Interconnects - Magnet wire?

Cat 5 and magnet wire - very good speaker cables . . For interconnects, I’ve read that Cat 5 can work very well.

I’m moving my sources & “pre-amp” next to where I sit . . it makes room up the more visible end of the room, for some big speakers

Interconnect lengths will be about 5 m.

I’ve got quite a few to do (including another room o0f similar lengths) so for speed of putting them all together *if sound quality is likely to be about equal, I’d prefer solid core over multi strand (cat 5).

I don’t know if any of this effects things:
the “pre-amp” is a Promitheus Audio transformer volume control ( The TVC’s output impedance depends on the volume setting, but is very low, typically c 2 - 25 R. As TVCs add virtually resistance, the impedance at it’s outputs effectively derives from the source equipment. The source is mostly CD, though in time it will become the PC. The power amps for the foreseeable future will be SS.

So . . would solid core magnet wire work as well as cat 5?

I’m considering a local magnet wire of outer size of 1.25 mm (core = 1.0? mm) ie cross sectional area about 0.8 mm^2. As cross sectional area (CSA) decreases, I think resistance less? Does the TVC’s lower output impedance mean less interconnect CSA is needed?

A search showed that some believe that if the is cables including cat 5 are Teflon-wrapped (C5 plenum grade), it’s better, not sure how. If such a thing might exist, is it worth seeking out Teflon-wrapped magnet wire?

Tho’ if cat 5 is better, how many strands should be enough for that length?

flat shield braid is also a good choice. flat conductors have very little skin effect since the magnetic fields are concentrated on the outside edges.

solid magnet wire has a lot of skin effect, stranded has less, litz wire(made of separate enameled strands of magnet wire) is better than stranded (and cat 5 cable behaves like litz wire), and flat ribbon conductors have the least skin effect.
Other way around there, as the cross sectional area INCREASES, resistance DECREASES.

At any rate, the resistance of the wire will be negligible compared to the source impedance.

Using shielded twisted pair wire could be beneficial, in the form of lower noise pickup. Some CAT5 wire is shielded, some isn't. If you don't ground the shield it won't do much good however.

The dielectric type is not significant unless you believe it is.

If you are very concerned about the long interconnect length, consider balanced connections far more important than the type of wire used.
I'm not contradicting you there UncleJed, and I fully understand skin effect as I deal with radio frequencies over long distances daily. I didn't see your post until after the reply was sent. The OP post did not reference skin effect and incorrectly stated that a decrease in cross sectional area would result in a decrease in resistance.
> the resistance of the wire will be negligible compared to the source impedance

Even over 5 m (16 feet) - I didn’t know that

flat shield braid? (flat conductors would fit better under carpet.)

I found this

but haven’t yet found a supplier or a price. My hunch is it would cost more, are there other brands/ who is your supplier?

Thank you
I design sonar transducers for a living, and have access to all forms of neat shielded twisted pair cables. We routinely transmit up to 800kHz over distances longer than 5m.

PM me if you are in the USA.

There are way more important things to worry about in audio than cables.
> There are way more important things to worry about in audio than cables.

Agree 100%. I simply want to select something non-deleterious, quick to build and economical, use it and forget them.

While I’m not in the USA (in Australia), US brands are usually readily obtainable over the net - if you could advise a major supplier/ brand, I’d appreciate it: rick57 at (or here)

mentioning Belden triggered me recalling Jon Risch's very handy (but slow) site.
Main page
diy Interconnects

His results are “compiled over the years while listening to cables” . .
though eg “the recommendations for insulator materials follows the ranking (more or less) of those same materials for their dielectric constants, i.e. Teflon (best sounding) has the lowest dielectric constant, and PVC (worst sounding) has one of the highest” . .

From Jon’s long article on interconnects there:

For a line level interconnect, I recommend Belden #89259 as my first choice.
It is of coaxial construction, with bare copper braid and shield coverage of 95%, an insulator of foamed Teflon around a 22 Ga. stranded bare copper wire, and an outer jacket of black tint solid Teflon. Capacitance is a low 17.3 pF/ft, almost half the capacitance of common interconnect cables.
It is superior sonically to many of the expensive/ high-end cables sold for hundreds of dollars, and can be terminated with your choice of RCA plug, anything from the nickel plated Switchcraft RCA to the Cardas/ Kimber/ Tiffany RCAs. It is good enough to be worth using the finest plugs. Available in 100 ($100) and 1000 ($950) foot spools.

