Speaker Cable Capacitance and Amplfier Stability

[JCLNV]

Hi All,

Sorry this isn't really a DIY question but I think a good few people on here really know there stuff so here goes.

I have some Kimber 8PR cables that, according to the Kimber website, has a (Cp) parallel capacitance:742.0 pF @ 20 kHz for a 2.5 metre pair. I'm a bit worried about using them with my little Lavardin amplfier as they recommend (like Naim etc) avoiding high capacitance cables. I know this isn't very high capacitance like a Goertz cable for example but just what is a 'high' capacitance cable ? And does the speaker impedance (8ohms in my case) play a part in this ?

Thanks for any advice in advance.

[Geek]

Hi,

Is this for line in?

Capacitance is one of the reasons I make my line cables from RF coax, like RG8 mini foam. That would give 140pF for 2.5M.

742pF seems a little nuts for 2.5M, IMO.

Cheers!

[SY]

I think that this is for speakers. Yes, there are (too) many amps which act oddly or even destructively when loaded with even moderate values of capacitance. I'm unfamiliar with yours- have you asked the manufacturer directly for advice? Does it use an output coil?

You can risk the amp but spare your speakers by hooking up an 8 ohm dummy load, then paralleling that with 1000pF or so and see if the amp screams and smokes. Or you could just chuck the high capacitance cables and use something less glamorous but more sensible.

[Johan Potgieter]

A rather late reply as I haven't visited threads widely lately.

742pF (if that is correct) is an extremely high capacitance for 2.5 meter loudspeaker cable - what misguided design route did these fellows use?

The lowly ripcord (1 mm) measures only 400pF for 4 meters. Of the more professional types measure < 200pF for 3 meters. Also I am very amazed that amplifiers exist which could not tolerate this sort of thing. (If Sy did not acknowledge that, I would be very hard-put to believe it.)

JCLNV, I was hoping that you might have a scope and square wave generator at hand or available. The best test is to hook the lot up and watch the scope for accentuated overshoot or oscillation (and shut down immediately in case of the latter). Sy mentions an 8 ohm resistor and 1000pF load; fair enough, but one must remember that at 20 KHz most loudspeakers are way above 8 ohms - more likely 20 - 30 ohms inductive, and increasing. My best thoughts would be that those amplifiers that do object, do so because of resonance with the loudspeaker. impedance/cable capacitance at some supersonic frequency rather than just cable C.

But explanations aside, you have what you have and I would suggest the above square wave test if feasible, to put your mind completely at rest.

[jneutron]

Originally posted by Johan Potgieter
[B]742pF (if that is correct) is an extremely high capacitance for 2.5 meter loudspeaker cable - what misguided design route did these fellows use?

Why is that high?? 2.5 * 39 = 98.5 inches / 12 = 8.2 feet

742 / 8.2 = 90.4 pf per foot. Sheesh, that's not bad.

Why is it misguided? If it's fully constrained with foam dielectric, the inductance will be 11.4 nH per foot, cable Z of 11.25 ohms....if it is a typical dielectric (2.7), it will be 31 nH per foot, and 18.5 ohms cable Z. If unconstrained, L and Z will be higher, but without L or the geometry, constrained is the lower limit.

It would appear that their design guidelines were to get the cable impedance down to within the speaker impedance realm, but not all the way down to 8 ohms, either due to capacitive concerns, or geometric constraints.

But choosing a design impedance that low is not misguided, it's an attempt to reduce the inductance as much as possible, to keep the cable energy storage minimal.

Cheers, John

[jneutron]

Looked at the kimber site.

L of 56 nH per foot.

LC/1034 = 4.87

Not very well constrained...geometry defined effective dielectric constant of 4.8..

Z = 25 ohms.

Well, it's better than zip spec wise..but I still don't know what their design criteria was.

Cheers, John

[Johan Potgieter]

Hi John!

Glad to hear from you; lately I am not here that often.

By high I meant relatively. Not to mention brands: For 2,5m length, a quite expensive cable I measured showed 180pF. RG58/U (a co-ax cable) had about 253pF. The lowly rip-cord had 250pF. Other moderately expensive types (not exotic) were 60 - 80pF. That makes the Kimber capacitance 3x higher than the highest out of about a dozen I checked.

In that sense they must have thought the use of whatever di-electric material/conductor diameter/inter-cable distance they used justified despite the popular concept that even loudspeaker cable must have low capacitance (not saying I share that). I simply expressed wonder at the rejection of such a "selling point" - in favour of what?

I am not checking or disputing your analysis (I know better!), but I think both you and I know better than to involve "speaker impedance" (another "selling point" with some) for cable lengths of the order of < 0,1% of a wavelength, and cable storage effects in the presence of loudspeaker impedances .... But yes, those are design goals with some; still, excuse my use of the term "misguided".

Not to re-re-re-open this can of worms here; just a reply to yours.

Regards

[phase_accurate]

According to NP the impedance mismatch at shortwave frequencies can lead to amplifier instability, in his opinion more so than the influence of the cable-capacitiy on phase-marging.

He recommends the use of a Zobel at the speaker end of the cable in order to achieve at least some impedance matching at high frequencies.

Regards

Charles

[jneutron]

Originally posted by Johan Potgieter
Hi John!

Glad to hear from you; lately I am not here that often.

Good to hear from you also... I've also been other places. Just got back, and trying to find my pulse..

Originally posted by Johan Potgieter

By high I meant relatively.

Ah, ok. No prob.

Originally posted by Johan Potgieter
I am not checking or disputing your analysis (I know better!), but I think both you and I know better than to involve "speaker impedance" (another "selling point" with some) for cable lengths of the order of < 0,1% of a wavelength, and cable storage effects in the presence of loudspeaker impedances .... But yes, those are design goals with some; still, excuse my use of the term "misguided".

If one is concerned with SWR concerns, then matching line to load does indeed appear to be of no consequence to audio. That's what they taught us all those many years ago.

When one analyzes the speaker run as a transmission line anyway, it is not as clear as one would like. Mismatching line and load does change the response in a fashion not alluded to by our schooling.

Originally posted by Johan Potgieter

Not to re-re-re-open this can of worms here; just a reply to yours. Regards

Again, not a problem..

Originally posted by phase_accurate
According to NP the impedance mismatch at shortwave frequencies can lead to amplifier instability, in his opinion more so than the influence of the cable-capacitiy on phase-marging.

He recommends the use of a Zobel at the speaker end of the cable in order to achieve at least some impedance matching at high frequencies.

Regards

Charles

Who is NP?

Cheers, John

[AJinFLA]

Originally posted by jneutron
Mismatching line and load does change the response in a fashion not alluded to by our schooling.

But is this audible with a music signal in a room with a typical multi (largely uncorrelated) driver loudspeaker? Enquiring minds want to know.

Originally posted by jneutron
Who is NP?

My guess would be Nelson Pass. Much nicer than our old pal JC

cheers,

AJ