Zip Cord VS Cross Connected Coax Article

[Jon Risch]

<font color='#000000'>Gene recently wrote yet another article on audio cables, this one specifically about one of my original and unique speaker cable designs, the DIY Cross-Connected 89259 speaker cable.

This was apparently done in an attempt to defend Gene's comments about my DIY speaker cable design, when he originally made the following statement:
&quot;Please refrain from posting links to Jon Rischs cable receipts as we feel that his designs can do more harm than good.&quot;

I replied, and Gene defended SOME of his comments, in this thread:

http://www.audioholics.com/cgi-bin....4;t=483

This post is my comments on his article regarding my speaker cable design.

The first point of contention, is the actual inductance of a typical 12 ga. zip cord. &nbsp;Gene's reference is the original Monster 12 ga., which is just a clear PVC jacket, bare copper type of zip cord. &nbsp;

Gene says my measured amount of inductance is incorrect.

This depends on the definition of what a typical 12 ga. zip cord is. &nbsp;Some zip cords have a fairly close center to center spacing of the wire conductors, this tends ot lower the overall inductance. &nbsp;Belden 9718 is one example of this.

Then there are the hardware store or electrical supply store zip cords, which is what I am calling typical, that do not have the very close spacing, but a more nominal amount of spacing.

Then there are other cords that have greater than typical spacing, such as some of the Radio Shack flattened zips, and some of the so-called 'premium' zip cords, that use a larger than normal amount of insulation and spacing, so as to make the cable appear larger than it really is. &nbsp;Some have a distinct spacing web down the middle, that tends to separate the two conductors even more than they would be if just the insulation was present side by side. &nbsp;I am not talking about the normal very small web that all zip cords tend to have, but a wider than normal web.

Since I do not have a sample of Original Monster Cable handy any more, I can not re-measure it, but earlier measurements indicated it fell into the middle group, and was not unusual in its LCR parameters.

My own measurements for it, and for most typical 12 ga. zip cords came up with approx. 0.25 uH per foot of inductance. &nbsp;Recent measurements by jneutron/John Escalier posted at AR of several different zip cords indicated that they measured around 0.23 uH per foot.

Fred Davis, in his postings on the newsgroups, and in publication of an article on Loudspeaker cables, &quot;Effects of Cable, Loudspeaker and Amplifier Interactions.&quot; JAES, Vol.39 #6, June 1991 (Also later reprinted in mildly edited form in Audio magazine, July 1993, as &quot;Speaker Cables: Testing for Audibility&quot
states that Belden 9718 12 ga. zip has 0.23 uH of inductance per foot.

This data was provided on the newsgroups, in this post:
http://www.google.com/groups?....xy=1997

So there is a disagreement between Gene's measured data, and my measured data. &nbsp;However, it is well known that the free-field equation for wire inductance is not directly applicable to zip cords, and so, the very fact that Gene's measurements seem to correlate exactly with his theoretical calcluations makes it more likely that his measurements are the ones that are off, and not mine. &nbsp;In any case, I am in the company of other engineer's reported measurements, so I can hardly to be said to be one who is unequivocally incorrect or completely wrong.

BTW, I have been suspicious of Nordosts zip cord inductance figures for some time now, as they do not jibe with the previously established measured data that has been available.

The next portion of his article goes over my measurements for the Cross-Connected 89259, in which he ultimately decides my numbers are reasonable and in the ballpark. &nbsp;

I note for the record, that the twisting of the two coaxes is not necessary for the reduction of the inductance down to levels of 0.067 uH per foot, where the single coax is at 0.092 uH per foot. &nbsp;Since the two cables are cross-connected, the amount of mutual inductance coupling is increased over the single coax, primarily where the two braids run next to each other through the thickness of two very thin teflon jackets. &nbsp;It is this which further reduces the total inductance, and not the twisting. &nbsp;

The twisting is done to render the coax pair more flexible, and to help keep the two conductor systems from being able to move too easily relative to one another, and thus, eliminate the need for the original early design version's adhesive lined heatshrink. &nbsp;When the two coaxes are twisted together moderately, they become much more flexible than when adhesive HSed side-by-side, and yet resist any tendency to be loose relative to one another. &nbsp;I suppose it is also tidier than two coax sections wanting to splay out wildly when run around a bend. &nbsp;

The whole second page of the article compares the two cables.

But there is one thing that is not quite right. &nbsp;Normally,when folks compare speaker cables, one of the primary cirteria is to compare like gauges, so that any difference in resistance does not skew the results. Gene originally, in his first comments about my CC89259, compared it to 10 ga., stating that the CC89259 had almost twice the resistance of 10 ga. zip cord.

In a reply, I stated some values and figures for 12 ga zip cord, because that is the one that is most commonly recommended as being all that is necessary to acheive audio nirvana in terms of speaker cables.

However, if a scientific analysis was to be performed, then why not use an equivalent gauge zip cord?

To take a 13 1/2 ga. equivalent cable to task for having a higher DCR than a 12 gauge cable is a trivial excercise, and does not really compare what the geometry is doing for the Cross-Connect coax vs. the zip cord. &nbsp;Isn't this the place that makes a point of claiming to be &quot;Pursuing the Truth In Audio&quot;?

Tautologies are all very well and fine, but who is going to argue that a smaller ga. (larger diameter) wire has less resistance?

So most of the second pages comparisons are not really relavent or valid, if you really want to compare the LCR parameters on an equal footing. &nbsp;Another way of saying this, is that for an equivalent ga. zip cord, that the CC89259 will indeed have a much lower inductance, and not that much of a capacitance penalty.

If Gene/AH feels that a much lower resistance is required, then a Star-Quad CC89259 variant using four coax runs per cable will provide the equivalent of a 10 1/2 ga. cable, with even lower inductance than before. &nbsp;
Note that the details of the Star-Quad CC89259 is at:
http://www.AudioAsylum.com/audio/cables/messages/19015.html
and is listed in the Annotated Cable Bibliography at my web site.

Some specifics from the article:
[ &nbsp;While their efforts should be applauded, they should also be cautioned to consider any deleterious effects that may result, &nbsp;... &nbsp;]

This implies that the CC89259 may have deleterious effects, and this is in line with what Gene originally said, &quot;Please refrain from posting links to Jon Rischs cable receipts as we feel that his designs can do more harm than good.&quot;

However, this is not shown to be the case.

