I am amazed that there is still so much voodoo & pseudoscience regarding speaker cable.
I still say > Buy the heaviest gauge OFC (now with good dielectric) you can afford.
If you feel the need to introduce C , L or R , you can add it with cheap components.
It also amazes me just how many people are apparently confused regarding AF vs RF.
The "Skin Effect" of heavy gauge OFC would amount to .1 dB at worst in the audio band.
( obviously I'm not talking about 50 meter cables )
I still say > Buy the heaviest gauge OFC (now with good dielectric) you can afford.
If you feel the need to introduce C , L or R , you can add it with cheap components.
It also amazes me just how many people are apparently confused regarding AF vs RF.
The "Skin Effect" of heavy gauge OFC would amount to .1 dB at worst in the audio band.
( obviously I'm not talking about 50 meter cables )
Not to detract from Jan's point, but skin effect causes a frequency-dependent resistance in cables, which can be avoided by using multi-stranded cables, each strand being sufficiently fine.
So I wouldn't use heavy gauge cable for audio frequencies.
My step sister is a nuclear. engineer, a nuclear plant operator. If skin effect is pronouinced at 60Hz. 8+ some fractons of inch iinches, has only been studied in high voltage/frequency applications, there certainly are some factors not very intuitive.
One of the most crippling handicaps of the ear is it's proximity to the copora quadrigemina, a brain element responsible fot active responses to routine, mundane srimuli in usually safely filtered from the conscsciou analyses.
In the audiophile genotype, when farhfully reproduce music begins--and this is true of a relatively narrow band of genotype, raher than react by having one's fill of the audio spectrum, move to a different location or start start binging on wallpaper paste, these clear products of evolutionalry adaptatiotation have formed into a critique of discrete PCB-level compnonents.commonly understood role as transforming recorded media, into a more convenient.
This has developed format that can record replaying a small cylinder pipe organ recording. Sympnony Orchestras worrked them same way; much smaller than an actual Orchestra, at a ==
Seeminingly involuntary, the music, the musicians' fees, even tunrntables in cars (Motorola), swamped the evolutionary window exactly the way manusctured infant formula doesn't. Men began showing a distinct disorder, which , if left unchecked, caused huge monetary expendurures on Hi Fi.
To prevemt the audio iindustry from ruling the Universe, subversives wrote farcical diatribe, and published it on news-groups. It eventually grew into the most authoritative collective of information on Hi Fi. This was evolution accellerating itself to accommodate new discoveries.
Men no longer came home to music. Th they can home to vacuum tubes, transistoroors, inductors, trans former, TUNEDsoind transfucers. Capacitors. Within a few years men were listening to discrete passives, transistors, Integrated surfaces. And, and, potentiometers!
I can't delve very far into interconnet cables. They are crafted by means thatnaooear ti ,u l
One of the most crippling handicaps of the ear is it's proximity to the copora quadrigemina, a brain element responsible fot active responses to routine, mundane srimuli in usually safely filtered from the conscsciou analyses.
In the audiophile genotype, when farhfully reproduce music begins--and this is true of a relatively narrow band of genotype, raher than react by having one's fill of the audio spectrum, move to a different location or start start binging on wallpaper paste, these clear products of evolutionalry adaptatiotation have formed into a critique of discrete PCB-level compnonents.commonly understood role as transforming recorded media, into a more convenient.
This has developed format that can record replaying a small cylinder pipe organ recording. Sympnony Orchestras worrked them same way; much smaller than an actual Orchestra, at a ==
Seeminingly involuntary, the music, the musicians' fees, even tunrntables in cars (Motorola), swamped the evolutionary window exactly the way manusctured infant formula doesn't. Men began showing a distinct disorder, which , if left unchecked, caused huge monetary expendurures on Hi Fi.
To prevemt the audio iindustry from ruling the Universe, subversives wrote farcical diatribe, and published it on news-groups. It eventually grew into the most authoritative collective of information on Hi Fi. This was evolution accellerating itself to accommodate new discoveries.
Men no longer came home to music. Th they can home to vacuum tubes, transistoroors, inductors, trans former, TUNED
I can't delve very far into interconnet cables. They are crafted by means thatnaooear ti ,u l
Last edited:
The wavelength at which an amplifier may oscillate may correspond to several MHz, if not higher. Also, a 1/4 wavelength line is already resonant. If an amplifier is oscillating continuously or bursting in and out of oscillation it can absolutely affect the sound in the audio band.
