Cable Long Runs.Experience,Theroy

the

Why not give us some typical examples then! Not hand selected exceptions. Or in your case, NOTHING at all.

I suggest you pull your head in or give us some REAL figures to prove what a pedantic prick you are being.

Measuring skin effect,

Yep, sure is if you have the right equipment, and the results at 20kHz are pretty trivial.

We do know typical figures for Hi-Fi equipment. Obviously you want to prove you are right by using outrageous examples.

And adding NOTHING usefull at all, simply grandstanding it would seem. I hope you are feeling suitably smug.

MrT.

Reply to
Mr.T
Loading thread data ...

Why don't you use hollow core tubes or waveguides?

MrT.

Reply to
Mr.T

"Jasen Betts"

** Bye Bye, Jasen, bye bye......

......... Phil

Reply to
Phil Allison

If the calculations on this website can be trusted, the effective skin depth in copper at 20kHz is 0.5mm. So with very large diameter cables (high powered systems) this could easily influence HF response with low impedance loads.

formatting link

There is some conjecture about stranded cables. Apparently if the strands are individually insulated (ie Litz wire) they are effective in avoiding skin-depth problems but if they are making electrical contact the *bundle* acts similarly to a single cable, with some differences.

-Andrew M

Reply to
Andrew M

"Andrew M

** Nope.

Simple, linear inductance has a *much* greater effect on cable impedance than copper resistance at high audio frequencies.

A 1 inch long copper wire of 1 sq mm cross section has as much inductive reactance as resistance at 3kHz.

....... Phil

Reply to
Phil Allison

I guess that's why they use paired conductors...

-

Bye. Jasen

Reply to
Jasen Betts

"Jasen Betts" Phil Allison

** Only works when constructed as a transmission line and terminated with the load equal to the characteristic impedance.

A twin or twisted pair cable does not match a 4 or 8 ohm speaker load not is a speaker a defined impedance.

........ Phil

Reply to
Phil Allison

Thanks... It dose produce regluated output however could be set not to by removing the voltage regulator. It is a car alternator,and requires about 920rpm to kick in as a guess as for power vairation above this speed I have not got there yet,just at finding out if I can do this cost vs. commercial so to speak. I like this idea to get started....upgrade to more volts(better trubine/alternator) down the track.

Reply to
Magic Mushroom Farmer

Yes....like this idea,however for first time setup it is alittle bit of an overhead due to some unkown factors such as turbine speed,I may be lucky to get 920 odd rpm out of it to power my car alternator however will try,I have seen people claim upto 3kva from 30rpm on homemade models using custom magnets and windings/parts.

Reply to
Magic Mushroom Farmer

Yeah........It is about $4k-$5k to get the commercial stuff,also already pay one bill and with a roaring water course,it is worth a good hard try,also have the option to pipe water for a possible 10foot fall,however gallons per minute would be limited to about 4 gpm. I will attempt to use water current to power an under running water wheel for a first time attempt.

Reply to
Magic Mushroom Farmer

yes they are pretty trivial.

delta = 1/sqrt(pi*f*mu*sigma)

mu = permeability of material = 4*pi*10^-7 * mu_r,

mu_r = relative permeability of material = 1 for copper

sigma = conductivity of Copper

bung in the numbers, and for copper at room temp you get:

delta = 66mm/sqrt(f) where f is in Hz. which aint that hard. at 100C it changes to 75mm/sqrt(f).

at 20kHz, delta = 0.467mm in Cu at 20C, and 0.53mm at 100C.

easy. a signal generator and an audio amplifier, along with a length of non-inductively wound cable (eg a single tightly twisted loop), say

4mm^2. Use a CT (or a small resistor) to measure the current, and slowly wind up the frequency. watch the current drop, proportional to 1/sqrt(f). 4mm^2 = 5mm OD. At 20kHz, it looks like a 5mm OD hollow pipe 0.5mm thick. the AC/DC resistance ratio at 20kHz will be about

Fr = R20kHz/Rdc = 5mm^2/(5mm^2 - 4mm^2) = 2.8x

At 100kHz, it will be 6.5x the DC value

in a choke, 4 layers (one atop the other) of 4mm^2 Cu winding will have Fr = 90 at 20kHz. proximity effect makes things get a lot worse, a lot faster

If you were making, say, a 2kW tweeter, you had BETTER pay attention to skin and proximity effect. normally there aint that much energy at the higher frequencies....

I'd expect to see proximity effect in mid-range and tweeter crossovers. Again, the severity is reduced by the lower power levels at higher frequencies.

HTH

Cheers Terry

Reply to
Terry Given

that I'd believe.

I did a pair of 15m cable runs in the loung of my last hous (it was a BIG lounge), 4mm^2 twin cable. I worked out the inductance (I forget the cable dimensions), and at 10kHz it was a sizeable chunk of my 4Ohm speaker impedance - about 10% IIRC, so perhaps 6uH. compare that with the 0.165Ohm DC resistance....

So I ignored it, and was still happy. I could have twisted the speaker cables, but that required effort, and climbing up a scaffold.

OK, assume resistivity of copper = 22nano-Ohm-meters (pure Cu is 18nOm).

R = rho*l/a = 22e-9 * 25.4e-3/1e-6 = 0.558 mOhms

0.558mOhms/(2*pi*3e3Hz) = 29nH

yep, thats about right, 1nH/mm

1mm^2 = 1.13mm diameter.

skin depth at 3kHz = 1.2mm, so the effect is negligible. bit of a straw man, really.

but if you were trying to build a 29nH inductor, it would be very lossy at 3kHz.

and if you were trying to build an 0.558mOhm resistor, it would be horribly inductive - the magnitude of its impedance is 41% higher than expected, and its step response would be evil :)

lets make it a real test, at 20kHz. skin depth = 0.47mm, so Fr = 2.5 and Rac = 1.35mOhms

Xl = 3.6mOhms

now that is a seriously piss-poor 0.558mOhm resistor, but only a moderately shitty 29nH inductor.

Xl starts to overcome Rac, because it increases linearly with f, whereas Rac increases as sqrt(f), so a lot slower.

Cheers Terry

Reply to
Terry Given

more likely because 2 wires is a pain....

rubbish. That wont help reduce the inductance or skin effect at all!

What it will do is prevent ringing due to reflections (electrically "long" cable runs = distributed circuits) or L-C resonance (electrically "short" cable runs = lumped circuits). and a simple RC damper would do that.

twisting the wires will reduce the inductance, it should be easy to get a tenfold reduction in inductance. It will increase the DC resistance (more cable), but make *NO* difference whatsoever to the skin effect.

no reason why you cant RC terminate a long speaker cable. not that you'd ever hear cable resonance....for my long cables, C = 2nF or so, and L =

6uH, so Fo ~ 2MHz.

conjugate matching is a bad idea for power delivery, as the source dissipates as much energy as the load gets. RF power amps dont do it, why would audio?

I presume you meant "NOR is a speaker a defined impedance"

thats an understatement.

do you have a schematic of a really, really good crossover? AIUI its quite tricky designing a good crossover, theres all the pseaker resonances, as well as those of the crossover itself.

Cheers Terry

Reply to
Terry Given

"Terry Given"

** Piss off, Terry - you have no idea what I even said.

** Of course a matched transmission line eliminates the effect of linear inductance.

What second rate NZ anti-ADHD drugs are you on now ?

** Not even faintly related to my remark.

Given is an extreme ADHD victim.

......... Phil

Reply to
Phil Allison

but it wont reduce the inductance. and the inductance governs the overall velocity of the cable.

the context was the reduction of inductance.

Reply to
Terry Given

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.