[ &nbsp;In any event, I am not sure where the .25uH/ft estimate that Jon Risch specified came from, but it seems like a bit of a stretch, actually quite a bit, 1.56 times reality! &nbsp;]

Nope, it was not an estimate, it was a repeated and 2nd source verified measurement consistent with other previously published data, and by another more recently posted engineer. &nbsp;I take exception to Gene calling his measurement &quot;reality&quot;, to further denigrate my reported data.

It would be one thing to merely comment on the disparity, but why is it necessary to do it in this manner, if it wasn't meant to be a way of casting even more doubt on my data?

[ &nbsp;As you go to thicker gauge wire (say 10AWG), the radius increases, thus inductance decreases which is exactly opposite of what Jon Risch implied. &nbsp;]

In the real world, the larger zip cords tend to use a thicker insulation, and so, the 10 ga. zip cords would have the centers of the two conductors spaced further apart due to this AND the increased diameter of the conductors themselves. &nbsp;The net result is higher overall inductance. &nbsp;12 ga. tends to sit at a &quot;sweet spot&quot; in terms of wire size, center to center spacing, and the thickness of the insulation used, so it does better than the larger diameter conductor wires.

[ &nbsp; Since Jon Rischs cables illustrate twisting, Ls should decrease somewhat depending on amounts of twists and assuming the twists are tightly packed. &nbsp;]

This comment indicates that Gene is not fully realizing the Cross-Connection paradigm. &nbsp;It is this, and not the twisting of the two coaxes, that causes the inductance to go below that of a single coax.

The close proximity of the two coax braids carrying opposite polarity signals is what does the trick.
Note that because of the two braids carrying opposite polarity signals, the jacket of the coax becomes a portion of the dielectric system, and the fact that 89259 has a teflon jacket is quite helpful here in maintaining a superior dielectric for the entire cable assembly as a whole. &nbsp;Use of another coax with a PVC jacket would then include this as a potion of the overal cable dielectric, and would tend to compromise the properties of the inner coax insulation.

RE the PSpice simulation, all I can say is that I made actual measurements, and that the model used does not necessarily include all the parameters.

[ &nbsp;Note: &nbsp;Skin Effect loses (Rac) were not accounted for either of these cables for simplicity purposes and also because they would generally represent only very minimal losses (&lt; .05dB @20kHz) within the audio band and thus DC Resistance (Rdc) in this analysis is the dominant metric of comparison for resistive losses in speaker cables. &nbsp;]

I note for the record, that despite the total equivalent ga. of the Cross-Connected 89259, that the skin effect will be much less than for a 12 ga. zip cord., as the dominant factor will be the 22 ga. center wire. &nbsp;The difference in the amount of skin effect between a 22 ga. wire, and a 12 ga. wire, is real, and could be factored in. &nbsp; Of course, adding this in would favor the CC89259, even if just a bit.

[ &nbsp;Unfortunately due to the added DC resistance inherent in this cable design, we see uniform insertion loss throughout the audio band ... ]

But only when comparing apples to oranges. &nbsp;A trivial tautology that is mentioned more than once.

[ &nbsp;Thus the benefit of reduced inductance of this cable design is greatly compromised by the additional uniform insertion loss within the entire audio band do to added DC resistance. &nbsp;]

But not if compared to an equivalent ga. zip cord, or if a Star-Quad CC89259 is used as the basis for comparison.

And again:
[ &nbsp;Yet the Coax Cable design, because of its increased DC resistance, resulted in a -0.1dB loss within the entire audio bandwidth, which would be even more apparent as cable length increases or speaker load impedance decreases. &nbsp;]

[ &nbsp;The added capacitance of the Cross Coax cable design can also represent stability problems as cable lengths increase, especially for esoteric tube amp designs with higher output impedance and lower unity gain crossing. &nbsp;]

And I addressed this in the reply to Gene, that it would take excessive lengths of CC89259 to even begin to pose a problem in the real world. &nbsp;49 pF per foot, or even around 122 pF per foot for a Star Quad version, is not that high, and not that &nbsp;limiting in terms of allowable speaker cable length.

[ &nbsp;Secondly, in the time domain, the step response may have a much higher overshoot, and exhibit excessive ringing (at about the unity gain frequency) due to loss of power amp phase margin from excessive capacitive loading. &nbsp; ]

This would only be the case IF a signal existed that high up to excite any possible resonances. &nbsp;With typical CD or DVD source material, this is almost impossible.
Again, I note that the CC89259 does not have enough capacitance to pose such a problem for any reasonable length of speaker cable, with most any power amp.

[ &nbsp; &nbsp;On a less serious note, some people may prefer the excessive frequency peaking due to overshoot that high capacitance speaker cables may cause, assuming rampant oscillations are not present, as the listener may possibly perceive it as sounding “brighter”. The question should be asked however, “Do you want your cables to act as tone controls, or be as transparent (accurate) as possible?” &nbsp; ]

Of course, none of this actualy applies to the CC89259, but the implication is that it might.

Another instance of stating a &quot;problem&quot; that is not really relevant.

RE Zobel Networks:
[ &nbsp;The speaker, if designed properly, usually has already been compensated for. ]

Not many have an overall Zobel at the speaker crossover input, if a speaker system has a Zobel, it is often only on the tweeter leg, at the tweeter terminals.

[ &nbsp;By adding the Zobel network at the speaker like Jon Risch suggested, you can actually increase shunt capacitance that the amplifier sees and possibly further increase the likelihood of amplifier instability or overshoot. &nbsp; ]

Nope, this is just not the case. &nbsp;The resistor is what makes the difference, and what helps damp the amp. &nbsp;

[ &nbsp; A Zobel network at the speaker end of a cable is (usually) next to useless. &nbsp;]

Experience over the years with the other super high C speaker cables has shown that placing the Zobel at the speaker does the most good, and is what works. &nbsp; So I have to disagree with this completely and totally. &nbsp; If one peruses the literature, you will find that in reported instances of unstable or marginally stable amps with a high C speaker cable, that the cure was often the placement of a RC damping network (or what we have been calling a Zobel) right AT the speaker terminals.

