Tube Tone

[trem]

Do tubes have a frequency response.?
I read about tubes being dark, shrill, lifeless, or whatever.
Is that possible.?
If two different AX7 have the same gain, can one of them be more "bright" than the other.?
Does it involve the capacitance of the tube itself, with the circuit that it is connected to.?
Thank You

[teemuk]

Do tubes have a frequency response.?

No.

I read about tubes being dark, shrill, lifeless, or whatever.

I've read people hearing differences between different coloured power cords. Doesn't mean everything you read is true.

[Chuck H]

Certainly different brand tubes sound different. At least WRT guitar amps.

[Jazz P Bass]

To me, that is a pretty open ended question.
Do they or can they is more specific.
Are you asking of preamp or power amp tubes?
NOS, new current production or old worn out, hey I dropped it tubes?
I don't think current production methods produce near the quality of tubes as in the heyday of tube manufacturing.
I also think they ship (& maybe 'grade' as such) whatever comes off the assembly line.
I had a EVH 5150 that I took every tube that I had & put it in V1 (the hiss stage).
Every tube affected the hiss differently.
A 'normal' gain stage should not be too picky about what tube you put in.
Extreme circuits (ie: EVH 5150 V1) may want a specific tube.
As to 'frequency response', there is a limit to what a preamp tube can pass (miller effect).
That is why there are specific tubes for radio.

[Chuck H]

WRT current production tubes alone, anyone can hear the difference between, say, a Shuguang 12ax7 and a Sovtek 12ax7WA. Especially in the first preamp spot. The OP was asking about 12ax7 types I think. JPB makes a good point about the noise issues too. The first preamp spot in any amp is going to be the most important and sensitive to differences. By a huge margin. In a moderate to high gain amp I can hear a difference between almost any two 12ax7's in the first preamp spot. The tubes I mentioned above easily demonstrate different high end characteristics.

As to the ideal that anything labled "12ax7" is the same because it must be within margined industry standards... It's not like that. Back when tubes were all there were I'm sure it was much more consistent. But, again WRT guitar amps, the circuits often use less NFB and less ideal operating parameters than hi fi. So the differences between tubes and how they respond to less than ideal circuits becomes much more obvious.

[trem]

Suppose you played the same amp, and changed the AX7 in V1.
What part of tube manufacture would make 1 tube sound darker than another to a player.
With a speaker, there is a frequency response chart that you can look at and see how it responds throughout its range.
Do tubes do that.? Do some sound brighter than others for some reason.?
I kind of thought they had a linear gain response to whatever frequency came into their grid.
Thanks

[Chuck H]

In a perfect world tubes have a linear response. The internal plate to grid capacitance probably increases with gain. And is anybodys guess when a tube is clipping. I don't know enough about tube construction to say what might affect the frequency response. It's an ideal that tubes simply mirror the grid signal. But it's not an ideal world. The absolute fact that different tubes sound different proves that. I don't know of any chart like graphing for tubes. And I really question the accuracy of many currently made tubes data sheets. Thing is, players like different speakers to sound different. But I think most players would prefere that different tubes sound the same. As noted above, a 12ax7 thould always be a 12ax7. Just like a 10k resistor should always be a 10k resistor. I can't imagine any tube manufacturers stepping up to say that their product is different in any way. Even though they often are. There are plenty of reviews and shootouts on line that can be researched to get a very opinionated evaluation of specific brand 12ax7's tonal character (did I really need to say that?).

[R.G.]

Frequency response can be measured. However tone, like pitch, happens inside the human head, and is not directly measurable.

That being said, humans do have some soggy general directions about relating frequency response to tone. We can measure frequency response to a fair-the-well. "Tone" depends on all kinds of things, including the listening humans' moods. For instance, lower relative amounts of high frequencies can be heard as "dull", "dead", "warm" or "brown", to quote just a few. Lower relative amounts of low frequencies can be heard as "shrill", "sharp", "tiring", "clear", "ringing", "glassy", and many others, all describing the person's emotional experience of the sound.

To abuse an aphorism, tone is like noses: everybody's got one, and they're all different.

