Speaker Cable Length

0 Members and 1 Guest are viewing this topic. Read 5142 times.

rollo

  • Industry Participant
  • Posts: 5440
  • Rollo Audio Consulting -
Speaker Cable Length
« on: 10 Nov 2010, 06:43 pm »
  Is there any sound reason why the return leg must be the same length as the supply leg ? Why can't it be shorter ?
  Second question, is the return signal put back into the signal or sent to ground ?

charles

Speedskater

  • Full Member
  • Posts: 2672
  • Kevin
Re: Speaker Cable Length
« Reply #1 on: 10 Nov 2010, 06:55 pm »
If the send and return wires in a speaker cable are not in close proximity to each other for the entire length of the cable, then the self-induction of the cable increases, this may reduce the high treble volume.
Any signal (speaker or inter-connect) consists of a send part and a return part, the return may or may-not be connected to ground. 

weitrhino

Re: Speaker Cable Length
« Reply #2 on: 10 Nov 2010, 07:29 pm »
For a solid state amp the return signal does not get backfed into the amp input again due to the low impedance output of the amp.  However, tube amps will 'hear the room' feedback.

*Scotty*

Re: Speaker Cable Length
« Reply #3 on: 10 Nov 2010, 07:40 pm »
Feedback loop?



Scotty

Speedskater

  • Full Member
  • Posts: 2672
  • Kevin
Re: Speaker Cable Length
« Reply #4 on: 10 Nov 2010, 07:41 pm »
I think that we are talking about wire legs, rather than the signal generated by the amplifier as opposed to the signal generated by the speaker acting as a microphone.

Johnny2Bad

Re: Speaker Cable Length
« Reply #5 on: 10 Nov 2010, 10:26 pm »
There are reasons, mostly dealing with how your amplifier deals with the speaker load. The cable is part of the speaker load, as far as your amp is concerned. Your amp wants to see a low inductance cable there. Twisting, braiding or pairing the cables lowers inductance. So, that is one answer to your question right there.

I think there is an unanswered question, which goes like this: why do you ask? What configuration of amp -> cable -> speaker do you want to use?

Depending on what you want to do, there may be solutions that will work, but if you're asking what is ideal, the answer is as others have already said.

You're free to use a non-ideal cable configuration, but there has to be a reason; otherwise I don't think you would be asking the question. Tell us that, and maybe we can help you do it the best way possible, given the limitations you (seem to) need to impose.

For a solid state amp the return signal does not get backfed into the amp input again due to the low impedance output of the amp.  However, tube amps will 'hear the room' feedback.
I'm not sure I follow you here. The back-emf comes from the speaker motor; it's more than enough to overcome the low output impedance of a typical SS amplifier, and would happen if the amp output was directly connected to the speaker with a theoretical "no cable at all". I don't understand what you're referring to (which isn't to say you're wrong ... I just don't know what back force you're referring to).

Speedskater

  • Full Member
  • Posts: 2672
  • Kevin
Re: Speaker Cable Length
« Reply #6 on: 10 Nov 2010, 10:42 pm »
I don't think that any modern well designed amplifier cares what the self-inductance of the speaker cables is!
But high self-inductance speaker cables do become another component in the cross-over network.

Johnny2Bad

Re: Speaker Cable Length
« Reply #7 on: 10 Nov 2010, 10:54 pm »
I don't think that any modern well designed amplifier cares what the self-inductance of the speaker cables is!
But high self-inductance speaker cables do become another component in the cross-over network.

I'm not convinced everyone uses a "modern well designed amplifier"; it's my personal belief that people use whatever they happen to have or what they can afford. Certainly in my case, most of my life, that's been the case.

And I've been able to send very highly regarded, "modern well designed amplifiers" that work fine on conventional motor loudspeakers into expensive piles of smoke and silicon by hooking up a pair of Dayton-Wright electrostatics, which is primarily a capacitive load.

Others have had similar experiences with Heil AMT's, and I have with Maggies and Arnie Nudell's version of Infinity speakers.

Surely the amps didn't just spontaneously take up tobacco, blow an oil seal, or become critically depressed and pick that exact moment to end it all with no suicide note and plenty of drama.

