Why use Monster as the reference cable?

[Ivor the Engineer]

One of the things that knackers the performance of home audio systems is your super-duper amplifier(s) with a flat frequency response from 1Hz to 200kHz and 0.0001% distortion, is driving speakers with a frequency response specified to +-1dB or worse and a distortion of 1% or higher. Add in room resonances and reflections and the cable in-between the amplifier and speaker is of virtually no consequence. 13A Mains flex is just as good as any "Audio-Grade" cable, and a LOT less expensive.

If the bi-wired speakers were also bi-amped, then it (in theory) could make a big difference to the performance, since the passive crossover networks in the speakers can be removed and the range of frequencies sent to each speaker split by low-level (i.e. more accurate) electronic circuitry before the power-amplifiers. It would also permit the low-frequency driver to be driven closed-loop, using motional feedback to correct for non-linearities.

 

[Dan O.]

I can find nothing in the article you link to support your claim. Perhaps you'd like to quote the relevant text? Remember, we're talking solely about a network inserted into the path from the amplifier to the speaker here (not networks in the amplifier or in the speaker crossovers). You're saying that this is necessary to avoid amplifier "melt down." You further claim that the linked site supports this (bizarre) contention. Care to show where?

Did you read (for instance) the "introduction":

Even though most power amplifiers are limited to at most a few hundred kHz or so, there can still be some energy at higher frequencies - typically noise. What often happens is that an amp can be quite stable with a capacitive load and no signal, but as soon as it is driven it "excites" the whole system, and it then bursts into sustained oscillation.

It is almost impossible for any amp to reproduce high levels at extremely high frequencies, and they are not present in the source material. This has never stopped an amp from oscillating though, usually at a frequency high enough to cause simultaneous conduction of the power transistors, since they cannot switch off quickly enough, and both will be turned on at the same time.

This simultaneous conduction is what causes damage, since the output devices heat up very quickly and may go into second breakdown - if that happens, then it's all over - the amplifier will fail with blown output devices. Anyone who has had an amplifier on a test bench and supplied it with an input signal at 100kHz or more will have seen this - even with no load, the amp will draw a lot of current even at low output levels. If maintained for any period of time, the amp will fail.

 

[Yoink]

Did you read (for instance) the "introduction":

And where, in that long quotation, is anything at all about networks in speaker cables?

 

[Ivor the Engineer]

Sounds like you guys are talking about Zobel networks, which usually consist of a parallel LR in series with the amplifier output and a C to ground, used to isolate the load (including the speaker cable) from the amplifier at high frequencies. Often these are simplified to a series RC network to ground from the amplifier output. The aim is to make the load look as much like a pure resistance as possible from the amplifiers point of view.

 

[Yoink]

Sounds like you guys are talking about Zobel networks, which usually consist of a parallel LR in series with the amplifier output and a C to ground, used to isolate the load (including the speaker cable) from the amplifier at high frequencies. Often these are simplified to a series RC network to ground from the amplifier output. The aim is to make the load look as much like a pure resistance as possible from the amplifiers point of view.

Right--and those Zobel networks are either in the amplifier or in the speakers, right? Dan O.'s claim was that if you don't have an additional Zobel network in the speaker cables themselves you risk having your amplifier melting.

 

[bjb]

Section 3.1 discusses 'special cable', which has a Zobel network at one end of the cable. The Zobel network is supposed to make the speaker look like a pure resistance at all frequencies.

I use them in custom crossover designs, but I design them for the particular woofer and tweeter used in the speaker. The idea is to put the Zobel network across the woofer and tweeter so they appear 'resistive' to the crossover. This makes the crossover work a little better because it is driving resistive loads instead of reactive.

That webpage says that 10 ohms and 100nF is the 'standard' network for all applications, which does not make sense to me at all. The resistor and capacitor must be chosen to balance out the reactance of the speaker, and since all speaker designs have different reactances, no single Zobel network 'fits all'. I base my calculations for the Zobel network parameters upon impedance vs. frequency sweeps but appears as though audiophiles simply pick their numbers out of the air.

