Go to home Aspen Amplifiers Go to home
products FAQs discussion buying contact













The Heavenly Connection – DIY Speaker Cables

(Our long time associate Lewis Muratori crafted and contributed the following article to the Melbourne Audio Club monthly magazine.  Lewis is an erudite, intelligent and experienced lifelong audiophile, who has experimented at considerable length and in considerable detail in this and many other DIY audio areas.  We thank him warmly for allowing us to reprint excerpts from his original work, as follows).

As we all know, there's an outrageous amount of hype, New Age Physics and, frequently, nonsense directed towards the paranoid, obsessive-compulsive, dollar rich set known collectively as Audiophiles.
I enjoy a laugh as much as the next guy, and I am sorely tempted to aid and abet this confusion by offering even more dross and magic.  However, following are the bland facts, wholesome and unembellished.  I sincerely hope my article encourages you to experiment, as cables do make a difference.

Andrew, a good friend and audiophile, recently had a visit from a young audio acquaintance who had constructed a set of unusual speaker cables.  The recipe for these cables is still freely available on the Net.  These cables sounded surprisingly good.  In fact, they were so good that Andrew immediately collected the vital ingredients and made up one of the three sets needed to fully equip his tri-amped system.
Quite soon he invited me to audition these cables.
Andrew never exaggerates - quite the opposite, in fact.   Like me, he has more than a couple of decades invested in this "hobby".  So, when Andrew calls and suggests something I listen.
These cables were very interesting.  They were unusual, yet compellingly cheap.   They looked like two lengths of 75 ohm co-ax connected between amp and speaker.
Imagine we have two pieces of 75 ohm co-ax lying parallel in front of us.  On the LEFT cable end, the PVC sheath has been trimmed back 2cm, and the braid pulled back and twisted together.  The central dielectric, (in this case the pentagonal "air" type rather than the foam) has been trimmed back about 1cm. There is 2cm of copper braid sticking out to the side and 1 cm of solid core copper sticking clear out of the end.  The RIGHT cable is trimmed in a similar fashion.

Now, here's trick number 1.  Take the braid from the LEFT cable and connect it to the central conductor of the RIGHT cable.  Take the RIGHT cable’s braid and connect it to the central conductor of the LEFT cable.  Repeat this at the other end of your cable run.
You can see that each wire of each speaker cable will be both an inner and outer conductor, but not from the same co-ax cable.

That takes care of the mechanics.  In the Net design however, the braid was instead a thin copper shield.  This shield was left covering the inner dielectric after an outer braid was removed from a particular cable designed, I think, for microwave use.  Andrew and I spent a couple of days refining this original design by adding a few ideas of our own.
Now, the methodology.  Every configuration and/or material tried was exhaustively auditioned and compared with a couple of references to ensure we were not imagining things.  Those who know me will be aware of my healthy skepticism concerning all things audiophile. I have just two audiophile requirements of any new idea:
* It can be demonstrated repeatedly both to myself and to a non-audiophile to be an improvement AND·  It delivers demonstrated value for money.

Ideas which pass these tests end up staying in the system until further notice.  If it is DIY, then even better!   Andrew's outlook parallels my own.  And it should go without saying that only one thing was changed at a time.  We listened in mono, as this halved our workload, yet still gave us all the information we needed.
The system configuration was:
1. Amplifier- solid state
2. Speaker- large, full-range 3 way dynamic type
3. Stuff that didn't work - I'll spare you
4. Opinions- our own
5. Credibility - you be the judge.The first area of improvement we discerned lay in the quality of the conductive elements.  Some judicious hunting unearthed an inexpensive coax with both a clean, and more importantly, annealed centre conductor.  This latter, in my experience, tends to sound less grainy than the "work hardened" alternative.  The copper used in the original design was of a fairly low quality, bad enough to leave even a neophyte audiophile unimpressed.  Both Andrew and I had experimented in the past with flat copper speaker cables, based on unrolled "alpha core" inductors that we cut into lengths and then shielded.  Could we use some of this good quality material instead?  Well, yes, but the pile of 14 gauge material in our stash was 30% too wide to wrap in a single layer around the dielectric.  Stanley knife and straight edge to the rescue, and we were ready to go.
Yes, it sounded far, far better, and yes, it was worth the effort.

The next problem was mechanical stability.  After wrapping the copper tape along the length of the cable it was obvious that we would need some extra support to avoid cable kinks when bent.  Of course, we also needed an insulated covering for obvious reasons.  The original design used heatshrink over the whole thing, which worked well.  But concerns about the effect on the sound due to the heatshrink, particularly over such a long length of conductor, led to a better idea.  We used 3mm nominal expandable cable mesh to cover the copper before the heatshrink was slid on and shrunk (use the clear stuff). Approximately 6.5 mm internal-diameter worked fine.  This gave two worthwhile benefits:
*   Kept the heatshrink off the surface of the copper ribbon; and·     Introduced a virtual air dielectric over the ribbon due to the open nature of the braid.