Call me crazy, but I think it had something to do with the speaker load, of which the cables are a part. I've also seen fairly ordinary speakers paired with high inductance cables like the Goertz pretty much have a "spinning, screaming 2-year old on too much sugar" reaction where the amp was fine with zip cord.

Resistance, Inductance and Capacitance of speaker cables certainly can affect an amplifier. Otherwise nobody would know what a zobel network was and no-one would use them with our modern, well designed amps.

The OP hasn't told us what amp, what speakers, and what cable configuration he wants to use. I'm going to avoid assumptions and wait for him to tell us what he really wants to do or know; there's too much missing from his original post to answer any of this.

Give me enough time and easy access to the world's commercially available speakers and cables, and I can blow up pretty much every amp out there just by playing music loud, and I don't mean "too loud", and do it just by altering the load (ie a single commercially available hifi speaker and a single commercially available hifi speaker cable per channel).

I know of a few speaker and cable models from real-world mainstream manufacturers that will make short work of half the amps in existence.

I also know a few amps that can drive pretty much whatever you throw at them, without distress whatsoever. Give me a Bryston 4B and a Quad 33, and I'll run them into any speaker and load combo you can throw at me. The Bryston will play into very difficult loads that kill other amps. There are others that are just as robust, just as there are those that are finicky.

The Quad 33 will handle it better than the Bryston will; it will run into a direct short all day without damage, and with anything even a teeny itsy bitsy less than a direct short, some music will come out, I don't care what the load.

You could even use a Quad 405, although after ten minutes at full power into a direct short, it might sustain some damage. Not at nine, though.

I don't see the point of generalizing about it; let's get the specifics and take it from there.

weitrhino

Re: Speaker Cable Length
« Reply #8 on: 11 Nov 2010, 12:09 am »
Johnny2Bad said...

Quote from: weitrhino on Today at 07:29 PM

For a solid state amp the return signal does not get backfed into the amp input again due to the low impedance output of the amp.  However, tube amps will 'hear the room' feedback.

End Quote.


I'm not sure I follow you here. The back-emf comes from the speaker motor; it's more than enough to overcome the low output impedance of a typical SS amplifier, and would happen if the amp output was directly connected to the speaker with a theoretical "no cable at all". I don't understand what you're referring to (which isn't to say you're wrong ... I just don't know what back force you're referring to).

Have a look at Bob Carver's explanation at 51 minutes into this video.
http://www.youtube.com/watch?v=LQ9USqpclWc

audioengr

Re: Speaker Cable Length
« Reply #9 on: 11 Nov 2010, 12:58 am »
If the send and return wires in a speaker cable are not in close proximity to each other for the entire length of the cable, then the self-induction of the cable increases, this may reduce the high treble volume.
Any signal (speaker or inter-connect) consists of a send part and a return part, the return may or may-not be connected to ground.

Quite right, except its mutual inductance I think.  There are two ways to get low inductance, low self-inductance using very large gauge wire or mutual inductance of lots of smaller twisted pairs.  They both work.  I'm not so sure about litz wire.  All of the currents going in the same direction will increase inductance.

You might find these interesting:

http://www.empiricalaudio.com/computer-audio/audio-faqs/short-versus-long-cables

http://www.empiricalaudio.com/computer-audio/audio-faqs/are-cable-resonances-real

http://www.empiricalaudio.com/computer-audio/audio-faqs/should-i-match-impedances

Steve N.


rollo

  • Industry Participant
  • Posts: 5440
  • Rollo Audio Consulting -
Re: Speaker Cable Length
« Reply #10 on: 11 Nov 2010, 01:49 pm »
 I'm using a SET amp and Pipedream speakers with sub. The reason I asked is that my favorite speaker cable [ return ] split and needs repair. So I inserted a 3' pair of Tara "Space & Time Return while still using an 8 ft. run of OmegaMikro [ active].
  Actually wasn't trying to accomplish anything however it sounds much better than using the 8 ft Omega return . Was puzzled as to why it sounded better . Thanks for the responses so far.


charles

Johnny2Bad

Re: Speaker Cable Length
« Reply #11 on: 11 Nov 2010, 07:41 pm »
Checking out the OmegaMicro cables, I see they are a ribbon type. This kind of cable has properties (due to it's geometry) that are somewhat different than conventional cable geometry.