On the topic of speaker cables, star-quad has lower inductance than twisted pair, and is very reasonably priced.
Last edited:
True, continuous oscillations or bursting, if either happens, is a real problem.
If an amplifier is oscillating at several MHz, then . . .
1. Either Fix it,
Or,
2. Get another amplifier.
Simple and effective solutions.
The wavelength at the speed of light for 30MHz is 10 Meters.
A slow speaker cable velocity will have a different wavelength, 30MHz wavelength can be as long as 20 Meters in the speaker cable.
Lower frequencies, like 3MHz, have wavelengths 10 x as long (100 Meters in air, 200 Meters in speaker cable).
These are approximations, your cable may differ.
But the principles apply.
If your speaker cables are 10 Meters long (33 Feet), get a different setup where the amplifier is closer to the speakers.
Just another elegant solution.
For those who have never done accurate Vector Network Analyzer measurements on speaker cables, what you would discover is that the biggest difference is due to the speaker at the far end of the speaker cable.
A quarter wave length cable that is open at the far end, is nothing more than an out of phase reflection.
A quarter wave length cable that is connected to a speaker at the far end is whatever phase reflection it "wants to be".
A quarter wavelength is always dependent on the frequency.
These are all about Complex problems, and the impedance is usually Complex too.
Just my opinions (and puns).
Some of your opinions will differ.
It is a Mean thing to say that Karl Friedrich Gauss was just an Average man at the Peak of his career, and who was at the Center of the Gaussian Curve.
If an amplifier is oscillating at several MHz, then . . .
1. Either Fix it,
Or,
2. Get another amplifier.
Simple and effective solutions.
The wavelength at the speed of light for 30MHz is 10 Meters.
A slow speaker cable velocity will have a different wavelength, 30MHz wavelength can be as long as 20 Meters in the speaker cable.
Lower frequencies, like 3MHz, have wavelengths 10 x as long (100 Meters in air, 200 Meters in speaker cable).
These are approximations, your cable may differ.
But the principles apply.
If your speaker cables are 10 Meters long (33 Feet), get a different setup where the amplifier is closer to the speakers.
Just another elegant solution.
For those who have never done accurate Vector Network Analyzer measurements on speaker cables, what you would discover is that the biggest difference is due to the speaker at the far end of the speaker cable.
A quarter wave length cable that is open at the far end, is nothing more than an out of phase reflection.
A quarter wave length cable that is connected to a speaker at the far end is whatever phase reflection it "wants to be".
A quarter wavelength is always dependent on the frequency.
These are all about Complex problems, and the impedance is usually Complex too.
Just my opinions (and puns).
Some of your opinions will differ.
It is a Mean thing to say that Karl Friedrich Gauss was just an Average man at the Peak of his career, and who was at the Center of the Gaussian Curve.
Last edited:
With much respect to Jan, I find coaxial speaker cables sound better in RF-polluted environments such as city centres. I like these: https://shop.sommercable.com/en/Cab...utsprecherkabel-SC-Magellan-SPK-440-0201.html
for pretty well everything above 160 Hz (my crossover point of choice).
For the end of the spectrum, I suggest something big, bold:
https://shop.sommercable.com/en/Cab...abel-Elephant-Robust-SPM440-490-0051-440.html
since black backgrounds matter less to subwoofers than low resistance.
Note these also come in fire-retardant types, denoted with an F at the end of the model number. The jacket is less flexible, not quite to my taste.
for pretty well everything above 160 Hz (my crossover point of choice).
For the end of the spectrum, I suggest something big, bold:
https://shop.sommercable.com/en/Cab...abel-Elephant-Robust-SPM440-490-0051-440.html
since black backgrounds matter less to subwoofers than low resistance.
Note these also come in fire-retardant types, denoted with an F at the end of the model number. The jacket is less flexible, not quite to my taste.
1/4 wavelength is long enough for line to be fully resonant. It doesn't have to be one wavelength long. Also, loading one end a line with a reactance such as output stage capacitance can substantially make the required length of cable shorter and still achieve full resonance. Is called, end-loading, IIRC.
Anyway, for transmission line behavior to be negligible, the standard rule of thumb is that the cable needs to be less than 1/10 of 1/4 wavelength long. This stuff is all very basic RF electronics.
Regarding amplifier stability, Cyril Batman wrote up an pretty good explanation long ago. For anyone interested, please see attached.
Anyway, for transmission line behavior to be negligible, the standard rule of thumb is that the cable needs to be less than 1/10 of 1/4 wavelength long. This stuff is all very basic RF electronics.