In fact, Goertz supplies free of charge an RC damping network to owner's of their ribbon cables, and guess where it is to be placed? &nbsp;Yup, right at the speaker terminals, not at the amp.

[ &nbsp;As we can see in the above analysis, there maybe no real apparent benefits to the Cross Coax cable design over ordinary 12AWG Zip Cord for high end audio speaker cable applications. &nbsp;]

This comment is based on some of the faulty premises I point out above, and so, is not really valid.

[ &nbsp; If inductance is truly a concern, then one could certainly choose a twisted pair variant of 10AWG or 12AWG Zip Cord, which will maintain low DC Resistance, critical for accurate and high performance realization. &nbsp;]

Not quite spot on either, as the comparison to 12 and 10 ga zip cords is not equivalent in resistance, and the inductance of an equivalent cable (the Star Quad CC89259) would be even lower still, down below 0.4 uH per foot.

[ &nbsp;When you consider the potential negatives of Cross Coax cable designs (IE. Increased DC resistance, excessive capacitive loading), and the hassles (IE. attempting to compensate for stability issues with Zobel networks, series inductance and/or resistance), determine for yourself if it’s really worth pursuing this effort, and if you have the time, patience and know how to proceed? &nbsp; ]

All red herrings, and not relevant to a scientific comparison between equivalent DCR cables.

The CC89259 does not need a Zobel, and it is not going to be hard to make or require undue patience or knowledge. &nbsp;Most folks assemble them, hook them up and just start loving the improved sound.

Of course, none of this simple analysis of the basic LCR parameters takes into account any of the secondary issues such as skin effect and it's related issues, or of dielectric absorption, conductor purity (lack of tin platings, etc.) and so on, and it is in these areas that the CC89259 offer improvements over 12 ga zip cord as well. &nbsp;However, given the policy of making every attemptto dismiss and ignore these types of cable aspects, I focused on mostly just the LCR parameters as did AH. &nbsp;I just wanted to point out that not everyone is as ready to dimsiss and ignore these other aspects, and they should be mentioned at least in passing.

Finally, I posted in reply to a post by Suppers_Ready about an alternative to 12 ga. zip cord, that costs about the same or less, and has superior materials and inductance over zip cords.
I a referring ot the use of Belden 8213. &nbsp;If we compare the cost of this coax and my recommended RS gold plated spade lugs to Dan B.s recent DIY article using 12 ga. zip as the basis for the cable, the 8213 approach comes out cheaper. &nbsp;The cable will have the equivalent round trip resistance of a 12 ga. cable, and have significantly less inductance. &nbsp;Plus, as an added bonus (for those who at least grant the possiblility), the cable uses foamed PE as the insulation, and all bare copper for the conductors.
Capacitance is slightly LESS than most 12 ga. zips, at 17.3 pF per foot. &nbsp;In addition, if the outer braid is connected as the ground lead (a logical thing to do), it would then provide some self-shielding.
I talk about a coup0le of other candidates fotr this type of use at:
http://www.AudioAsylum.com/audio/cables/messages/7637.html

So why would it not be a good idea to use this as a speaker cable, as it would do no worse than zip cord, have potentially superior materials, and lower inductance for those long runs to the rear speakers in a multi-channel DVD-A or SACD system?

Jon Risch</font>

 

[Audioholics]

<font color='#000000'><table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Quote </td></tr><tr><td id="QUOTE">Since I do not have a sample of Original Monster Cable handy any more, I can not re-measure it, but earlier measurements indicated it fell into the middle group, and was not unusual in its LCR parameters.

My own measurements for it, and for most typical 12 ga. zip cords came up with approx. 0.25 uH per foot of inductance.  Recent measurements by jneutron/John Escalier posted at AR of several different zip cords indicated that they measured around 0.23 uH per foot.</td></tr></table>Wow. I'm not sure where to begin... Jon, the fact that you &quot;do not have a sample of Original Monster Cable handy any more&quot; means that naturally, we should take your numbers over Gene's which were just measured and confirmed by two independent sources?

Next, Gene gave you the benefit of the doubt by using your numbers throughout the written article in addition to his own, thus additionally proving, the limited benefit of cross-coax cables even using your numbers.

Thirdly, the comparison was created to compare Monster zip to your cross-coax speaker cables, so to try and deviate the topic by insisting on a different zip guage is simply silly. The measurements and cables are provided as is, there's nothing to hide.

Fourthly, DC resistence in your cable design is not &quot;apples to oranges.&quot; The measurements are first hand and clear. Read them again, and think about the long runs that may be asociated with using cross-coax cables for surround speakers - or did you design them primarily for the main and center channel speakers?

Fifthly, when fairly presented with hard data that presents an opposing viewpoint to your designs, it would be best to respond in a factual, data-backed manner. Your response reads like someone whose pride was hurt, not at all like an engineer who has critical feedback to the article. remeber, Gene is critiquing cross-coax, not Jon Risch.

Finally, for the benefit of everyone, please consider reposting a response that includes either first-hand contrary measurements, or a fact-based response to Gene's article [unmeasurable skin effect as a defense?]. I believe that utilimately Gene provided a fair review of your cables, without slamming your design or insulting your methods.  It was, in fact, a peer review - which, as an engineer, you should encourage. The fact that your cables are not proven to be overtly better than standard zip cord (as very clearly defined by Gene in the article) should not be a shock to you if you indeed did the same measurements yourself.

Have a nice day.

[edit: spelling]</font>

 

[Bprest0n]

<font color='#000000'>It's funny how Jon fails to realize that Gene actually endorsed his cable design as an alternative to &quot;exotic&quot; cables and I quote:

<table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Quote </td></tr><tr><td id="QUOTE">We fully encourage audiophiles or hobbyists that wish to experiment with &quot;exotic&quot; cables to first try a design such as this Cross Connect Coax or a Cat 5 variant. Although many of them are higher in capacitance than standard Zip Cord, and sometimes higher in DC resistance, they are an excellent way to &quot;try before you buy”, and Jon Risch’s receipts are reasonably good and cost effective to produce, even if his claims of their performance are a bit overstated. </td></tr></table>

Also; one major problem I see is that Gene took Jon's word on his cable metrics without measuring them to verify, especially since all of Jons attenuation #'s were very exaggerated.

