He did predict he's going to get called out on it.
Wouldn't this be very easily tested with very long wires, source and maybe a led? surely if the led lights up faster than the speed of light would take through the cables, then something is clearly happening between source and led, across the space between them.
Last edited:
So for a 2MHz signal the phase shift would be let's say 3.6º for one meter offset? That could be measured, same wire length just have the wire to the closest source snaked.
Hi johnmath,
Coaxial cables are a capacitance by definition, and that can often make amplifiers very unstable. This ought to keep you up at night. I have repaired more than a couple blown amplifiers over the years due to the use of coaxial cables used to connect speakers. Guess what? This voids warranties on amplifiers.
By far and away the best cables for speaker connections are a fine strand copper in a "zip cord" configuration. Special cables with distance between the conductors will reduce capacitance but this is not critical. Think of the old 300 ohm TV antenna wire with large conductors.
I have tested just about every design of speaker cable over the years, especially in the mid to late 80's when there was an explosion of interest in the subject. I still get asked to test "new designs" (they aren't) and the results are generally typical of other like designs tested before. The cable design I mentioned earlier, a zip cord type cable, has always turned in the best results.
I have no horse in that race, just years of experience. Telling it like it is.
-Chris
A couple of points should be made here. Coupling between cables is incredibly weak, we have to use wound coils to couple energy. At speaker impedances you are looking at things like output transformers in tube amps or McIntosh type amplifiers. I think you can sleep comfortably at night if you ignore this aspect.
Coaxial cables are a capacitance by definition, and that can often make amplifiers very unstable. This ought to keep you up at night. I have repaired more than a couple blown amplifiers over the years due to the use of coaxial cables used to connect speakers. Guess what? This voids warranties on amplifiers.
By far and away the best cables for speaker connections are a fine strand copper in a "zip cord" configuration. Special cables with distance between the conductors will reduce capacitance but this is not critical. Think of the old 300 ohm TV antenna wire with large conductors.
I have tested just about every design of speaker cable over the years, especially in the mid to late 80's when there was an explosion of interest in the subject. I still get asked to test "new designs" (they aren't) and the results are generally typical of other like designs tested before. The cable design I mentioned earlier, a zip cord type cable, has always turned in the best results.
I have no horse in that race, just years of experience. Telling it like it is.
-Chris
Ah, but are you a trained listener?
That's the "qualifier" someone tries to use as a boogeyman.
LOL!
The audio industry has been my life, so I have listened more than most while being paid for it.
So to answer your question, obviously not!
The audio industry has been my life, so I have listened more than most while being paid for it.
So to answer your question, obviously not!
That depends on the application. I know of an installation of electronically steered loudspeaker arrays where the directionality was completely screwed up by cross-coupling between incorrectly paired conductors in the multi-core speaker cables.
The coaxial speaker cables I am aware of have capacitance of ~160pF/m, compared with Canare's range of professional quad-core speaker cables of 147pf/m and Belden 14AWG jacketed 2-core speaker cable of 118pF/m.
That depends on which performance parameter is being deemed ‘best’? All designs play off various compromises between different performance criteria and price. Try to get 13AWG (2.5mm2 x 2) speaker cable into a DIN plug as some high performance consumer audio has traditionally used; you can do it with professional Eurocable 02N25C coaxial speaker cable which is <6mm diameter.
An amplifier with the typical ~10µH inductor and Zobel will normally be stable into some capacitance, after all a lot of loudspeaker loads are capacitive. On the other hand, some amplifier designs e.g. older Naim amplifiers, omit the output inductor and require special spaced apart cables to remain stable, because the cable performs the role of the output inductor which the designer left out.
I have also performed hundreds of loudspeaker cable tests over four decades, including many in conjunction with other respected AES engineers and many double blind, which I have described in other threads on this and other sites. I have made somewhat different observations to your emphatically expressed views. I am retired and have nothing to gain except to help people get value by spending money wisely.
Last edited:
It's audio cable so audible parameter. Why ask, unless you have too much free time on your hand.
Then what do you recommend for speaker cables and why?
I sincerely doubt that!
Most normal cables have lower capacitance, I can check that again of course. Coax cables have much, much higher capacitance than zip cord style as a rule. A special low capacitance coax cable would be huge and stiff. I just don't see any reasonable installation using something like that.
Totally open ended statement. Fact: I have repaired more than a few amplifiers blown up by capacitive speaker cables that are normal, well designed amplifiers. There are plenty that do not have wide phase margins that are really sensitive.
Since when ? Most are inductive.
I'll go on record to say they were not good designs, and the Nate was pure garbage. Very bad example to use.
... and we also ran electrical tests and monitored the amps for instability too.
After all the smoke cleared, the best wire was always the zip cord style with fine copper strands. The gage depended on distance and power levels, you don't always need 16 ga or heavier. 20 ga was too light.
I have nothing to lose by recommending a cost effective wire. No friends to avoid embarrassing or company associations to worry about. But the truth remains, expensive wire at best buys you nothing. At worst, may have characteristics that can make an amplifier unstable. A coaxial cable would be in a special use situation and would probably require a series network to isolate the high capacitance from the amplifier. I can only think of a high RF field. In that situation I used filters at each end and standard wire. I had two radio stations within 1 KM transmitting around 1 MHz (just over) at 50 KW each. CD players often didn't work properly in that area. The application used tons of exposed wire running a distributed paging speaker system over a large campus. Government job. That and a few normal stereo systems in the area as well.
mods, I really was curious about the statements in the video, and wanted to see discussion on that subject. yet somehow I got dragged into reading about speaker cables. great job
Me either. You won't find a post of mine where I suggest anything else.
