Was reading a bit about SPICE http://www.normankoren.com/Audio/Tubemodspice_article.html and was surprised by this throwaway line.
How can speaker cables affect an amp’s performance, poor phase margin or not? I have always been a cable skeptic but this author hardly seems like the kind of guy that spouts audio woo. Is there really something to speaker cable quality other than gauge? If you have a higher feedback amp what should you look for in speaker cables to avoid problems?
How can speaker cables affect an amp’s performance, poor phase margin or not? I have always been a cable skeptic but this author hardly seems like the kind of guy that spouts audio woo. Is there really something to speaker cable quality other than gauge? If you have a higher feedback amp what should you look for in speaker cables to avoid problems?
If an amp has poor phase margin, it can become unstable with a capacitive or inductive load.
Many speaker cables are deliberedly constructed with lots of extra capacitance or inductance because it increases the audible differences.
Such cables can make an amp with excess phase shift unstable.Note that is isn't the amount of feedback, it is the gain/phase margin that determines the sensitivity of an amp to this. Heavy gauge mains cable used as a speaker cable is the most neutral and uncritical 'speaker cable' you can get.
Jan
Many speaker cables are deliberedly constructed with lots of extra capacitance or inductance because it increases the audible differences.
Such cables can make an amp with excess phase shift unstable.Note that is isn't the amount of feedback, it is the gain/phase margin that determines the sensitivity of an amp to this. Heavy gauge mains cable used as a speaker cable is the most neutral and uncritical 'speaker cable' you can get.
Jan
Any speaker cable has equivalent lumped shunt capacitance, whether it is cheap or expensive.
The amplifier mfr should provide specs on any instability problems due to capacitive loading.
Then you determine whether the particular speaker cables of a specific length will cause problems
for that amplifier.
The amplifier mfr should provide specs on any instability problems due to capacitive loading.
Then you determine whether the particular speaker cables of a specific length will cause problems
for that amplifier.
I think that the fundamental issue is that we live right in the middle, between the microscopic and macroscopic worlds. At either end of the scale it's clear which approach to take but here in the middle the choice is less clear.
If we look at speaker cables using a basic Ohm's Law approach we end up with a different design than we would if we look at them as a wave guide. I have come to the conclusion that no one cable can do it all: the cable design that has the best treble can't really control the woofers sufficiently and by the time they can, the cables no longer have the best high end.
I have been bi-wiring with different cables for the woofers and tweeters for decades. I get the best results when the two cables are made of the same materials and fundamental design and only differ in size.
You guys who mention "mains wire" are you referring to what we in the US would call Romex? The 12 gauge solid core wire that is used in the walls? If that is what you are using and your speakers lend themselves to bi-wiring, use 18 or 20 gauge solid core bell wire for the tweeters and tell me that the combination doesn't sound (MUCH) better than either one alone. By only using each cable for what it's good at and not for what it is bad at, you really do get the best of both.
Pete
If we look at speaker cables using a basic Ohm's Law approach we end up with a different design than we would if we look at them as a wave guide. I have come to the conclusion that no one cable can do it all: the cable design that has the best treble can't really control the woofers sufficiently and by the time they can, the cables no longer have the best high end.
I have been bi-wiring with different cables for the woofers and tweeters for decades. I get the best results when the two cables are made of the same materials and fundamental design and only differ in size.
You guys who mention "mains wire" are you referring to what we in the US would call Romex? The 12 gauge solid core wire that is used in the walls? If that is what you are using and your speakers lend themselves to bi-wiring, use 18 or 20 gauge solid core bell wire for the tweeters and tell me that the combination doesn't sound (MUCH) better than either one alone. By only using each cable for what it's good at and not for what it is bad at, you really do get the best of both.
Pete
By "mains wire" we in the US would mean lamp cord, SPT, "zip cord" - two conductor flat flex cable, also available as "speaker wire" in gauges from 22 to 10 or more.
Speaker impedance deviations from the nominal value are at least an order of magnitude larger than those created by the speaker cable, so I wouldn't worry about it too much. For transmission line effects to come into play, the length would need to be a significant fraction of a wavelength - the wavelength is 13 km at 20 KHz.
Speaker impedance deviations from the nominal value are at least an order of magnitude larger than those created by the speaker cable, so I wouldn't worry about it too much. For transmission line effects to come into play, the length would need to be a significant fraction of a wavelength - the wavelength is 13 km at 20 KHz.
The potential problem is what the cable does to super audible frequencies, not what it does to audio frequencies. The amplifier feedback does not stop at 20KHz, and transmission lines may form a high-Q resonance at high frequencies. A series inductor+resistor is important to isolate the feedback from HF loading effects, and a Zobel network is a good idea.