What's interesting is that the simulated speaker system design does not include Zobel networks for the woofer and tweeter! If the speaker system was properly designed, the Zobel networks would be in there already and there would be no need for 'special' cable. One major criticism of the article is that a high-frequency model is not used for the speaker. The speaker inductances should have a capacitance across them to simulate winding-to-winding capacitance. This turns the inductors into short circuits at high frequencies instead of open circuits as would occur in the simulation as written. In other words, the whole simulation is wrong because the speaker simulation is not valid.

I wonder if the Transparent Opus cable isn't simply a 'fix' for speaker designs that don't have the Zobel networks already built in? Another effect I've noticed with Zobel networks is that they suck away some high frequencies. They have to because they are a resistor and capacitor in series so they affect the highs more than the lows. I've even experimented with a Zobel network for my guitar amp to tame the highs. This may account for the Transparent Opus cables 'helping' the bass frequencies. They may simply be throwing away some highs to make the lows stand out. In effect, these cables are acting as tone controls, which is something audiophiles tend to avoid, at least consciously.

 

[Ivor the Engineer]

Right--and those Zobel networks are either in the amplifier or in the speakers, right? Dan O.'s claim was that if you don't have an additional Zobel network in the speaker cables themselves you risk having your amplifier melting.

You are correct. If you need to add extra parts outside the speaker [enclosure] or amplifier to ensure stability, one or both have been incorrectly designed or have a fault.

 

[bjb]

Right--and those Zobel networks are either in the amplifier or in the speakers, right? Dan O.'s claim was that if you don't have an additional Zobel network in the speaker cables themselves you risk having your amplifier melting.

Amplifiers don't have built-in Zobel networks because they need to be designed for the speaker. Some speaker designs don't include them, either. The simulated speaker design from that webpage is one example.

This caught my attention:

"This simultaneous conduction is what causes damage, since the output devices heat up very quickly and may go into second breakdown"

The author sees that the amplifier draws more current at high frequencies so he assumes it is due to 'simultaneous conduction' of the output transistors. In the power electronics world, this is called 'shoot-through', and it is a well-documented failure mode. However, there may be another explanation for the increased current draw.

A power transistor will be less efficient at high frequencies than at low. The transistors generate heat every time they switch from off to on and back again, so the faster this occurs, the more heat is generated in a given amount of time. Since more heat is generated at high frequencies, the efficiency drops. The same thing happens in a computer that is overclocked. I think the amplifier is drawing high current at high frequency because this is normal behavior!

I'm afraid the author is observing perfectly normal circuit behavior and has proposed a rather elaborate mechanism to explain it. I notice he does not say that he tried a Zobel cable and it kept the amplifier from drawing high current at high frequency. An observation like that would convince me there was something to this amplifier instability issue, but he never gets around to actually verifying that the Zobel network stabilizes the amp.

I'm not even sure this guy knows how to identify an oscillating amplifier. In real life, a technician can test for oscillation by looking at the amplifier output with an oscilloscope and applying a squarewave to the input. Ringing or fuzziness on the output will indicate oscillation within the amplifier. Once again, the stability added by a Zobel network is not verified by a test like this.

 

[Yoink]

Amplifiers don't have built-in Zoble networks because they need to be designed for the speaker. Some speaker designs don't include them, either. The simulated speaker design from that webpage is one example.

This caught my attention:

"This simultaneous conduction is what causes damage, since the output devices heat up very quickly and may go into second breakdown"

The author sees that the amplifier draws more current at high frequencies so he assumes it is due to 'simultaneous conduction' of the output transistors. In the power electronics world, this is called 'shoot-through', and it is a well-documented failure mode. However, there may be another explanation for the increased current draw.

A power transistor will be less efficient at high frequencies than at low. The transistors generate heat every time they switch from off to on and back again, so the faster this occurs, the more heat is generated in a given amount of time. Since more heat is generated at high frequencies, the efficiency drops. The same thing happens in a computer that is overclocked.

I'm afraid the author is observing perfectly normal circuit behavior and has proposed a rather elaborate mechanism to explain it. I notice he does not say that he tried a Zobel cable and it kept the amplifier from drawing high current at high frequency. An observation like that would convince me there was something to this amplifier instability issue, but he never gets around to actually verifying that the Zobel network stabilizes the amp.