</font>

 

[Guest]

<font color='#000000'>I measured a home depot #16, #18, and #24 awg zip. &nbsp;All fell in at .233 microhenries per foot.

As stated in an AR post, I have not measured any #12 zip, my lowest was a twisted pair of #10..

Two LCR machines were used, both correlated, both traceable to national standards.

The formula became inconsistent for conductor spacings below ten diameters..

Given the consistency of error between the equation and measurements for all the tests I performed, I find Jon's #12 zip measurement to be in order, though I did not test a piece.

Terman formula (courtesy of &quot;Leisure7&quot:
L=.01016*length*{2.303*Log(2*D/d)-(D/length)+(mu*delta)}

At low freq, delta (skin factor) is about .25, goes down with freq (courtesy of &quot;Leisure7&quot:, which is consistent with my tests at 100hz, 1Khz, 10Khz, and 100Khz, the measured values asymptotically approaching about +25% error over equation..

&quot;we should take your numbers over Gene's which were just measured and confirmed by two independent sources?&quot;Hawke

My measurements seem closer to Jon's.. &nbsp;I am of course, open to the possibility of testing error on my part. &nbsp;However, I did use two different HP LCR meters..

&quot;&quot;Recent measurements by jneutron/John Escalier&quot;&quot;JR

Two l's.

&quot;&quot;twisting of the two coaxes is not necessary for the reduction of the inductance&quot;&quot;JR

In both my #24 and #10 twisted measurements, the twist pitch played no DIRECT role in inductance. &nbsp;When the twist is tight, it helps keep the wires together to lower inductance; pulling the wire pair with little or no twist had the same effect.

Cheers, John</font>

 

[Bprest0n]

<font color='#000000'>Hi John;

I don't understand how you could imply that skin effect changes inductance of a cable?  The phase angle of the impedance change resulting in skin effect losses is zero, thus skin effect can only be responsible for change is resistance, NOT inductance.

Take a look at this interesting link from Belden.  They calculate Ls = .06uH/ft for 10AWG cable, certainly much lower than you, Jon and even Gene claim.

Belden Cable Parameters



As for your equation, not sure what those last terms mean, but I doubt they relate to skin effect.

Also, take a look at this article. &nbsp;It shows that skin effect does not even come into play until the skin depth is less than the conductor diameter. In the case of 12AWG wire, thats at about 5kHz as shown here.


Skin Effect Relevance in Speaker Cables</font>

 

[Audioholics]

<table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>

Guest : <font color='#000000'>My measurements seem closer to Jon's..  I am of course, open to the possibility of testing error on my part.  However, I did use two different HP LCR meters..</font>

<font color='#000000'>It could also be that your measurements are correct, but that you didn't measure the same type of cable (Monster brand zip cord) OR that your measurement technique is not the same. In either case, the point is nearly moot as both measurements were given and the conclusions are the same.</font>

 

[Guest]

<font color='#000000'>Hi Bprest0n
Not THE B Preston of &quot;outta space&quot; fame??

I'll start with this:
&quot;&quot;As for your equation, not sure what those last terms mean, but I doubt they relate to skin effect.&quot;&quot;

My apologies...I kept it a tad too brief...they are skin effect terms.

From Leisure7's post:

Radio Engineer's Handbook (F. E. Terman, McGraw-Hill, 1943) is the source of the equation..

&quot;mu*delta&quot; is the skin effect correction to be applied to the equation...Mu is the magnetic permeability of copper...delta is a nomograph-derived skin effect factor. (my note:I've no info as to how the nomograph was derived)

He goes further to state that the skin effect at high frequencies decreases inductance about 2 percent for long wires.....End of Leisure7's post info....

In examining the link you provided stating .06 uhenries per foot?...It's a typo, I believe he meant 0.16 calculated...As my testing, Jon's, and Gene's go against that number..And .16 was the ballpark for my calculations as well..Too bad the zip cord and LCR meters I used don't understand the math.


&quot;&quot;I don't understand how you could imply that skin effect changes inductance of a cable?&quot;&quot;

I'm not implying it...I (leisure7) just provided an equation which does so, as printed in 1943.  Terman says so...And if you go through the numbers, the effect is very low..

&quot;&quot;The phase angle of the impedance change resulting in skin effect losses is zero, thus skin effect can only be responsible for change is resistance, NOT inductance.&quot;&quot;

Yes, the transport current phase angle will shift only a little as a result of skin, the longline inductance only shifting up to 2%...Not zero mind you, but very little..I'm not concerned nor speaking of the orthogonal lagging eddy currents.

And until proven otherwise (I'm tryin), I can only concur that the only effect would be current density profile resistive effects..

My testing results..#18 guage wire...twisted pair..

freq         distance  calculated  actual  error     meter
Hz             inches        L             L      %

1000          2.5        0.587      0.626    6.64    HP4162A
1000          1.75      0.545      0.556    2.01     HP4162A
1000          0.5        0.392      0.434   10.7     HP4162A
1000          0.08      0.169      0.208    23.0    HP4162A

120            0.08      0.169      0.278    64.49   HP4263B
1000          0.08      0.169      0.233     37.86   HP4263B
10000         0.08      0.169      0.231    36.68   HP4263B
100000       0.08      0.169      0.22      30.17   HP4263B

As is seen, the inductance drops as the frequency increases with the HP4263B...Graph would imply settling in at about 25% error..

And the first set....shows the equation error as the conductor spacing decreases..


Cheers, John</font>

 

[Bprest0n]

<font color='#000000'>Perhaps you should qualify how you are measuring the cable (IE. are you measuring one leg, or shorting one side and measuring at the other end?)</font>

 

[Guest]

<font color='#000000'>&quot;&quot;Perhaps you should qualify how you are measuring the cable (IE. are you measuring one leg, or shorting one side and measuring at the other end?)&quot;&quot;

My fault, sorry..

The cord under test is stripped at both ends...The far end, away from the meter, the two wires are shorted..

At the meter, they are just connected to the force terminals, there are jumper strips to the sense terminals...So the measurement is four wire up to &nbsp;the binding posts..

The second meter, the 4263, actually has a set of large alligator clips, with two wires to each clip..giving full Kelvin connection at the wire.

For the non zip spacings, I controlled the wire geometry, and kept the resulting loop away from conductive objects.