Oh, certainly - the speaker wires will form a high Q network at 200 KHz if they're 1/4 wavelength - 530 meters long... oops, not likely. Or 2 MHz if they're 53 meters long though a tube amp won't have any gain / output at 2 MHz... A transformerless tube or solid state amp might. Practical lengths of speaker cable will have a minuscule effect compared to speaker impedance.
The gain bandwidth product of many amps exceeds 2MHz, where a quarter wave ~= 37.5m. Not common in a residential location but easy for institutions and pro sound. But I think the comment "sensitive to cables and loudspeakers." is directed to complex loudspeaker impedances including crossover inductor self-resonance. The cable is a small but real part of the situation, including it's affect as an antenna.
steveu,
A.) a vacuum tube power amplifier is one thing.
B.) a solid state Hi Fi Stereo amplifier is another thing.
C.) a solid state IC op amp is a third thing.
The specifications are Different.
1. Vacuum tube power amplifiers list bandwidth (where a low level signal or medium level signal drops by 3 dB at the low frequency and high frequency limits, versus the mid frequency signal output). "Frequency Response" specification.
A very well specified vacuum tube amplifier also specifies the -3dB power bandwidth.
Example a 30 Watt amplifier at 1 kHz, that only puts out 15 Watts at 35Hz and 25kHz. "- 3dB Power Bandwidth" specification
3. An IC op amp may have a Gain Bandwidth Product of 100MHz.
At 100MHz the gain is 1, and at 100kHz the gain is 1000; at 1kHz the gain is 100,000.
What does Gain Bandwidth Product specification have to do with a vacuum tube amplifier?
What does Gain Bandwidth Product specification have to do with a Sony, Marantz, Kenwood, or Denon solid state Hi Fi Stereo amplifier?
Please find marketing's brochures that list those specifications, and post them in your answers.
Thanks!
4. I think that IC op amplifier subjects and discussions are best transferred to another area of diyAudio.
A.) a vacuum tube power amplifier is one thing.
B.) a solid state Hi Fi Stereo amplifier is another thing.
C.) a solid state IC op amp is a third thing.
The specifications are Different.
1. Vacuum tube power amplifiers list bandwidth (where a low level signal or medium level signal drops by 3 dB at the low frequency and high frequency limits, versus the mid frequency signal output). "Frequency Response" specification.
A very well specified vacuum tube amplifier also specifies the -3dB power bandwidth.
Example a 30 Watt amplifier at 1 kHz, that only puts out 15 Watts at 35Hz and 25kHz. "- 3dB Power Bandwidth" specification
3. An IC op amp may have a Gain Bandwidth Product of 100MHz.
At 100MHz the gain is 1, and at 100kHz the gain is 1000; at 1kHz the gain is 100,000.
What does Gain Bandwidth Product specification have to do with a vacuum tube amplifier?
What does Gain Bandwidth Product specification have to do with a Sony, Marantz, Kenwood, or Denon solid state Hi Fi Stereo amplifier?
Please find marketing's brochures that list those specifications, and post them in your answers.
Thanks!
4. I think that IC op amplifier subjects and discussions are best transferred to another area of diyAudio.
Last edited:
As was mentioned, low capacitance and inductance. I've a set of cables that look like a giant version of the 300 Ohm antenna wire of ancient times, conductors spaced well apart along the run. One guy back in the days of DEC audio notes made his cables out of multiple strands of backplane wire; they looked to be 1/2" in diameter. He said they were low inductance. They looked like welding cables running to his speakers.
So - possibly - wide separation between +/- conductors, if not completely separate cable instances for the + and - runs. Lots of insulated, parallel conductors making up each cable.
I've made the cables that weave over and under an inert (plastic...) core, approximating 90 deg crossings of the +/- conductors at each weave. I've seen that style with multiple parallel conductors too.
I suppose those would be the kinds of things I'd look for, before giving them a listen. I basically stay away from fancy cables - unless I DIY them to experiment and for fun.
isaacc7,
A few high feedback amplifiers are not stable with certain loudspeakers, regardless of the cable.
A few high feedback amplifiers are stable with those same certain loudspeakers, and with most non-sense cables.
A few medium feedback amplifiers are unstable with certain loudspeakers.
Generalizations fall short of predicting what you will get.
An amplifier, a cable, and a loudspeaker is a part of a system.
Different products in the system . . . some match better, some match worse.
Without more definition of the various products in the system, we are blindfolded archers trying to hit the target.
Do you have an amplifier in mind to purchase, design and build, or that you have now? Describe it.
Do you have a loudspeaker in mind to purchase, design and build, or that you have now? Describe it.
Thanks!
A few high feedback amplifiers are not stable with certain loudspeakers, regardless of the cable.
A few high feedback amplifiers are stable with those same certain loudspeakers, and with most non-sense cables.
A few medium feedback amplifiers are unstable with certain loudspeakers.