Thanks for the clarification. I have some further questions if you wouldn't mind elaborating:

1/ Does it make any sense to put Zobel networks into the speaker cables? That is, if the Zobel network has to be designed for a specific speaker, would a "generic" Zobel network do any good at all?

2/ The implication of Dan O.'s article is that you might want a Zobel network on the speaker cables if the speaker cables are of particularly high capacitance. Does that make sense to you?

3/ What would be the effect, if any, on the sound produced by your speakers of a Zobel network in the speaker cables?

 

[Ivor the Engineer]

Amplifiers don't have built-in Zoble networks because they need to be designed for the speaker.

<snip>

Yes, they do.

Check out figures 3 & 4 on the LM4765 datasheet:

http://cache.national.com/ds/LM/LM4765.pdf

They are not always required, but most amplifiers have some kind of network to isolate the load from the amplifier output and/or provide a load for the amplifier at high frequency.

 

[bjb]

Thanks for the clarification. I have some further questions if you wouldn't mind elaborating:

1/ Does it make any sense to put Zobel networks into the speaker cables? That is, if the Zobel network has to be designed for a specific speaker, would a "generic" Zobel network do any good at all?

It sort of makes sense to put a non-optimal Zobel network into a speaker cable. The idea it to present a more resistive load to the amplifier to help stabilize it and the generic Zoble network should help. Not enough compensation would still have some benefit and too much would unnecessarily suppress the highs. Of course, I wouldn't be surprised if someone came up with a case where this did not apply but this is just a guess based upon my own experience.

2/ The implication of Dan O.'s article is that you might want a Zobel network on the speaker cables if the speaker cables are of particularly high capacitance. Does that make sense to you?

Maybe, but it depends upon the amplifier. Some are more stable than others when driving capacitive loads. I'm fortunate to have a copy of "Audio Cyclopedia" from the early 70's and there's some pretty good information concerning capacitive loads. Electrostatic speakers were the big thing back then and because of their high capacitance, it was difficult to design an amp that could drive them without becoming unstable. However, those loads were on the order of 10uF, much more capacitance than seen in a speaker cable.

Another problem with this question is that someone can always design a speaker cable with arbitrarily high capacitance. In fact, some audiophiles seem bent on accomplishing this! Trying to compensate for someone else's stupid design is like trying to win an argument with the Mad Hatter. The rules of reason are of no help because the situation can always be made worse by avoiding reason! There might be a 'best' way to deal with excessive capacitance, but but my answer is to avoid capacitance in the first place. A normal piece of speaker wire is good enough to do the job without introducing unnecessary problems like high capacitance.

3/ What would be the effect, if any, on the sound produced by your speakers of a Zobel network in the speaker cables?

If the network is made up of a 10 ohm resistor and 100nF capacitor, its critical frequency is about 160kHz. This is pretty darn high and it shouldn't create an audible effect. Maybe this is why these values are used by audiophiles. It only has a stabilizing effect well outside the range of human hearing but it does 'something' at frequencies where an amplifier might oscillate.

Please notice that this particular Zobel network does not make the speaker more 'resistive' at audio frequencies, where the speaker is meant to work. This is the assumption both Ivor and I made when discussing Zobel networks but it turns out audiophiles have their own definition of what a Zobel network is supposed to do. These values don't do anything to compensate for the reactance of the woofer or tweeter so it really isn't a Zobel network in the first place. Of course, don't try and tell this to an audiophile...

For comparison, the values I used in my guitar amp were 5 ohms and 33uF for a critical frequency of about 1kHz. This definitely had an audible effect on the highs, which may or may not have been a good thing. I really would like to know what is inside those Transparent Opus cables.

 

[Yoink]

I really would like to know what is inside those Transparent Opus cables.

Well, buy some and take them apart!

Seriously, though, would it be a simple thing to find out what's in them simply by measuring input/output? Or would you need to actually explore them physically?

And would it be a fair summation of your post (for which I thank you) to say that a very high capacitance speaker cable with a Zobel network in it might conceivably have the effect of slightly dampening audible high-frequency sounds?

 

[bjb]

Yes, they do.