And, yes, we did try it both ways, and yes, the braid did sound better. Smoother, clearer, less background noise.  The difference was surprisingly obvious.  Incidentally, the Alpha Cores should be marked for directionality.  Sounds best with the outer end at the destination i.e., speaker.
Design refinement completed, and both of us satisfied we had tried all reasonable configurations and had finished with something pretty special, we called it quits, compared grins, and had another chuckle at the expense (pun intended) of high-end cable manufacturers.

The sound?  This is a supremely musical cable.  Its overall character in the context of our systems is silky sweet, totally grainless and harmonically rich, especially in the lower mids.  Treble information is extremely clean, aided by the cable's surprising dynamic capabilities both at micro and macro levels and from top to bottom.  Ah yes, the bottom!  Taut detailed, articulate, lacking any boom or bloat and with superb pitch definition.
In fact, this cable contrives to amaze me even after enjoying it in my system for some months.  Many of its attributes remind me of the effect that a good mains filter can have in a system.

Theories?  Nothing concrete this end.  Constructive input from the more technically inclined is warmly invited 'cause I for one would love to know why it does what it does so well.

Lewis Muratori

Just to hand (June 2005) has come an addendum to Lewis' cable, from Andy Redwood, in Melbourne, Australia. Andy, a good friend, has an AKSA/ Magneplanar system with active crossover, a very impressive system indeed. Here's Andy's article offering enhancements to Lewis' original recipe, and it is both easier to make and should give some sonic improvements. If you'd like to contact Andy to discuss his design, he can be reached at Andy.Redwood [ampersand]bigpond.com

After first building a 4m bi-wired set of Louis Muratori's variant of the Jon Risch cross-connected speaker cable (for my "second" system), I started to think about ways the cable could be improved. Then the time came to replace the speaker cables in my main music system, so I set to work!

First of all, with my active 3-way Maggie speakers, Net folklore has it that it's beneficial to have very thick cables for the bass panels - I know of some who uses 9awg cables here. As Louis's variant on the CC cable uses foil strips cut from Solo/Alpha Core crossover inductors in place of the standard coax braid, at first thought, making up "extra low gauge" speaker cables appears to be problematic as the copper foil is very thin (about 0.075mm by my calcs, if a 45mm wide strip is 12awg) and there is a limit as to how thick a foamed-teflon insulated, solid-core central conductor can be.

However, with some help from Audio Asylum inmates, I was directed to Belden 89292 which has a 14awg, solid-core central conductor with FFEP dielectric. Of course, if you're happy to use stranded conductors, you have more choice!

For the mid-panel and ribbon cables, I used Belden 1506a, which has a 22awg, solid-core annealed central conductor.

The beauty of Louis's variant, apart from making the whole cable out of "solid-core" components, is that the width of the foil can be adjusted to add more overall copper thickness. For instance, 14awg has a cross-sectional area of 2.08mm sq. Add a 30mm wide strip of 0.075mm foil and you end up with a total area of 4.29mm sq - which is just thicker than 11awg!

However, a 30mm wide strip will wrap round the foamed-teflon central conductor about 1½ times so, to uphold "solid-core" principles (adjacent conductors must be insulated from each other), you need to make use of the thin plastic insulation strip which is used to separate the layers in a foil inductor, to prevent the overlap from making contact with the underneath layer of copper foil.

The second improvement which can be made to Louis's design is to do away with the heatshrink! After all, the reason for the heatshrink is to:
1. hold the foil in place round the central conductor
2. stop air getting to the copper foil surface, and
3. electrically insulate the foil.

Instead of using heatshrink, I spiral-wrapped plumber's teflon tape tightly around the foil strip - with about a 2/3rds overlap. The very thin foil does roll easily around the central conductor's dielectric so all that's needed is to keep the top layer of the overlap "tied down" tightly … and the teflon tape is certainly strong enough to do this. Thus:
· the teflon tape now holds the foil in place quite adequately - possibly even tighter than the heatshrink!
· the foil is now covered with a couple of layers of teflon tape, so it won't oxidise, and
· the foil surface is already electrically insulated,

… so there is no need for heatshrink over the top. This must be better, sonics-wise!

Then I put two layers of 1/8" polyester braid over the teflon. This provides some additional "constriction" and I felt made the cable look a bit more elegant. And the 2 layers, in conjunction with the Teflon tape, will certainly ensure complete electrical insulation of the copper foil surface.

Andy Redwood