Generalizations fall short of predicting what you will get.
An amplifier, a cable, and a loudspeaker is a part of a system.
Different products in the system . . . some match better, some match worse.
Without more definition of the various products in the system, we are blindfolded archers trying to hit the target.
Do you have an amplifier in mind to purchase, design and build, or that you have now? Describe it.
Do you have a loudspeaker in mind to purchase, design and build, or that you have now? Describe it.
Thanks!
Speaker cables are a very, very high retail margin item, so a lot of spin gets devoted to the topic. In the 1980s, 1990s ideas were floated like matching cables' transmission line impedance to load impedance (and ignoring source impedance, but there it is) which had a commercial upside of sounding different, and different is always better. Don't believe me? just ask any reviewer. The resulting large lumped-sum capacitance gave the marginally stable tweaky amps of the era hissy fits, but those that didn't self destruct at least sounded different, so mission accomplished. Sign here please.
My good friend M. has a really really nice audio system, all Spectral electronics, Avalon speakers and a pair of MIT cables made just for this combination. The cables are bigger than a garden hose, not that unusual these days, but they have a "magic box" built into them at the speaker end. All potted and sealed up, and separate wiring to the speakers' high and low frequency input terminals, but it's not hard to imagine the network inside. The box is big but not big enough for a crossover of that quality. So what would we include in such a box (which could, of course, be located inside or immediately outside the speaker box)?
All good fortune,
Chris
My good friend M. has a really really nice audio system, all Spectral electronics, Avalon speakers and a pair of MIT cables made just for this combination. The cables are bigger than a garden hose, not that unusual these days, but they have a "magic box" built into them at the speaker end. All potted and sealed up, and separate wiring to the speakers' high and low frequency input terminals, but it's not hard to imagine the network inside. The box is big but not big enough for a crossover of that quality. So what would we include in such a box (which could, of course, be located inside or immediately outside the speaker box)?
All good fortune,
Chris
Some amplifiers were notorious for this effect, often exacerbated by omission of a series inductor and zobel network on the amplifier output. Early Naim amplifiers back in the 80s needed specific cables to prevent the amp from blowing up.
NVA gear is the same. In fairness on their site they say:
"IMPORTANT - there is no protection circuitry or filtering on the output of our amplifiers (that's one of the reasons why they sound so good!), so care must be taken in use. Do not short circuit the output; doing so will fatally damage the amplifier boards. Do not use high capacitance speaker cables as these could damage the amplifier by creating a virtual short-circuit at high frequencies. Do not use bi-wiring with non-NVA cables. Do not use Litz, Goertz or woven cables. Do not twist cables together. Avoid cables with a capacitance of greater than 200pf per metre (failure to do so will invalidate our warranty) and do not use non-NVA cables over 10 metres in length"
https://nvahifi.co.uk/products/nva-s80-stereo-power-amplifier
Again in fairness to NVA their specific cables are not expensive https://nvahifi.co.uk/products/nva-ls3-loudspeaker-cable-various-lengths
NVA gear is the same. In fairness on their site they say:
"IMPORTANT - there is no protection circuitry or filtering on the output of our amplifiers (that's one of the reasons why they sound so good!), so care must be taken in use. Do not short circuit the output; doing so will fatally damage the amplifier boards. Do not use high capacitance speaker cables as these could damage the amplifier by creating a virtual short-circuit at high frequencies. Do not use bi-wiring with non-NVA cables. Do not use Litz, Goertz or woven cables. Do not twist cables together. Avoid cables with a capacitance of greater than 200pf per metre (failure to do so will invalidate our warranty) and do not use non-NVA cables over 10 metres in length"
https://nvahifi.co.uk/products/nva-s80-stereo-power-amplifier
Again in fairness to NVA their specific cables are not expensive https://nvahifi.co.uk/products/nva-ls3-loudspeaker-cable-various-lengths
Agree Chris. So there's pressure to offer something that shows an audible difference, and manipuating C and L is one way. I'm pretty sure that's the purpose of those boxes at the end of the cable. It looks impressive, it justifies the price, and it makes it sound different. Win-win for the manufacturer.
Jan
The box usually contains a battery and the circuit that is used to bias the dielectric in the cables. Audioquest, for example, uses a 72v system.
Pete
Decades ago, Audionics in Portland Oregon designed and built a solid state amplifier with specifications that were "the cat's meow", when loaded with a non reactive power resistor. Lots of negative feedback gave super performance
Then they were sold, and most of them blew up on real world loudspeakers, & had to be repaired.
I hope I got the story 100% right.
I know somebody who will probably read this and correct me if I am wrong; or will add more details to the story.
Then they were sold, and most of them blew up on real world loudspeakers, & had to be repaired.
I hope I got the story 100% right.
I know somebody who will probably read this and correct me if I am wrong; or will add more details to the story.