More specifically, it tends to have high inductance, low capacitance, and low resistance. The low capacitance is typically a design goal and I don't know of any ribbon cables that don't address this.


Broadly speaking this type of cable tends to present a moderate to severely difficult load to the amplifier. A transformer-coupled amp may tolerate it better than a direct coupled stage as found in most SS amps. Better quality amps tolerate it more than a more "ordinary" amp ... ie your Pioneer receiver probably won't like it at all (a zobel network would help there), while separates may be OK with it. Similarly, a given amp with ribbon cables may work fine with one loudspeaker but balk if you switched to a speaker with it's own somewhat difficult load characteristics. Finally, cable length tends to be somewhat more critical with ribbons than conventional construction; short lengths may be fine but with long lengths there will be a point where the conventional cable is the only version that will work at all without audible or protection circuit-based complaints.


By configuring the cable as you have, you may well have lowered the inductance; ie your new configuration is a higher inductance configuration over conventional non-ribbon paired but you already had a high inductance configuration with your original ribbon cables. There will of course have been other changes to the cable's contribution to overall properties but I would expect them to be less significant overall.

When people with the requisite test equipment test cables, they tend to find a conventional cable will show it's highest (mutual) inductance when separated by about 4" or more. Often this resulting (total) inductance value is lower than ribbon types (self inductance) will show when paired as per manufacturer's recommendations.


Perhaps that helps with regard as to why it seems to sound better. Maybe your amps would prefer a more conventional design for both channels, or shorter lengths of ribbon cables to both speakers.

Speedskater

  • Full Member
  • Posts: 2672
  • Kevin
Re: Speaker Cable Length
« Reply #12 on: 11 Nov 2010, 07:48 pm »
Steve N. you're correct it's mutual inductance.  Self-inductance is for a single conductor.

Johnny2Bad

Re: Speaker Cable Length
« Reply #13 on: 11 Nov 2010, 08:58 pm »
Johnny2Bad said...




Quote from: weitrhino on Today at 07:29 PM




For a solid state amp the return signal does not get backfed into the amp input again due to the low impedance output of the amp.  However, tube amps will 'hear the room' feedback.




End Quote.








I'm not sure I follow you here. The back-emf comes from the speaker motor; it's more than enough to overcome the low output impedance of a typical SS amplifier, and would happen if the amp output was directly connected to the speaker with a theoretical "no cable at all". I don't understand what you're referring to (which isn't to say you're wrong ... I just don't know what back force you're referring to).




Have a look at Bob Carver's explanation at 51 minutes into this video.
http://www.youtube.com/watch?v=LQ9USqpclWc


Thanks for the link. I see Bob talking about back-EMF and how tubes and solid state amps react to Back-EMF differently via the feedback loop. I didn't hear anything from Bob that conflicts with what I already understood.

Assuming it's as Bob says ... that the SS amp will tend to dampen the Back-EMF signal while the vacuum tube amp will be modulated by it, through the feedback loop, due to the inherent lower output impedance of direct coupled SS designs compared to vacuum state designs.

Bob implies that Back-EMF is completely suppressed in a SS design. I suggest he is making a pitch on a radio station rather than addressing the IEEE, and that he really would not say that a SS amp is unaffected by Back-EMF to an engineer, but instead would say he believes that output impedance plays a role in how well it is suppressed.

Going back to what Mr Carver says, note also that his talk does not address every possible source of Back-EMF. For example, the mass of the driver will tend to contribute regardless of room interaction; ie in an anechoic environment or in a boundary free environment that the anechoic chamber attempts to replicate.

If we remember our basic physics ... a body in motion tends to remain in motion, and any failure to perfectly control the driver results in Back-EMF applied to the amplifier.

The magnitude of that Back-EMF signal can, and does affect SS amps. I know of no SS design capable of completely controlling a large mass driver in motion. I would not assume that because Bob suggests a tube amp does it relatively poorly it follows that a SS amp does it perfectly, or is unaffected by it.