Regarding amplifier stability, Cyril Batman wrote up an pretty good explanation long ago. For anyone interested, please see attached.
With RF and transmission lines:
The amplifier output impedance = the transmission line impedance = the load impedance.
Anything else has a Reflection, and transfers less power.
The same holds true for Waveguides, or there is a transition adapter at one or both ends to match impedances to amplifier, antenna, etc.
500 GHz is the highest frequency I ever measured. That is still far short of near infra red.
I see very few amplifiers that are 8 Ohms output impedance (damping factor 1.0, Zo = 8 Ohms).
I see very few speaker cables that have 8 Ohm impedance from 20Hz to 20kHz.
I see very few speakers that are 8 Ohms impedance [resistive] from 20Hz to 20kHz.
All this is required, or we are not talking about proper and efficient transmission line theory, as always applies for RF systems.
Any speaker cable, no matter how long, or how short, is reactive at some frequency.
if you are interested, look up transmission line theory, including Rho, reflection co-efficient.
Then continue the discussion on this thread.
My $0.02
Your money may vary.
I think that 1/10 of a 1/4 wavelength (0.025 wavelength) at 20kHz is much longer than almost everyones speaker cables.
My speakers do not work very much over 20kHz, and my one working ear stops at about 13 - 15kHz.
Does your audiologist test your ears to 30kHz?
By the way, 1/4 wavelength occurs at 'one, and only one' frequency.
The frequency at which the line is 1/4 wavelength is dependent on 2 things:
1. The velocity factor of the line (only free space and waveguide are at or near the speed of light).
2. The length of the line.
All of the above is basic electronics for some, and is very complex electronics for others
The "Simple" article is 32 Pages
I still believe:
"If an amplifier is oscillating at several MHz, then . . .
1. Either Fix it,
Or,
2. Get another amplifier.
Simple and effective solutions."
. . .
Do not fix the speaker cable; do not fix the speaker. Fix the amplifier.
The amplifier output impedance = the transmission line impedance = the load impedance.
Anything else has a Reflection, and transfers less power.
The same holds true for Waveguides, or there is a transition adapter at one or both ends to match impedances to amplifier, antenna, etc.
500 GHz is the highest frequency I ever measured. That is still far short of near infra red.
I see very few amplifiers that are 8 Ohms output impedance (damping factor 1.0, Zo = 8 Ohms).
I see very few speaker cables that have 8 Ohm impedance from 20Hz to 20kHz.
I see very few speakers that are 8 Ohms impedance [resistive] from 20Hz to 20kHz.
All this is required, or we are not talking about proper and efficient transmission line theory, as always applies for RF systems.
Any speaker cable, no matter how long, or how short, is reactive at some frequency.
if you are interested, look up transmission line theory, including Rho, reflection co-efficient.
Then continue the discussion on this thread.
My $0.02
Your money may vary.
I think that 1/10 of a 1/4 wavelength (0.025 wavelength) at 20kHz is much longer than almost everyones speaker cables.
My speakers do not work very much over 20kHz, and my one working ear stops at about 13 - 15kHz.
Does your audiologist test your ears to 30kHz?
By the way, 1/4 wavelength occurs at 'one, and only one' frequency.
The frequency at which the line is 1/4 wavelength is dependent on 2 things:
1. The velocity factor of the line (only free space and waveguide are at or near the speed of light).
2. The length of the line.
All of the above is basic electronics for some, and is very complex electronics for others
The "Simple" article is 32 Pages
I still believe:
"If an amplifier is oscillating at several MHz, then . . .
1. Either Fix it,
Or,
2. Get another amplifier.
Simple and effective solutions."
. . .
Do not fix the speaker cable; do not fix the speaker. Fix the amplifier.
Last edited:
Are you aware of the frequency where this becomes into play?
Skin effect is a non-issue in speaker cables. It does exist (of course) but invisible for audio.
To @DSP_Geek's point, yes speaker cables can and do pick up RF pollution.
A good amplifier has provisions to avoid ingress into the amp and the feedback circuit.
As was mentioned before, get a competently designed amp instead of trying to fix it with speaker cables with no guarantee that it actually helps.
Jan
500 GHz is the highest frequency I ever measured.
.........for myself around 50GHz. that puts both of us into the wonderful copper-smith & plumbing world. In 1972 as a Uni project I made a receiver/ dish and picked up Skylab signals on now lowish 2.28GHz.