The main cost of a cable is the copper content. In fact when I was the Technical Manager - Sound of a performing arts center from 1984 for many years, we bought cable by net weight of copper, irrespective of it's configuration. That was a government-wide negotiated contract price for the entire state, called a State Supply Contract. During that period I started doing exhaustive tests of cable configurations for loudspeakers. The biggest installation I oversaw included twenty kilometres of speaker cable in one auditorium. You bet I wanted it to be cost effective!
The installation in question was a house of parliament. The objective in this kind of system is to put every listener in the chamber an effective acoustic distance of a metre or so to the orator. In other words, every listener should be able to hear what any person is saying in the chamber as though that person is talking directly to the listener from a close distance.
To achieve this objective in a voluminous auditorium and keep an adequate gain margin (i.e. freedom of ringing or feedback) requires highly directional loudspeakers. These are built of large numbers of drivers, where each individual driver is fed with a processed signal that varies the timing/phase relationships between the drivers to create the exact desired polar response for the array, with a flat frequency response at the target seats and a sharp cut-off outside of that zone.
Because these large arrays are flown above the listeners, and for serviceability, onboard amplification is eschewed for remote racks of amplifiers. An single array may need 18 channels of amplification or more, which means at least 36 individual wires from the amplifiers to the array.
These individual wires if not carefully located with respect to which active circuit they are in, can form air coupled transformers inside the speaker loom, which in turn introduces crosstalk between the drivers and causes the array not to create the desired coverage pattern.
In the case of one installation where the wiring contractor did not follow the system diagram, all of the (+) wires were in one loom and all the (-) wires were in a seperate loom, which because of the air spacing between the two looms completed stuffed the performance of each array (there are several such arrays for each installation). This really had the system designers scratching their heads as to why they could not get the system to meet its performance specification for commissioning, until they worked out the problem - air coupled transformers between the different circuits.
Here's an interesting response video to the initial one that started this thread, and it includes a simulation:
Response To Veritasium - Electricity Propagation Time Problem - YouTube
Response To Veritasium - Electricity Propagation Time Problem - YouTube
^^^
That video makes more sense. Obviously the wire formed a transmission line and sort of an inefficient LF loop antenna at the same time. Since the wire was sitting on the ground, probably not very good for launching much of an LF wave into space. Anyway, since there was a transmission line of sorts, there was both some capacitive and some inductive coupling between the wires. As an interesting twist, suppose the wires were one light year long but open at the ends. If the lightbulb was sensitive enough to light when the ends of the wire formed a closed loop, would the bulb still light up if the ends of the wire were cut open (a) before the switch was turned on?, (b) 3-months after the switch was turned on?
That video makes more sense. Obviously the wire formed a transmission line and sort of an inefficient LF loop antenna at the same time. Since the wire was sitting on the ground, probably not very good for launching much of an LF wave into space. Anyway, since there was a transmission line of sorts, there was both some capacitive and some inductive coupling between the wires. As an interesting twist, suppose the wires were one light year long but open at the ends. If the lightbulb was sensitive enough to light when the ends of the wire formed a closed loop, would the bulb still light up if the ends of the wire were cut open (a) before the switch was turned on?, (b) 3-months after the switch was turned on?
Last edited:
If you had signal amplification at the speakers, I can see that occurring. Cable shielding then becomes a determining factor. If you are driving speakers from remote amplifiers you couldn't couple enough energy to create a problem given the impedance of a speaker. Not unless you had serious issues with coupling (current dependent).
Why wouldn't you drive the array using a 70 volt line? Energy loss is greatly reduced and you would only require a single cable to drive the array. I can't see the need for time correction difference between the drivers. Not unless you were emitting sound in all directions which would place some arrays facing each other. That is certainly not good practice. Normally you start at one end facing all arrays or speakers in the same general direction. At each location you would delay the sound so that it would be in acoustic phase with the first location, but that delay is the same for all speakers in each location. that is the normally accepted way to cover large areas and maintains maximal intelligibility across the area. "Flying" the areas is the only reasonable way to do this as well, so that makes perfect sense.
Just trying to understand what you were doing. Driving an 8R speaker (or thereabouts) makes zero sense at all. But a low impedance speaker is insensitive to coupling due to the low impedance, then the distances involved also reduce the sensitivity to coupled energy between speaker feeds.
-Chris
Why wouldn't you drive the array using a 70 volt line? Energy loss is greatly reduced and you would only require a single cable to drive the array. I can't see the need for time correction difference between the drivers. Not unless you were emitting sound in all directions which would place some arrays facing each other. That is certainly not good practice. Normally you start at one end facing all arrays or speakers in the same general direction. At each location you would delay the sound so that it would be in acoustic phase with the first location, but that delay is the same for all speakers in each location. that is the normally accepted way to cover large areas and maintains maximal intelligibility across the area. "Flying" the areas is the only reasonable way to do this as well, so that makes perfect sense.
Just trying to understand what you were doing. Driving an 8R speaker (or thereabouts) makes zero sense at all. But a low impedance speaker is insensitive to coupling due to the low impedance, then the distances involved also reduce the sensitivity to coupled energy between speaker feeds.
-Chris
For a DC source I don't think it would turn on at all before flipping the switch. But for an AC one it could, no? (edit: he's showing a DC source in the video though)
The whole idea in the video is ambiguous in the absence of concrete values. In the real world with all the imperfections there's all kinds of interactions/losses/couplings etc
The whole idea in the video is ambiguous in the absence of concrete values. In the real world with all the imperfections there's all kinds of interactions/losses/couplings etc
There would be no difference between AC and DC. Barring any other effects, energy introduced in the cable would reach the other end and for the period of time the switch was closed.
I meant AC source would have some emissions from the terminals. I was talking about real world parasitic effects.