Going back to speaker cables for a moment, it was simply the way you described it that I didn't understand. I thought perhaps you were talking about the negative cable affecting something on the positive cable due to geometry or electro-magnetic interaction. The Back-EMF does travel along the entire cable length (and does not always manifest itself only on the - cable side), any suppression happens at the amplifier.

Finally, and this is not against Mr Carver personally, but not everyone agrees with his idea of how Back-EMF via the Feedback Loop accounts for a preference to vacuum tubes. (Some amps, for example, have no feedback loop for this mechanism to work within).

I would not assume everyone agrees with his explanation of why tube amps are preferred by some, or, alternately, that it is the only possible difference to explain a preference of SS over tubes, or vice versa. After all, we refer to transistors as "switches" and vacuum tubes as "valves", and it's for a reason. Also, there are some who will will argue that a SS amp's low output impedance does not directly translate to a high damping factor, because the real-world speaker load consists of more than a simple 8 ohm resistor and an infinitely short cable.

jneutron

  • Full Member
  • Posts: 557
Re: Speaker Cable Length
« Reply #14 on: 12 Nov 2010, 09:47 pm »
A piar of wires will have two inductances associated with them.

1.  Self inductance of the conductors.  A cylindrical conductor will have 15 nH per foot of self inductance.  This is INVARIANT of the diameter of the conductor, and is a consequence of the integral of the magnetic field from the center of the cylinder to the outer surface.

To lower the self inductance of a conductor, it is necessary to change the cross sectional profile.  For example, a coaxial outer braid is a thin wall cylinder of current, and there is NO magnetic field within that hollow cylinder, therefore there is NO self inductance associated with the outer shield.

Another example is the use of flat wide conductors.  A copper foil that is 100 times wider than it is thick will have a far lower self inductance because the current has been spread over a wider area, and that lowers the magnetic flux in the bulk of the foil.

Skin effect will also lower the self inductance of a solid conductor.  At the limit, skinning will drop the self inductance a maximum of 15 nH per foot over the DC inductance value.

The larger the conductor, the lower the surface magnetic field.  This should not be confused with the self inductance of the wire.  Wire diameter does not change the self inductance of a conductor.  It's in the math..

2.  External inductance.  The energy stored in the flux lines outside of the conductor.  This is heavily dependent on the conductor spacing.  Further away, more inductance.

Steve is correct in that many twisted pairs, such as cat5e, can lower the inductance of the aggregate heavily.  But only if each wire in the pair has the opposing current in it.

Litz wire will not reduce self inductance, and if it is spaced as another pair of solid, it will have the exact same external inductance.  What is different about litz is the fact that at frequency, the randomly arranged, independently insulated wires guarantee uniform current density.  It also guarantees that proximity of the return conductor will not distort the current distribution such that less of the copper is conducting..

Cheers, John

rollo

  • Industry Participant
  • Posts: 5440
  • Rollo Audio Consulting -
Re: Speaker Cable Length
« Reply #15 on: 13 Nov 2010, 04:04 pm »
 The amp is transformer coupled. Mapleshade insists on a 8 foot length for both supply and return. The load of the speakers is approx. 6 ohms. I can use a minimum of one meter since the monoblocks are close to the speakers. Have always used 8 ft. cables until recently. Actually the Tara Space & time cable is about 1 1/2 mtrs long.
   So what you are saying is that I may have gotten lucky with the LCR of the combo of the two different cables being used together ? Would there be a particular type of conductor or assembly that would better suit my needs as oppossed to the ribbons ?
  Thanks for all the time you took to respond. Very helpfull and educational for me.  :thumb:


charles

Johnny2Bad

Re: Speaker Cable Length
« Reply #16 on: 15 Nov 2010, 08:37 pm »
The amp is transformer coupled. Mapleshade insists on a 8 foot length for both supply and return. ...

charles
I am quite surprised to hear that; it doesn't really jive with most amplifier manufacturer's recommendations (who generally suggest a shorter cable is preferred, with the condition that both legs (+ and -) be of the same length and construction). Perhaps you should eMail Mapleshade and ask them directly whether a shorter cable would be fine, and what they recommend, and take it from there.