Check out figures 3 & 4 on the LM4765 datasheet:

http://cache.national.com/ds/LM/LM4765.pdf

They are not always required, but most amplifiers have some kind of network to isolate the load from the amplifier output and/or provide a load for the amplifier at high frequency.

That's great! I'm more familiar with vacuum tube amps and we don't use those sorts of networks. Some tube amps did use an R-C across the output transformer to straighten out the load line for the power tubes, though, and R-C networks are often found in the feedback loop.

The components are in Figures 2, 3, and 4 on pages 5 and 6. The values are 4.7 ohms and .1uF. This works out to 100nF so the network is very similar to what the audiophiles like. But since these are meant to suppress oscillation, and not compensate for speaker reactance, I'm still not sure if they should be called Zobel networks. The networks do stabilize the 'load' impedance (speaker + wire + stray capacitance) so I suppose they're Zobel-like after all.

The amplifier simulation contains these sorts of components. The values aren't given in the article, but Figure 1 shows C3/R7 are in parallel with the output while L1/R8 are in series with the output. Even though the simulated amplifier contains the built-in networks, the simulation indicates an improvement with a speaker-end network. Maybe there's something to this after all but I'd be more convinced if I could see some test data showing suppression of oscillations.

 

[krelnik]

Oh, you remember Monster Garage on Discovery Channel? Monster Cable went after them and now they own all the trademarks for Monster Garage.

Do you have a link for that? The only ones I can find indicate that they sued in 2004, and there was a "confidential" settlement some time after that. The show ended in the middle of 2006, and Discovery still sells it on DVD in their store with the original name and logo on the package.

There does seem to be quite a cottage industry of critizing Monster on the web, both for their product quality and their legal shenanigans. They have lots and lots of detractors.

--Tim Farley

 

[bjb]

Well, buy some and take them apart!

Seriously, though, would it be a simple thing to find out what's in them simply by measuring input/output? Or would you need to actually explore them physically?

And would it be a fair summation of your post (for which I thank you) to say that a very high capacitance speaker cable with a Zobel network in it might conceivably have the effect of slightly dampening audible high-frequency sounds?

Assuming the Transparent Opus cables contained an R-L-C network, it would be possible to take some measurements and guess at what was inside. There's a way to do this sort mathematical reverse-engineering but it would enough to take some measurements and show if/how the network affects the frequency response into a purely resistive load. All you need is a signal generator (or a stereo and a test CD), a DMM, and a non-inductive resistor (Radio Shack has them) and you're in business.

I would need some numbers so I could run a PSpice simulation. One problem I have with audiophiles is that they propose all sorts of interesting mechanisms but never do any analysis or measurements to show that the mechanisms will have an audible effect. If I answered the question without doing a proper analysis, I'd be guilty of the same thing.

I can say that the 10 ohm/.1uF network probably does not have an audible effect, but I know the network I tried (4.7 ohms/33uF) did suck out some highs. Without knowing the values used in a particular Zobel network, there is no single answer.

 

[Yoink]

Assuming the Transparent Opus cables contained an R-L-C network, it would be possible to take some measurements and guess at what was inside. There's a way to do this sort mathematical reverse-engineering but it would enough to take some measurements and show if/how the network affects the frequency response into a purely resistive load. All you need is a signal generator (or a stereo and a test CD), a DMM, and a non-inductive resistor (Radio Shack has them) and you're in business.

I would need some numbers so I could run a PSpice simulation. One problem I have with audiophiles is that they propose all sorts of interesting mechanisms but never do any analysis or measurements to show that the mechanisms will have an audible effect. If I answered the question without doing a proper analysis, I'd be guilty of the same thing.

I can say that the 10 ohm/.1uF network probably does not have an audible effect, but I know the network I tried (4.7 ohms/33uF) did suck out some highs. Without knowing the values used in a particular Zobel network, there is no single answer.

To get at the heart of my question (relative to the MDC that Mike Lavigne hopes to mount): given that it is possible (theoretically) for these networked cables to introduce audible distortion into the sound, would it also be possible for JREF to propose a straightforward set of benchmark tests that the Transparent Opus cables would have to pass (impedance, capacitance etc.) before accepting them as "neutral" competitors with regular speaker cables--or is it simply unsafe to accept them as viable options for a blind test?