For the zip, I was able to wrap the wire around any metal I desired, the readings did not waver..as was expected..

To Hawke: &nbsp;agreed, I did not measure the exact same cable...

You confused me with &quot;we should take your numbers over Gene's which were just measured and confirmed by two independent sources?&quot; , as my measurements are consistent with Jon's, while Gene's was lower..

I am unable to explain why my measurements and Jon's do not agree with the equation, while Gene's is dead nuts on..I can only note the error diminishes with increasing spacing, and does not diminish at frequencies up to 100Khz.

Cheers, John</font>

 

[gene]

<font color='#000000'>Hey John;

Nice to see you participating in our forums!

Unfortunately the Wayne Kerr Analyzer has been moved to an area of my building I don't have clearance to enter and will be there for at least 2-3 weeks

However, I did manage to get my hands on another LCR meter. &nbsp;This one is the HP 4275A. &nbsp;Unfortunately it is a high bandwidth meter, but the low end goes down to 10KHz.

When I say your posted measurements with inductance decreasing as frequency goes up in the audio band, it raised a red flag in my head.

Thus I went to the 4275A with my 1ft 12AWG Monster Zip cord and began taking measurements (without calibrating the device).

What I found was inductance measurments of .268uF @ 10kHz and decreasing with frequency up to 1MHz. &nbsp;This can't be right. &nbsp;Thus I shorted the leads together without any wire and measured .1uH of inductance. &nbsp;Alas, I found it. &nbsp;We need to calibrate.


Calibration procedure:
I first ran open circuit calibration. &nbsp;Next, using a paper clip (YES I am scientific &nbsp;

), I performed a short circuit calibration.

New measurements:


Frequency &nbsp; Ls (1st time) &nbsp; &nbsp; Ls (2nd time)
10kHz &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.151uH &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.149uH
20kHz &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.164uH &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.167uH
40kHz &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.172uH &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.170uH
100kHz &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; .178uH &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.176uH
1MHz &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.170uH &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.169uH
5MHz &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.165uH &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.162uH

This pretty much correlates with my prior measurements from the Wayne Kerr. &nbsp;You should not see inductance decrease until you go beyond a decade past the audio band.

As for low frequencies, inductance at say 120Hz should be even less than my 10kHz measurments. &nbsp;In fact you sometimes will find you have to switch between Ls and Lp, and/or change your measurment speed to make an accurate measurement because inductance is so low.

I am very familiar with the LCR meters you are using. &nbsp;I used them for several years at my former place of employment. &nbsp;They are great machines and very easy to use. &nbsp;The model you are using where you have to jam the wire into a slot can be quite tricky. &nbsp;I suspect it doesn't like the thickness of 12AWG wire and that can be hindering your measurments. &nbsp;As for the one with the probes, I suspect you need to calibrate out the probe inductance.

Just to further my case, I called MonsterCable today and asked them for thier RLC #'s, sure enough they confirmed .16uH/ft is reasonable at 20kHz.

Once I get ready for my cable face off comparison and I have the Wayne Kerr back in my possesion, I will redo all of these measurements. &nbsp;Perhaps we can chat at that point.

One more thing, I also note that moving around your probe tips, or not making tight contact with them to your test wire, while also keeping the test wire stripped ends in close proximity, can also throw off your results.

To confirm my results, set-up, etc I had an independent engineer at my company make these measurements for me and he came up with very similar results.

Also, please note that I used Jons inductance figures for 12AWG Zip cord in my analysis as well and showed that his attenuation #'s were not correct.

In the end, we are splitting hairs here and a 1/2 to 1dB difference at 20kHz is trivial, especially when you consider hardly anyone hears up that high, only harmonics of music are perhaps at that frequency, and the human ear is very insensitive past 10kHz based on the flecher munson curve.

Fletcher&amp;Munson Curve

As I concluded in the article, Jon's cables are quite good, very comparible to 12AWG Zip Cord, and certainly better than many costlier &quot;exotic&quot; cables out there. &nbsp;When time permits, I may construct his cables to use in my cable faceoff comparison.

I am not sure about the rest of that equation you quoted from &quot;Leisure 7&quot;, not even sure who that person is. &nbsp;As far as I know the inductance equation I used is valid to approximately calculate inductance. &nbsp;It is not exact, which is why I used measurments and vendor data, as well as Jons extreme inductance measurement.


Jon;

You are correct, I should have stated that your inductance figures are more of a result of mutual inductance and not from twisting. &nbsp;If time permits tonight, I will update the article accordingly.

As for skin effect changing the results of the comparison, I say read my article again where I addressed that topic. &nbsp;The science is sound, it has been peer reviewed, even by Dr. Howard Johnson. &nbsp;Skin effect is NOT an issue for speaker cables and only represents slight resistive losses at the frequency extremes.</font>

 

[Guest]

<font color='#000000'><table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>

gene : Thus I went to the 4275A with my 1ft 12AWG Monster Zip cord and began taking measurements (without calibrating the device).

Also, please note that I used Jons inductance figures for 12AWG Zip cord in my analysis as well and showed that his attenuation #'s were not correct.

Jon;

You are correct, I should have stated that your inductance figures are more of a result of mutual inductance and not from twisting. &nbsp;If time permits tonight, I will update the article accordingly.

[ &nbsp; Thus I went to the 4275A with my 1ft 12AWG Monster Zip cord and began taking measurements (without calibrating the device). &nbsp;]

You are using only one foot of cable to make the measurements?!?!?!

I have been using 10 foot sections, and dividing by ten, which I feel reduces any termination errors and still gives valid numbers.

[ &nbsp;Also, please note that I used Jons inductance figures for 12AWG Zip cord in my analysis as well and showed that his attenuation #'s were not correct. &nbsp;]

Well, this is a matter of outlook. &nbsp;You used a MODEL, and CALCULATED some numbers, I made MEASUREMENTS, and saw that they correlated with exisitng literature/data that is out there. &nbsp;Perhaps you have shown this to your own satisfaction, but not to a true scientist, and certainly not to me.

I have remeasured a lot of cables recently, due to jneutron/John Escallier pestering me about the accuracy of my measurements. &nbsp;All recent results are within measurement error limits of the old measurements, and were what I expected them to be.

