 11A Perfect Tube Preamp
SAS 25 W Ref Triode/UL Monoblocks
2 way Floor Standing Test Speakers

Offline tmazz

  • Out Of My Speaker Cabinet
  • ******
  • Posts: 12088
  • Just basking in the glow of my tubes.....
Re: Paralleling speaker wires and advantages.
« Reply #2 on: January 11, 2017, 01:37:30 PM »
Square wave can be very useful test signals because of the high amount of harmonic energy they contain. The upper harmonics are what make a square wave look square as compared to a sine wave of the same frequency and when those harmonics are attenuated the square edges of the wave will quickly start to get rounded.

Looking at a square wave at the input and output of a wire (or any other component)  can give you a very quick and simple reference as to whether it is passing all frequencies equally. You may need more sophisticated measurements to quantify the attenuation in terms of amplitudes and frequencies, but the square wave test can certainly give you a quick and dirty "if" type answer.
Remember, it's all about the music........

• Nola Boxers
• Sunfire True SW Super Jr (2)
• McIntosh MC 275
• ARC SP-9
• VPI HW-19 Mk IV/SDS/SME IV/Soundsmith Carmen Mk II ES
• Pro-Ject Pre Box S2 DAC/Rasp Pi Roon Endpoint
• DigiBuss/TWL PC&USB/MIT Cables

Offline rollo

  • Industry Participant
  • *
  • Posts: 6982
  • Rollo Audio - Home demo the only way to know
Re: Paralleling speaker wires and advantages.
« Reply #3 on: August 12, 2017, 06:26:55 AM »
Steve how does a "Litz" winding affect the inductance and capacitance ?

charles
contact me  at rollo14@verizon.net or visit us on Facebook
Lamm Industries - Aqua Acoustic, Formula & La Scala DAC- INNUOS  - Rethm - Kuzma - QLN - Audio Hungary Qualiton - Fritz speakers -Gigawatt -Vinnie Rossi,TWL, Swiss Cables, Merason DAC.

Offline steve

  • Audiologist
  • *
  • Posts: 1238
Re: Paralleling speaker wires and advantages.
« Reply #4 on: August 17, 2017, 10:28:48 AM »
Steve how does a "Litz" winding affect the inductance and capacitance ?

charles

Twisting wires will reduce inductance and increase capacitance. However, the twisting mimicks a coil, so the initial inductance, before adding the second twisted wire is higher. Since the centers of each wire cannot be occupy the same location, cancellation is not complete.

For interconnect cables (IC) dealing with high impedances, one should not worry about the inductance, but concentrate on the capacitance. My belief is that IC capacitance should be as low as possible.

For speaker wires, while capacitance should be as low as possible, we are just as concerned with the inductance. This is because the inductance is in series with the low speaker impedance. Examples. I will use 5 foot length, 18 gauge, 20khz. (Larger wires have slightly less self inductance.)

At 20khz, a 5 foot long twisted pair 18 gauge wire was measured as ~0,8uh inductance and .10 ohms inductive reactance. 

Enamel 5 foot insulated wire, twisted pair ~0,22uh inductance and 0,027 ohms inductive reactance at 20khz. Really good in that aspect.

At 20khz, 5 feet of 10 parallel 18 gauge wire has an inductance of only ~.241uh and 0,03 ohms inductive reactance. The twisted enamel and 10 parallel wires are about equal. The heavier insulated twisted is not so hot.

However, capacitance wise, the 10 parallel wires have virtually no capacitance. The enamel wire has a whopping 700pf. Any inductance in the output circuit will react with the capacitance to create a resonance easily in the audio band.

Of course the skin effect will be less as the frequency increases.

As an extra attraction, the graph shows different materials and the skin effect at different frequencies. Notice at 20khz, the skin depth of copper wire is ~0,15mm. It would take a small round 28 gauge wire to not be affected by skin effect. 28 gauge wire is 0.323mm diameter, so the resistance of the wire will not be affected by skin effect at higher frequencies. With larger diameter wire, the resistance of the wire will vary by skin effect. (The skin effect is not inductive reactance.)



Cheers

Steve
« Last Edit: August 17, 2017, 10:40:17 AM by steve »
Steve Sammet (Owner, Electron Eng, SAS Audio Labs, Ret)
SAS "V" 39pf/m 6N copper ICs,
SAS Test Phono Stage
Acutex 320 STR Mov Iron Cart
SAS 11A Perfect Tube Preamp
SAS 25 W Ref Triode/UL Monoblocks
2 way Floor Standing Test Speakers

Offline rollo

  • Industry Participant
  • *
  • Posts: 6982
  • Rollo Audio - Home demo the only way to know
Re: Paralleling speaker wires and advantages.
« Reply #5 on: August 17, 2017, 01:05:49 PM »
   You Da man, thanks got it now.


charles
contact me  at rollo14@verizon.net or visit us on Facebook
Lamm Industries - Aqua Acoustic, Formula & La Scala DAC- INNUOS  - Rethm - Kuzma - QLN - Audio Hungary Qualiton - Fritz speakers -Gigawatt -Vinnie Rossi,TWL, Swiss Cables, Merason DAC.