2nd choice: Belden #82259
Same as above except with a Flamarrest jacket. May be more flexible than the above, have not handled a sample. Available only in 1000 foot spools ($750).

He later updates with several labour intensive, more expensive interconnect recipes for 89259, with braided shield, 89248, Teflon plumbers pipe thread tape etc.

He goes on to cover best digital ICs, etc etc etc
For sonar transducers, we are receiving signals at the microvolt level from the transducer and transmitting this passively over long cables (averaging 6m) to the sonar receiver. The entire chain from the transducer to the receiver typically has 90dB of dynamic range and 60dB of channel separation.

For cables, capacitance is the main factor we look at aside from proper dielectric strength (we transmit over 1000V so that is an issue for us, not so much in audio at 2V or so). We tend to use individually foil shielded twisted pairs, with polypropylene insulation and PVC jacket. For prototypes I usually purchase Belden or Alpha Wire, though we have cable custom made for production use.

The reason I asked about your location is that I have plenty of spare wire around, and would be glad to send you some if you pick up the shipping tab.


The comments from the folks above are correct regarding types of cable.

To me 5m is a long interconnect and I have found that with that length you need to take some precautions that would not be so important with a 1m interconnect. If your interconnects are 5m long then I have found that a shielded cable is best. You need to connect the shield at the pre amp so that you effectively extend the pre amp shield around the interconnect to shield the signal wire inside. For a shorter length this would be less important. I have found stranded cables better than solid for interconnects. However this is a contentions area and you will get many views.

If you can work it out then matching the pre amp and power amp impedances is probably the most important issue in getting the sound right.

Speaker cables need to be short and thick.

Hi Rick (Otto),

I found Jon's information (after the fact) and agree that dollar for dollar you can't beat it! If you want to dress them up get a braid material (like chinese handcuffs) and some channel (right and left) appropriate heat shrink at each end.

After reading his info, I never felt that making his "better" ones was necessary. (diminishing returns make me look elsewhere for better places to improve my system).

My cousin spent $350 on two Silver Interconnects and was ****** when mine was easier to listen to (less fatiguing) at a fraction of the cost.

I don't see any reason for me to look elsewhere for ICs at this point.

However, I do use Magnet wire for my speaker cable and I think it is another under-rated great value (for speaker cable). And again, my Eichmann Bayonets (banana plugs) were what cost the money. I replaced Onix Bi-Wire speaker cable and there is no comparison.

For ICs, maybe you can use Cat-5 and Jon's recipe for shielding and insulation for your own recipe. From what I understand, all insulation material affects capacitance and sometimes inductance and resistance too and so does twisting, so expeimenting to find the best balance is what this is going to take. Then you will just use the Belden because it is already there! LOL.

Good luck!

Cat5 is not stranded.
Cat5 patch cables are usually stranded.
Magnet wire = enameled copper, has very thin insulation.
A twisted pair will be closer together than a Cat5 twisted pair and so for the same core diameter the Cat5 will have slightly lower capacitance. However the dielectric varies between the various insulators that are available and this will have an effect.

In general, the twisted pair when used unshielded has good rejection of interference when at least one end (usually the send end) is connected to a lowish impedance
the original question was speaker cables. cat5 is usually 4 pairs of 24ga solid copper. since it's insulation is thicker than magnet wire, it will have less proximity effect than litz wire(and less skin effect), and with all 8 strands wired in parallel, has the same current capability as 16ga wire. actually, i wouldn't worry about skin effect in speaker wires, unless a) you've got very long speaker wires, or b) you're running 2 ohm loads. as far as capacitive effects go, the largest effect cable capacitance will have is whether or not your amp can drive the additional lag without oscillating, otherwise you're talking millidb's of loss at 20khz..... look sometime at what is used INSIDE amplifiers to connect the pc board to the speaker terminals
Joined 2004
Paid Member
Hi unclejed613,
the original question was speaker cables.
Yes, but you didn't "get" the rest of the opening statement ...
Cat 5 and magnet wire - very good speaker cables . . For interconnects, I’ve read that Cat 5 can work very well.

I do that sometimes too.

You know what I think?
They have different wire designed for different jobs. Cost is also an issue some times. That does not mean that expensive industrial cables make good wire for audio.

Knowing what I know about instrumentation, I fully agree with David on this. I also work with Cat-5 a lot. I do not use it for audio cables.

Proper impedances also go a long way to getting the signal from one place to another without too much "damage". :)