Skin Effect in Wires.

Status
Not open for further replies.
Jump to Latest
#221
mchambin said:
Yes, but could we have estimates ( based on usual cables and routed on the ground ).
Click to expand...
Yes, you can do a rough estimate. Start by assuming that the RF characteristic impedance of a typical pair of cables is about 80 ohms (less if tightly twisted, more if spaced apart by a web of plastic). Next assume that the velocity factor will be around 0.8, as the dielectric is partly plastic and partly air.

We know Zrf = sqrt(L/C). We know v = 0.8c = sqrt(1/LC), where L/C are H/F per metre (or feet if you are in America). Hence L = Zrf/v and C = 1/(Zrf x v). Plug in numbers to get L = 334nH/m and C = 52pF/m.

Both vary as the logarithm of the ratio of conductor size to separation, so won't vary much i.e. you can assume that most cables will be within a factor of 2 of these results.

simon7000 said:
Turns out the design of loudspeaker cable is not trivial resistance limited once you start getting longer than about 50 feet and full range to 20,000 Hz.
Click to expand...
Yes. For almost all domestic audio, which is usually what is being discussed here, resistance dominates but is still small so skin effect is negligible.
 
#228
DF96 said:
If the plait is equivalent to a tight twist (which I doubt) then it will be quite good at rejecting RF. You have simply spent extra time making a cable which looks nice. If it is not equivalent to a tight twist then you have spent time making a cable which performs worse than a much simpler cable. Fortunately in many cases RF rejection is not required.

Plaiting for audio cables is surprisingly popular; it looks nice, but has no special electrical properties and in many cases it will be inferior to a simpler arrangement. I cannot think of any circumstances in which it might be better than a simpler arrangement.
Click to expand...

Whatever ...... I have been to John Inlow's website many a time to study what he does, and eventually came across his own take upon diy twisted speaker cables. I am averse to verbatim copy other folk's ideas, and decided that a plait, rather than a twist, would work better for my needs.

The teflon sheathing on the auto loom wire used meant that I was able to plait the wires, (two positive, two negative), together very tightly and evenly along the entire length. I compared the visual differences between a 2P + 2N = 4 wire plait and a 2P + 2N = 4 wire twist, how the cancellation process actually works, and decided neither one was better than the other - that a plait is like a twist, and a twist is like a plait, and both achieve the same objective.

So no, I don't think I have wasted my time making would you consider a nice looking but inferior cable, and cannot see why you suggest that something simpler would work better.

How much simpler than a 4 wire RF rejecting plait can you get? ToS
 
Last edited:
#233
scottjoplin said:
That's ok then, often people plait 3 wires which don't
Click to expand...

Yep, dat's right, which is why I ruled out a 3 wire plait! ....... glad we agree.

Also, the very slight discrepancies at each hand overlap of the wires means the RF cancelling effect is enhanced in a way similar to hand scatter wound guitar pickups which have a smoother tonal response - if - compared with machine wound pickups. Of course, I could be completely wrong about this, but I still have all my fingers and toes.

I have seen folks do a 2P + 2N x 4 = 16 wire plait, but I don't get it. ToS
 
#235
tapestryofsound said:
I puzzled over the physics of speaker cables for quite awhile before deciding to make my own from 24 strand teflon coated automotive loom wire, using a four wire plait - mainly because it is of reasonable quality and cheap. Theory has it that a 4 wire plait enables the resultant cable to more or less self-cancel out any radio frequency interference, or other forms of electromagnetic modulation well beyond my ken.
Click to expand...

I don't care if it cancels RFI. (it does btw). I love how it looks. Nice work.

jn
 
#236
Plenty of gimp, lanyard, scoubi, scoobie, boondoggle, color wraps, neckspressions, or rex-lace
Scoubidou - Wikipedia

220px-Scoubidou_6_strands.jpg
220px-Scoubidou_3_strands.jpg

I search for a Moebius scoubidou....Not invented yet.
I know a Moebius ring can be done with knitting, and there is no cheating, no sewing, usual knitting with one wire.
 
Last edited:
#238
".I suspect your measurement equipment, your spool material, and technique. For a reasonable jacket thickness, I can get the spool number to match the unspooled. The only caveat to that is when the spooled distributed capacitance starts to cause resonance. That is also why I never recommend a simple one frequency measurement, I do 20, 50, 100, 200, 500.....for ease of log graphing. "

JN,

I have a nice General Radio Digibridge and reference capacitors and inductors. Resistors are just too common to worry about. The spools are wood. I do find a bit of difference between spooled and laid out straight four different measurement methods. I should have a short length around 100' I can try both straight and coiled.

And what do you get for inductance of any given cable?
 
#239
A while back the conversation was about Teflon insulated silver plated copper stranded wire. I randomly gobbed together the hundred feet of wire and shoved into a steel box and did a log sweep for the measurement.

The silver plated wire had just measurably more distortion than tin plated stranded wire.

Thanks DT
 
Status
Not open for further replies.
Top Bottom