BTW, I find it interesting that Belden has a spec for their 9718 zip cord, the very same cable that Fred Davis measured as having 0.23 uH of inductance, at:
http://ecom.belden.com/static/ZZBLDNTD01FROMCATA.HTM?P0=9718
(note: this site has to be accesed during bussiness hours in the daytime)

To spare the suspense, they rate it as having 0.20 uH of inductance. &nbsp;This is lower than what I and Mr. Davis (and Jneutron) measured, but it is also higher than what you measured. &nbsp; Note that this is one of the category of zip cords with very close spacing, due to a thin insulation layer, and so, represents the lowest possible amount of inductance for a 12 ga. zip cord style cable.

In the past, I had measured some 9718 as having 0.23 uH, so my measurements seem to be running a tad high, and yours seem to be running a bit low. &nbsp; If we use the Monster Cable numbers you derived (which is giving the beneift of the doubt, as it is highly likely that the Monster Cable has higher inductance than the 9718 due to greater center to center spacings), that would have your measurements about 25% low, while mine are running about 15% high. &nbsp;

Perhaps neither one of us has it nailed down to that last tenth of a uH? &nbsp;That might sum up the whole cable argument in a nutshell, but who will give ground first?

I still feel (as do MANY others) that the Cross-Connected 89259 speaker cable has benefits that 12 ga. zip cord does not, and the simple expedient of LISTENING to it provides all that most folks need to be convinced.

Jon Risch</font>

 

[gene]

<font color='#000000'>Jon;

I updated the article regarding the mutual inductance leading to lower overall inductance of your cable design as you commented in your first reply. &nbsp;I also showed this update on the last page. &nbsp;Thanks for your feedback.</font>

 

[Guest]

<table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>

hawke : <font color='#000000'>Thirdly, the comparison was created to compare Monster zip to your cross-coax speaker cables, so to try and deviate the topic by insisting on a different zip guage is simply silly. The measurements and cables are provided as is, there's nothing to hide.

Fourthly, DC resistence in your cable design is not &quot;apples to oranges.&quot; The measurements are first hand and clear. Read them again, and think about the long runs that may be asociated with using cross-coax cables for surround speakers - or did you design them primarily for the main and center channel speakers?</font>

<font color='#000000'>Hawke,

Invariably, when folks who are convinced that there are no differences between audio cables talk about comparisons between zip cord and 'exotic' cables, &nbsp;they discuss cables that have the same overall resistance, as this is one of the basic parameters for a speaker cable. &nbsp; &nbsp;They do this, so as to eliminate the trivially simple factor of a different DCR from becoming an issue, which would just confuse the whole comparison.

However, rather than doing this, Gene has chosen to compare a 13 1/2 ga. cable to a 12 ga. cable. &nbsp;One is OBVIOUSLY going to have a lower DCR than the other. &nbsp;This is a given.

Yet he has taken the CC89259 to task for having a higher resistance than 12 ga. zip cord!!!! &nbsp;Gee, I can't imagine how that happened!

The fact is, this kind of a difference is a trivial and insignificant kind of difference, and to then take the comparison further and compare the inductance and capacitance of the 12 ga. to a different effective gauge cable, does not make any sense, especially if he is going to use phrases like this:

&quot;Thus the benefit of reduced inductance of this cable design is greatly compromised by the additional uniform insertion loss within the entire audio band do to added DC resistance.&quot;

He is saying that the reduced inductance is not as noteworthy, because the 13 1/2 ga, cable has an higher DCR than a 12 ga cable?! &nbsp;Isn't this a bit off the beaten path of true science?

If we really wanted to compare the CC89259 and it's reduced inductance, it should be compared to an equivalent zip cord, say a 14 ga. zip (as the closest normal size), and see what it's inductance and resistance do for losses into a 4 ohm load. &nbsp;

Otherwise, the simple expedient of doubling up on the 89259 coaxes, creating a Star-Quad style Cross-Connected 89259 speaker cable, with the equivalent ga. of a 10 1/2 ga. cable, reasonable levels of capacitance, and an inductance lower than even the braided and ribbon sandwhich cables, will now outperform 12 ga zip, be close enough on DCR for 10 ga. zip, and have far lower inductance then either.

I note for Dan's benefit, that at this point, the capacitance of the SQ CC89259 speaker cable will still be less than 1/5th that of the other types of low L cables, allowing it to be used in the real world lengths. &nbsp;This despite the L being even lower than those other low L cables.

As for use of either with HT rear speakers, I believe I have made this perfectly clear. &nbsp;I do not believe that Home Theater (as opposed to multi-channel DVD-A or SACD) requires a high performance speaker cable, I feel that 12 ga. zip is plenty good enough for the rear channels of an HT system. &nbsp;The rear channels of all but a SOTA HT system just do not warrant a high performance cable. &nbsp;I define a SOTA or near SOTA HT system as one that has a separate component decoder, at a minimum, and a set of speakers to match that level of performance and sophistication.

Given the dismissal of certain audio cable performance parameters at AH, I find it absolutely amazing that in the same breath, one can say that 12 ga. is as good as CC89259, and yet, that a rear channel speaker needs something more than 12 ga. zip cord.

Where is the logic here, and what in the world would AH recommend for a rear HT system speaker cable?

I have pointed out the existence of another Belden part number, 8213, which has a resistance very close to 12 ga. for the round trip, slightly lower capacitance than 12 ga. zip, and uses superior materials, such as foamed PE for the insulation, as opposed to PVC. &nbsp;It has all bare copper, as opposed to possibly being tinned. &nbsp;And it has lower inductance than 12 ga. zip, at 0.097 uH per foot, vs. (from 0.16 uH to 0.25 uH, you choose).

On top of that, if the outer braid is used as the ground or negative connection, a logical choice, then the cable will also be self-shielding, something that zip cord does not have.

I have noted in another post that the cost of 8213, and a set of adequate gold-plated spades from RS AND a crimp tool to crimp them, costs less than Dan B.s DIY 12 ga. zip cord project, as his parts list and costs spell out.

So if one is concerned about the length of a rear channel run, then it looks like 8213 will outperform 12 ga. zip, using only the numbers that AH believes is relevant. &nbsp;Other folks will appreciate the foamed PE and the all bare copper conductors.

