If "Tube Sound" Is a Myth, Why Tubes?

[mocenigo]

Exactly the same. Silver, contrary to popular belief, doesn't work as an equalizer favoring higher frequency (frequencies between 5kHz and 20kHz or higher). Silver cannot possibly sound 'brighter/sharper' even if it really wanted to.

short braided silver cable vs original 3m coiled Beyerdynamic cable with DT1770.
As the DT1770 is 250 Ohm cable resistance is not of any (audible) influence.
For those looking for the purple trace... it is hiding behind the teal trace so that's why you don't see it.
We are looking at acoustical measurements here b.t.w.
Oh.... harmonic distortion and CSD also did not change.

But, but, but, it is silver coloured, it MUST be bright!!!!! ;-)

 

[mocenigo]

Just to clarify, impedance you read about cables are characteristic impedance, which accounts for resistance, capacitance, and inductance of a cable. So it’s not a (DC) resistance you can read using a DMM. So it’s not that simple as you illustrated.

meanwhile, analog audio only uses up to 20-40khz depending on application, the frequency is just too low for characteristic impedance to affect frequency response. So in practice, pro cable makers make cables with 40-150 ohm impedance for analog use and no problem would occur.

even if you are unhappy, impedance matching circuits (active compensation) is possible using buffer amplifier.

Re: characteristic impedance: you are right. I was oversimplifying, but it is not that as materials copper and silver conduct in a significant different way. And, yes, I use buffers

 

[Wes]

Why Tubes?

... because phaat J-Fets don't glow

 

[Veri]

Why Tubes?

... because phaat J-Fets don't glow

Anything can glow when you have colorful LEDs protrude from within the case ((just leave out the tubes))

Actually, scrap that and don't give anybody any ideas. Next thing we know there's gamer-LED-infested audio gear on the market.

 

[mocenigo]

Anything can glow when you have colorful LEDs protrude from within the case ((just leave out the tubes))
View attachment 65397

Actually, scrap that and don't give anybody any ideas. Next thing we know there's gamer-LED-infested audio gear on the market.

You can have green glow for far less: https://www.mcintoshlabs.com/products/merchandise/LB100

 

[Veri]

You can have green glow for far less: https://www.mcintoshlabs.com/products/merchandise/LB100

Oh my... A must for every McIntosh owner!

 

[mocenigo]

I sometimes think too much is as worse as too little, unless it’s necessary.

Anyway, since both the DC and AC impedances of copper and silver are the same within 7% IIRC, I really would like to see where they can measurably have an effect when used in topologically otherwise identical signal or power cables...

 

[mocenigo]

I sometimes think too much is as worse as too little, unless it’s necessary.

(As for buffers: yes I have one between DAC and power amp, and it does make a difference)

 

[Julf]

If a 24awg copper quad sound the same with 28awg silver plated quad, obviously silver plating is doing something.

Not sure I get your point. "If X and Y sound the same, there must be a difference"?

 

[mocenigo]

Not sure I get your point. "If X and Y sound the same, there must be a difference"?

24awg and 28awg are significantly different cross sections (do not ask me about the ratio, american measurement systems are insane), but if the impedance of the material is the same, so argues the poster, and the section is different, then the material must do something to balance the difference.

The actual explanations is simpler: signal cables are used to transfer a low signal, which is usually driven in voltage, not current. So who cares about the cross section? Lore is in fact “the thinner the better”.

Speakers are current driven, so this is a different matter, and cable section counts.

This said, I was talking about pure silver, not silver plated copper.
Regarding silver plated copper, there is an interesting interview to Tommy Jenvig of Supra on TNT Audio that explains why silver plating (not pure silver) is detrimental to sound (and why he preferred tinning). It is quite interesting.

 

[Blumlein 88]

24awg and 28awg are significantly different cross sections (do not ask me about the ratio, american measurement systems are insane), but if the impedance of the material is the same, so argues the poster, and the section is different, then the material must do something to balance the difference.

The actual explanations is simpler: signal cables are used to transfer a low signal, which is usually driven in voltage, not current. So who cares about the cross section? Lore is in fact “the thinner the better”.

Speakers are current driven, so this is a different matter, and cable section counts.

This said, I was talking about pure silver, not silver plated copper.
Regarding silver plated copper, there is an interesting interview to Tommy Jenvig of Supra on TNT Audio that explains why silver plating (not pure silver) is detrimental to sound (and why he preferred tinning). It is quite interesting.

Here is that part of that interview. The old diode hopping business.

LC > Silver and OFC copper are the most used materials for Hi-Fi cables. New materials such as carbon fiber, titanium or metal matrix seem to have interesting audio properties.
Is R&D at Supra moving towards new conductors?