 

[bjb]

The resistor and capacitor are passive linear elements and should not introduce measurable amounts of distortion. It is true that some resistors can produce distortion under certain condition but a speaker cable isn't an application that overstresses components.

I have not seen good evidence that capacitor distortion exists in speaker cable applications but it might be happening in vacuum tube amps, where high DC and AC voltages exist. A few years ago, I had a technician test a set of very different capacitors with a very, very expensive digital bridge. The results came back and they all tested about the same! This was very strange to me because I included a faulty cap in the test. It turns out the bridge only tested at low voltage and the faulty cap was only 'faulty' at high voltage. Since speaker wire does not carry high DC voltages those exotic forms of distortion that audiophiles love to fear should be absent.

A benchmark for speaker cables would be to require that the AC impedance of the speaker wire should be 1/10 the impedance of the speaker load. This meets the condition of impedance bridging:

http://en.wikipedia.org/wiki/Impedance_bridging

It's an old rule-of-thumb that ensures that the load impedance is much larger than the source impedance in order for the signal to reach the load. Our case is a little different because the speaker cable places elements in parallel with the source as well as in series. Some calculations would need to be made to ensure the impedance bridging rule is met.

I'm not sure if the impedance bridging would have to be valid over all frequencies or only the the audible frequency range. A cable that added a Zobel network that only affects ultrasonic frequencies but stabilized the amp might actually do some measurable good, in which case the paranormal aspect of the challenge would not apply. In other words, I agree with Randi when he says his challenge is for speaker wire only, not for speaker wire that includes circuit-correcting components like Zobel networks.

 

[Dan O.]

References for the current system were not available online but the closest match I could find gave an output impedance of the amp of .76 ohms and a speaker impedance of 4 to 8 ohms. So we may not exactly be in the impedance bridging realm.

 

[Ivor the Engineer]

Some people may be interested to read this Doug Self article:

http://www.dself.dsl.pipex.com/ampins/pseudo/subjectv.htm

Concern over cables is widespread, but it can be said with confidence that there is as yet not a shred of evidence to support it. Any piece of wire passes a sinewave with unmeasurable distortion, and so simple notions of inter-crystal rectification or "micro-diodes" can be discounted, quite apart from the fact that such behaviour is absolutely ruled out by established materials science. No plausible means of detecting, let alone measuring, cable degradation has ever been proposed.
The most significant parameter of a loudspeaker cable is probably its lumped inductance. This can cause minor variations in frequency response at the very top of the audio band, given a demanding load impedance. These deviations are unlikely to exceed 0.1 dB for reasonable cable constructions. (eg inductance less than 4 uH) The resistance of a typical cable (perhaps 0.1 Ohm) causes response variations across the band, following the speaker impedance curve, but these are usually even smaller at around 0.05 dB. This is not audible.

 

[bjb]

References for the current system were not available online but the closest match I could find gave an output impedance of the amp of .76 ohms and a speaker impedance of 4 to 8 ohms. So we may not exactly be in the impedance bridging realm.

What current system are you talking about? Here's another amp with an output impedance of 0.22 ohms:

http://www.stereophile.com/amplificationreviews/815/index5.html

At 0.22 ohm, the output impedance was a little higher than is usually found with a solid-state design, presumably, like the high gain, due to the absence of loop negative feedback.

The author believes 0.22 ohms is 'high' for a solid state amp, but the amp doesn't have negative feedback. Negative feedback is used in most amplifier designs but it is forbidden in some versions of the audiophile religion.

Other amps do use negative feedback and their output impedance is much lower. For example, the Parasound JC-1 mentioned in the Stereophile article claims a damping factor of 1200. Using the equation from the Wiki article on impedance bridging, this means the output impedance is 8/1200 ohms, or .007 ohms.

In either case, I don't believe the speaker cables would make a difference in the sound of the amps. Even so, I would prefer to use an amplifier that meets ordinary design practices. Michael Fremer already reviewed the Parasound amp and they cost $6,000 for a pair, so I'm sure a test using these amps (or something like them) would be acceptable to the average audiophile.