I note that this has been my recommendation to folks seeking a performance higher than zip cord for rear channel HT situations, a simple search of the AA Cables forum will verify this, and the date on the post where I first discuss the use of 8213 (May 20th, 2000) should indicate that this is not some sort of last minute response to the AH controversy.

I do not believe that any one here can come up with a downside to this, it is either nearly the same or &quot;out numbers&quot; 12 ga. zip, so what will be the reaction to this one?

Jon Risch</font>

 

[Audioholics]

<font color='#000000'><table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Quote </td></tr><tr><td id="QUOTE">I do not believe that any one here can come up with a downside to this, it is either nearly the same or &quot;out numbers&quot; 12 ga. zip, so what will be the reaction to this one?</td></tr></table>

That's just it, Jon... You nailed it. the differences are so minimal it is uncertain how the trouble of constructing the corss-coax design benefits people. While it is a good design, as noted by Gene, the benefits are not clearly superior to (choose your own gauge here) zip cord.

How long would you say it takes to make that cable?

Your cable design was praised, Jon, but it doesn't appear to be superior to Zip cord. Even the statements you make here don't serve to contradict this. I think that's the only response you will get regarding this.</font>

 

[Guest]

<font color='#000000'>Gene:  the equation from Terman has a correction factor for the length of the cable...D/length, and a correction for skin effect, mu * delta...I unfortunately do not have access to the nomograph yet..

Leisure7 is a poster at &quot;cable&quot;.

If you have a shorted reading of .1 uH, then reading a cable at .15 uH is not going to be accurate..  Are you locking it out of auto-ranging?  If you show .1 offset, then re-cal, and show readings that are .02 different, then the cal didn't really change anything.. I suspect there's a ranging/resolution issue with your meter.  (hmm, I better check mine for the same thing)..

I agree with Jon..You should be running at least 5 to 10 feet of wire to reduce the residual error.

Yes, the holes in the 4261A are a pain in the arse..But I ran the 10 guage with no error, used a 5 foot twisted pair, and the measurement was not susceptible to movement or configuration..

If you examine the Terman formula w/r to my #18 guage...

D=.08 (tightly twisted), r=.020, mu=1

delta=.25 (lf, no nomograph for exact number, but using the number provided by Leisure7).

For one foot:

L=.01016*12*{2.303*log(8/2)-(.08/12)+(1*.25)}

L=.12192*{(2.303*.602)-(.006)+(.25)}

L=.12192*{(1.386)-(.006)+(.25)}
note the length correction is -.4%, trivially small
note however, the skin correction is 18%, not small at all, and quite close to the errors I've been seeing..

L=.12192*{1.63} = .1987 uHenries per foot..

Compare to my test results:

freq        reading        error of reading    meter

1000       .208              +4.6%             HP 4162A

120       .278               +40%              HP 4263B
1000      .233               +17%              HP 4263B
10000     .231               +16%             HP 4263B
100000   .22                 +10%             HP 4263B
note:the 100k row is comparing measurement against a higher than should be skin correction, and the error % number should be considered suspect..possibly also the 10k..

Clearly, the 4162 produced the expected reading, one in good standing with the equation...

Now, as to the 4263B?  In setting it up again, I can change the reading by dressing the leads...For whatever reason, HP decided to separate the force and sense leads about two inches from the large clip...This is the source of the movement susceptibility..I'm looking up the spec on the clip assembly, I suspect it is not rated for inductances we are measuring..While the probes are not presenting an inductance to the instrument, the instrument is not designed to accomodate these clips at this low a measured value..

I recommend you review your setup for the same thing I'm finding..

I also recommend using longer lengths of cables, to get the reading out of the mud..If you measure 1 foot, then ten or 20, I think you'll see the readings drop...as you're dividing the residual out..termination errors, as Jon called them..Let me know..  Thanks.

Cheers, John</font>

 

[gene]

<font color='#000000'>John;

I realize with you I must be very clear and specific as you pay close attention to details

When I shorted the leads of the LCR meter, they didn't stay fixed at .1uH, but jumped around. &nbsp;After calibrating, it nearly zerod out. &nbsp;Even if cable length were an issue, one should still be able to accurately measure Ls at much higher frequency with only a 1ft length.

Howeverd, Today I brought in an 8ft Monster Zip Cord. &nbsp;If I &nbsp;have time during lunch, I will retake measurements 10KHz and above (since I dont have access to the Wayne Kerr LCR meter) and verify the measurments are consistant.

Without seeing the book you reference, I cannot verify that equation. &nbsp;I never heard any reference in any book that demonstrates skin effect changes inductance of a cable. &nbsp;Proximity effect will, but to my knowledge, not skin.

In actuality, proximity effect will reduce inductance, which is probably what you are seeing at high frequencies.

BTW, I used an 8ft legth of cable to measure Rdc awhile back and then divided by 8 to get Rdc/ft using a mohm meter. &nbsp;It is very difficult, if not impossible to measure Rdc correctly with only a 1ft length until you get to high frequencies where you can measure Rs accurately.</font>

 

[Guest]

<font color='#000000'>Well....found one big error source...

The HP 4263B is designed for 4 wire measurements...Two current terminals, and 2 sense terminals, just like standard high precision measurement setups..Unlike standard setups, however, is how they handle the shield question..

For standard measurement techniques, the shield can carry individually to each of the four terminals..And it is an electrostatic shield...This meter, however is just a tad different..

The connectors on the front panel are BNC..to eliminate the di/dt (current magnetic field) of the current wires from being picked up by the sense wires, the return current of the test is in the shield runs..In other words, the ac current used for the measurement comes out the positive current center pin, through the load, in the negative current center wire, through an internal current meter, then BACK OUT the shield of the negative current wire, then crosses to the shield of the positive current wire, to return to the meter positive pin shield point..The reason for this is both current sheilded wires carry coaxially the out and in current, thereby nulling out any magnetic field generation by the current..

The downside to this is you must connect the current shields as close to the dut as possible...the clip fixturing I used connects the shields a &nbsp;FOOT away from the clips...DUHHH!! &nbsp;Then has heat shrink holding the bare (non shielded) current/sense wires together for 10 inches....

Gee, maybe that explains a lot...It certainly helps explain why my high bandwidth 4 ohm load resistor measured 1.5 HENRIES, instead of the microhenries it actually is..