---------------
So, as I see it, (only my opinion) to sum up this speaker cable business, I just get thick copper cross section and be content. Oxygen free is a sales fad. Ohms for Ohms.
I´ve known close folks who believe in it, only to discover when it was too late that their SS amp was already marginally unstable trying to deal with the complex phase issues of their Loudspeakers with complex 3 way 24dB/oct LS crossovers. There was a period when mobile phones and nasty phased-controlled lighting could easily upset SS audio systems which is a good indication of troubles ahead.
I´ve worked with music studio equipment where "rush to finish" wiring repairs often resulted in dodgy connections and yet posed no change in audio quality. Contentus esto
BenchBaron
.........for myself around 50GHz. that puts both of us into the wonderful copper-smith & plumbing world. In 1972 as a Uni project I made a receiver/ dish and picked up Skylab signals on now lowish 2.28GHz.
---------------
So, as I see it, (only my opinion) to sum up this speaker cable business, I just get thick copper cross section and be content. Oxygen free is a sales fad. Ohms for Ohms.
I´ve known close folks who believe in it, only to discover when it was too late that their SS amp was already marginally unstable trying to deal with the complex phase issues of their Loudspeakers with complex 3 way 24dB/oct LS crossovers. There was a period when mobile phones and nasty phased-controlled lighting could easily upset SS audio systems which is a good indication of troubles ahead.
I´ve worked with music studio equipment where "rush to finish" wiring repairs often resulted in dodgy connections and yet posed no change in audio quality. Contentus esto
BenchBaron
Yes, of course. You're not the only one with a background in RF and microwaves. Mine was mostly with various particle accelerators used for radiation oncology and radio-neurosurgery. Non-uniform TL and waveguide structures in standing wave and traveling wave linacs, and D-structures in AVF cyclotrons.
Since this is a matter of opinion and getting closer to bickering, I will leave off here with one final observation: there is more to speaker cable than has been discussed here. There are other threads in the forum with more information on the subject. There is also some evidence that Zo can matter for audio even with reasonably modest length cables and stable amplifiers. IIRC @Hans Polak did some work on it in another thread, "zip cord for speaker test."
Last edited:
If you have a couple of hours to burn, you can build this speaker cable.
Two cross-linked coaxial cables
https://www.tnt-audio.com/clinica/ubyte2e.html
Good quality 75 Ohm satellite dish cable (*). Solid copper center, thick copper braid+copper foil. Foam or hollow dielectric.
Very low inductance, almost no capacitance.
Attached measurements are for 2.5m of this dual-cable loop.
Immune from antenna (RX/TX) effect.
George
(*) In my area I can get CAVEL Sat 752 cable for close to 1Euro/m
Two cross-linked coaxial cables
https://www.tnt-audio.com/clinica/ubyte2e.html
Good quality 75 Ohm satellite dish cable (*). Solid copper center, thick copper braid+copper foil. Foam or hollow dielectric.
Very low inductance, almost no capacitance.
Attached measurements are for 2.5m of this dual-cable loop.
Immune from antenna (RX/TX) effect.
George
(*) In my area I can get CAVEL Sat 752 cable for close to 1Euro/m
Attachments
Last edited:
I tried cross-linked coax in quad as suggested by Jon Risch in a now-departed post. It was okay but not a screaming improvement over the status quo ante. Simply running the coax in parallel (cores together, shields together) sounded better to me.
To Jan: a Quad 405 was the amplifier for the system under test, decent piece of kit. Or one could use a class-D amplifier, where the output lowpass filter has a sizeable film capacitor often bypassed by a C0G ceramic. That'll show the RF where to go!
To Jan: a Quad 405 was the amplifier for the system under test, decent piece of kit. Or one could use a class-D amplifier, where the output lowpass filter has a sizeable film capacitor often bypassed by a C0G ceramic. That'll show the RF where to go!
Reverse Transfer Impedance of Coax.
Common Mode Rejection of Common Mode EMI
Differential Mode Rejection of Differential Mode EMI
And one of the worst Oxymorons ever, EMC, Electro Magnetic Compatibility.
In many situations, there is No Compatibility.
Passing EMI regulations hardly ever insures Compatibility.
I love reading Marketing's statement that their super low distortion amplifier is "Perfectly Stable Under All Conditions".
I have this bridge in Brooklyn I want to sell you.
Common Mode Rejection of Common Mode EMI
Differential Mode Rejection of Differential Mode EMI
And one of the worst Oxymorons ever, EMC, Electro Magnetic Compatibility.
In many situations, there is No Compatibility.