TJ > Even here we have practiced lateral thinking and done the opposite to the common designing.
The facts that silver has a lower resistance than copper and that the silver oxide is conductive do not make it a suitable material for wire plating in speaker cables. The OFC copper is fine, but how long time does it stay Oxygen Free, unless you protect it by plating? A PVC- insulation is for example emitting chloride ions that make copper-chloride on the copper wire surface, a process that accellerates with current flow through the cable in use. (Is that the burn-in of a cable?) This problem is increased by the fact that the strands of a wire are twisted, leading to the current jumping from strand to strand when trying to go the straightest runway. Thus, the current passes the oxide of the the surfaces of the strands thousands of times per metre run in a cable.
So, what is the use of the OFC? Silver plating is a no good solution to this as it even increases the surface run, by means of its lower resistance. And the silver oxide, although conductive, makes diodicity thresholds and non-linearity. So, again, we choosed the opposite way. We tin-plated the strands, instead, as tin has a higher resistance, hence, making the current mainly be kept inside the copper which stays OFC, with less jumps and less oxide. Again, a better result at a lower price.
In application this is the Supra Ply speaker cable. Silver is, however, good for high frequency applications. Regarding other materials we use carbon fibers in our screens on a strong nylon base, but Titanium we have not tried in production, as yet.

 

[scott wurcer]

24awg and 28awg are significantly different cross sections (do not ask me about the ratio, american measurement systems are insane),

Please explain how the SWG is any less "insane" than AWG?

 

[Julf]

Please explain how the SWG is any less "insane" than AWG?

Both are arcane. The sane way is to use cross section in mm2.

 

[Julf]

The actual explanations is simpler: signal cables are used to transfer a low signal, which is usually driven in voltage, not current. So who cares about the cross section?

That was my point.

 

[mocenigo]

Please explain how the SWG is any less "insane" than AWG?

See Julf's reply

 

[mocenigo]

Here is that part of that interview. The old diode hopping business.

Why facepalm? Is this insanity? (Sincerely asking for a scientific rebuttal, I have no preconception here, I just thought this was meaningful.)

 

[solderdude]

IF a cable changes something in a 'diode junction' kind of way this would mean you would be measuring non-linear distortion.
Instead of theorizing they could have easily provided proof using measurements.
Where are the measurements showing silver, tin or non plated copper cables showing different (and audible level) distortion products (including noise) ?
If it was impactful we would be seeing tons of plots with rather substantial distortion differences between cables.
Certainly if they would also be able to show peaks or dips in FR (not induced by Ohmic losses and wildly varying impedances).

It's all sales-pitches and bla bla disguised as 'TOTL tech talk'

To get it back to tube talk... there are probably audible differences between the steel wires and bonds with plates/anodes/cathodes which are part of the magic between tubes. It's crazy most of the times copper traces, goldplating, internal wiring, copper leads of parts are usually left out of discussions... or is the magic difference between components caused by exactly this.

 

[Blumlein 88]

Why facepalm? Is this insanity? (Sincerely asking for a scientific rebuttal, I have no preconception here, I just thought this was meaningful.)

What solderdude wrote. Saved me the trouble.

 

[dfuller]

To get it back to tube talk... there are probably audible differences between the steel wires and bonds with plates/anodes/cathodes which are part of the magic between tubes.

I'd hazard a guess that barring tolerance for mu, gm etc (i.e. if you have two bogie spec tubes from different manufacturers in the same circuit), most of the differenc in the linear region is the inter-electrode capacitances of the tube , as well as how microphonic the tubes are (and make no mistake, all tubes are microphonic to some degree - part of the advantage of transistors is that they essentially don't have any microphony).

I figure that's why, for example, an RCA 12AX7 sounds somewhat different from a Tesla E83CC despite them being built to the same electrical specs. The Tesla in my experience (albeit sighted and most definitely not in a high fidelity system) is a bit less lively/jangly but is somewhat brighter despite that - which if I had to guess has something to do with construction differences. The RCAs all have longer plate designs than the Tesla, and the Tesla is built with a frame grid instead of a traditional design. That tells me that the Tesla simultaneously has lower inter-electrode capacitance and microphony. Of course, cathode coating and plate material choice do play into it as well, so that's part of it too, not to mention all the welds vs crimps for joining plates etc.

 

[SIY]

I figure that's why, for example, an RCA 12AX7 sounds somewhat different from a Tesla E83CC despite them being built to the same electrical specs. The Tesla in my experience (albeit sighted and most definitely not in a high fidelity system) is a bit less lively/jangly but is somewhat brighter despite that - which if I had to guess has something to do with construction differences. The RCAs all have longer plate designs than the Tesla, and the Tesla is built with a frame grid instead of a traditional design. That tells me that the Tesla simultaneously has lower inter-electrode capacitance and microphony. Of course, cathode coating and plate material choice do play into it as well, so that's part of it too, not to mention all the welds vs crimps for joining plates etc.

Depends a lot on the circuit. Maybe principally on the circuit.

Have you measured the interelectrode capacitances? I don't have a Tesla E83CC on hand, but I've compared long plate Telefunken ECC83 with some of the very short plate Eastern Europe 12AX7/ECC83 and they weren't very far apart. Cgp is the most critical for common-cathode topologies.