I'll have to make my own design test lead setup before I can trust the 4263B.

Cheers, John</font>

 

[gene]

<font color='#000000'>John,

I don't have your email address here at work, please send it to me at: gds@audioholics.com.

I have been speaking with Henry Ott and have some very interesting info to share with you. &nbsp;If he gives me permission, I will post it as a follow-up in this forum as well as my article.</font>

 

[Guest]

<font color='#000000'>A wee bit of the technical....

Inductance is the measure of the energy stored in the magnetic field as a result of current..Higher the inductance, the higher the energy storage...

Within a wire carrying DC, there is a uniform current density profile.. The magnetic flux within the wire is zero at the geometric center, and increases linearly in value as you move towards the surface of the wire...Outside the wire, the field drops off as 1/R....

When a conductor is skinning heavily, as in RF, all the current has moved to the outside surface of the conductor..

From the field equations, the field within an infinitely thin cylindrical sheet of current is zero....So, at infinite frequency, the internal portion of the wire has no field, hence, no energy stored, and no inductance....That is how the skin effect alters the self inductance of the wire. The more skinning that occurs, the less energy and field within the wire.. that's the &nbsp;.0152 microhenries per foot per wire, as described by the &quot;outside consultant&quot;.

That is also why the skin factor of the Terman equation is needed at low and audio frequencies, but can be ignored at RF...



Cheers, John</font>

 

[gene]

<font color='#000000'>It appears we have all been hit with the “error bug”.

Jon – overstated attenuation and inductance figures for 12AWG Zip and understated attenuation figures for Cross Coax.

John – set-up errors of LCR meters yielding too high inductance measurements.

Me – Using too short of cable lengths for measurements resulting in too low inductance figures, and not accounting for internal inductance of cables in inductance equation for frequencies below skin effect relevance.

With help from field expert Henry W. Ott and all of the back and forth debating between all of us, we have successfully increased accuracy in our analysis by a significant factor. I would like to thank both John and Jon for providing such excellent feedback.  Going forward, we can assure absolute accuracy and fair comparative methods for measuring all audio cables.  

Ok, lets go back to 12AWG Zip Cord and redo our calculations , measurements and simulation:

Calculations and Theory:

Ls = 0.281*Log(B/A) uH/ft     Eq(1)
where B is the space between two conductors and A is the Radius of each conductor.

This equation is valid for the total loop inductance (both conductors) of a parallel wire transmission line. It represents the EXTERNAL inductance of the conductors. However at audio frequencies, where skin effect is not a major factor, the total inductance becomes (external + internal) where external is listed in the above equation and internal inductance (self inductance) of a straight wire of circular cross section carrying a uniform low frequency current is 1.27*10^-3 uH/in, independent of wire size.  (Source, Noise Reduction Techniques in Electrical Systems, 2nd Edition, Henry W. Ott)

Thus to calculate total Inductance (Ls) the complete equation is:

Ls = 0.281*Log(B/A) + 2*(12)*1.27*10^-3 uH/ft     Eq(2)

For 12AWG Zip Cord, we select B = .150 and A = .0404

Thus Ls = .191 uH/ft



Measurements:

As previously mentioned, I currently don’t have access to the low frequency Wayne Kerr 3260B Precision Magnetics Analyzer and will have to wait a few weeks until I do.  Thus I will limit my measurements to the HP high frequency LCR meter 4275A starting at 10KHz and working up to show how skin effect reduces self inductance and thus increases Rs which in turn negates the added internal inductance figure from the above modified equation Eq(2). In theory, as frequency becomes larger, we should see inductance return to the calculated .16uH/ft as originally calculated.  As frequency becomes infinite, as John states, “the internal portion of the wire has no field, hence, no energy stored, and no inductance.”

For the measurement data below, an 8ft 12AWG Original Monster Cable product was used.  All values were divided by 8 to normalize them for a per foot specification.

Measured on: HP LCR 4275A
Frequency         Ls                         Rs
10 kHz         0.192 uH/ft                Beyond measurability
20 kHz         0.190uH/ft                         &quot;   &quot;
40 kHz         0.190 uH/ft                         &quot;   &quot;
100 kHz         0.187uH/ft                         &quot;    &quot;
400 kHz         0.185 uH/ft                  20.5 mohms/ft
1 MHz         0.180 uH/ft                  54.8 mohms/ft
4 MHz         0.162uH /ft                  145 mohms/ft
10 MHz         0.120 uH/ft                         Infinity

As you can see, as skin effect increases Rs, inductance drops.  As we approach 4 MHz, Rs becomes significantly large resulting in extreme reduction of self inductance and thus Ls approaches the value of the external inductance Eq(1).  As we go higher in frequency, we see Ls reducing to zero and Rs infinitely increasing as John also said in the previous post.  

*Note Rs at DC = Rdc = 3.4 mohms/ft thus as 20kHz, Rs should = 1.34*3.4 = 4.56 mohms/ft  
See: Skin Effect Relevance in Speaker Cables

Any readings on the LCR meter significantly than this for frequencies 20kHz or greater are rejected since the LCR meter cannot accurately measure Rs for those frequencies (value too low for measurability).  As we approach 10MHz, Rs becomes so large that it cannot be accurately measured on this LCR meter.  


PSPICE Analysis:


As we can see, Ls for 12AWG Zip Cord within the 20kHz audio band should be about .19uH/ft, not .16uH/ft as I originally states and confirmed with Monster Cable.  However, this is still significantly lower than the .25uH/ft that Jon Risch has stated previously.    


Corrected Analysis (Ls = .19uH/ft)
Cable Length: 10ft  Total Loss @ 20kHz = -.09dB &lt;&lt;&lt; than .25dB stated by Jon Risch and only .01dB worse than I originally stated.

Cable Length:  50ft:  Total Loss @ 20kHz = -0.70dB &lt;&lt;&lt; than 1.3dB stated by Jon Risch and only 0.1dB worse than I originally stated.


While the total loss changed by  -.01dB for 10ft and by -0.1db for 50ft at 20kHz seems insignificant, this exercise was completed to improve accuracy.  We encourage Jon Risch to remeasure his results since they are significantly off from both proven calculations and measurement tabulations.  


I will be updating the article tonight, time permitting, to reflect these changes.

[attempted to reformat tables]</font>