Passing EMI regulations hardly ever insures Compatibility.
I love reading Marketing's statement that their super low distortion amplifier is "Perfectly Stable Under All Conditions".
I have this bridge in Brooklyn I want to sell you.
Here are the loudspeaker cables that I use - pretty cheap and quite standard...
From the cheapest to the costiest :
They all share the same parameter : 2.5mm2 section, and all are fine for me. For the rest, I must confess that I'm not very obsessed. That said, the QED have a problem of stiffness, making cracks in the sheath - that's why I discard this model now.
My possible preference would be the Van Den Hull Clearwater and the Real Cable BM250T, just because of the spacing of the conductors, which reduces inductance and capacitance of the cable, nonetheless already very small, by the way, even with the other cable models above.
T
From the cheapest to the costiest :
They all share the same parameter : 2.5mm2 section, and all are fine for me. For the rest, I must confess that I'm not very obsessed. That said, the QED have a problem of stiffness, making cracks in the sheath - that's why I discard this model now.
My possible preference would be the Van Den Hull Clearwater and the Real Cable BM250T, just because of the spacing of the conductors, which reduces inductance and capacitance of the cable, nonetheless already very small, by the way, even with the other cable models above.
T
I scrolled all the posts of this thread, not a single capacitance per meter number announced. Ordinary cable 9mm wide 4.5mm thick, 2mm diam. copper measures 200pf/m.
Flat cable 15mm wide 3mm thick with 4.5mm wide 2mm thick copper measures 330pf/m. To note that if the same cross section is used for round standard cable the capacitance would be much higher than the flat version.
These flat cables on the other hand don't cancel the magnetic field of each other, results higher inductance.
If you have installed Thiel network, you need a zobel after the Thiel if flat wire is used.
10ft cable hardly reaches 600pf with ordinary cable, even without Thiel, you don't need much phase margin.
Tube amps, once upon a time we're designed by gain margin of 6db. The designer will increase the feedback ratio until oscillate than decrease it by 6db.
Flat cable 15mm wide 3mm thick with 4.5mm wide 2mm thick copper measures 330pf/m. To note that if the same cross section is used for round standard cable the capacitance would be much higher than the flat version.
These flat cables on the other hand don't cancel the magnetic field of each other, results higher inductance.
If you have installed Thiel network, you need a zobel after the Thiel if flat wire is used.
10ft cable hardly reaches 600pf with ordinary cable, even without Thiel, you don't need much phase margin.
Tube amps, once upon a time we're designed by gain margin of 6db. The designer will increase the feedback ratio until oscillate than decrease it by 6db.
Last edited:
DSP_Geek,
You said:
"Shoot, almost any op-amp would be stable driving that load."
Thanks for reminding me of the LM301 Op Amp.
In order to drive a capacitive load . . .
The Op Amp output pin was connected to a 100 Ohm resistor, and through the 100 Ohm resistor to the capacitive load.
There were 2 negative feedback loops:
The AC feedback loop came from the Op Amp output pin (the near end of the 100 Ohm resistor).
The DC feedback loop came from the far end of the 100 Ohm resistor.
If a similar thing were done with tube amplifiers, it might be something like this . . .
Connect the output transformer secondary to a 1 Ohm resistor, and the other end of the 1 Ohm resistor to the output terminals on the back of the amplifier.
The global feedback would come from the output transformer secondary.
There would be no DC feedback loop (output transformers do not pass DC to the output).
And, the damping factor would be reduced because of the 1 Ohm series resistor (or from a lower series resistor value, such as 0.25 Ohm).
You said:
"Shoot, almost any op-amp would be stable driving that load."
Thanks for reminding me of the LM301 Op Amp.
In order to drive a capacitive load . . .
The Op Amp output pin was connected to a 100 Ohm resistor, and through the 100 Ohm resistor to the capacitive load.
There were 2 negative feedback loops:
The AC feedback loop came from the Op Amp output pin (the near end of the 100 Ohm resistor).
The DC feedback loop came from the far end of the 100 Ohm resistor.
If a similar thing were done with tube amplifiers, it might be something like this . . .
Connect the output transformer secondary to a 1 Ohm resistor, and the other end of the 1 Ohm resistor to the output terminals on the back of the amplifier.
The global feedback would come from the output transformer secondary.
There would be no DC feedback loop (output transformers do not pass DC to the output).
And, the damping factor would be reduced because of the 1 Ohm series resistor (or from a lower series resistor value, such as 0.25 Ohm).