Do tubes have a frequency response?
Yes, but it has meaning only within the context of the larger circuit it's connected into. Tubes are **mechanical** devices, which rely on physical separation of the internal parts and the electrical fields put on them to divert the flight-paths of electrons through the internal vacuum. The physical positions of the parts put limits on how fast electrons can be diverted, but the circuit outside the tube puts on other limits as well.

From a physics point of view, any two conductors separated by an insulating medium have capacitance, which is an electrical field effect. That's fixed by the positions of the metal parts inside tubes. However, the effect on the electrons passing through is determined by the electrical fields, which is varied by the external circuit's effect on the voltages on the metal parts.

The same tube will have a different frequency response depending on whether you stick signal into the grid or cathode (both work, just differently) and if the grid is used, what the voltage gain set up by the external parts is.

I read about tubes being dark, shrill, lifeless, or whatever.
Is that possible.?
Yes, but bet on the external circuits and how well they extend or retard the internal capabilities of the tube. Can one horse run faster than another? Sure. But different horses perform differently on mud, grass, sand, in rain, etc. A winner is not always clear. Worse yet, humans do the judging by blipping a few guitar chords or songs, then saying how that made them feel. That's not the best way to measure things.

If two different AX7 have the same gain, can one of them be more "bright" than the other.?
Yes. Gain is different at different frequencies. "Gain" is often measured at DC. It's different as you go to AC and higher frequencies.

Does it involve the capacitance of the tube itself, with the circuit that it is connected to?
Yes, both. And the resistors. If you have a tube set up with gain, determined at low frequencies by the resistors and big caps, then you will get some frequency response. If you increase the low frequency response, you at the same time reduce the high frequency gain by a quirk called the "Miller Effect", which multiplies the effective value of the internal capacitors by the AC gain.

[trem]

Chuck
R.G.
Everybody
...Thank you so much. I think that pretty much clears up that aspect of tubes for me.
One down. One million to go.

[exclamationmark]

I think there are 2 distinct concepts at play here that relate to 'tube tone'. One is the actual concept of frequency response (as in gain vs frequency response plots - I think fully bypassed triode are pretty flat in that regard, except for the HF rolloff), and the other has more to do with non-linearities (ie, distortion, even harmonics, odd harmonics and all that jazz). It's quite easy to see a difference between different tubes in an identical circuit with a spectral analyzer or a high quality sound card (hi-fi nuts do it all the time) - the differences become even more drastic when driven to the point of clipping. No two tubes are ever made the same either. I remember trying to match some triode sections via their AC gain, and it was surprisingly difficult. I had these 2 triode sections that were vastly different in gain at low signal levels, but became almost identical when driven by a large signal. It just goes to show how the grid curves vary even within the same bottle!

Whilst the structure of a tube isn't that complex, the physical interactions between the various components of the tube certainly are. The grid current drawn by different models of 12ax7's vs bias voltage is fairly unique, and it differs between manufacturers (though I imagine some of it may have to do with the degassing of the tube). The same applies to the cut off characteristics. Even a 'clean' amplifier with 2 gain stages will undergo significant clipping under transient conditions, and I think this is probably where most of the apparent 'mojo' from specific tubes comes from, rather than from their small signal characteristics. I'd also expect that some manufacturers 12ax7's grid curves are vastly different to their competitors, which account for some of the small signal differences, which may or may not be audible (I once tried tube rolling a single gain stage boost pedal - couldn't hear a damn difference and it was re-amped through my computer).

[Steve Conner]

There are two ways that a tube can affect the frequency content of a signal that it's amplifying.

1. By amplifying some frequencies more than others, the traditional concept of "frequency response". The only mechanism for frequency-dependent behaviour in a triode is Miller capacitance, which could roll off the high end somewhat. Miller capacitance is the tube's plate-to-grid capacitance multiplied by the gain, so it can vary from tube to tube.

2. By creating new frequencies that weren't present in the input signal, in other words distorting it. IMO, this is the most significant effect. A tube that generates more high order harmonics will sound "brighter", but only when it's overdriven, because the amount of harmonics generated depends strongly on the drive level.

Having said that, the tubes are overdriven to some extent even in the clean channel of a guitar amp, so I would expect the distortion signature to be audible pretty much anywhere.

And, the distortion signature depends on the details of the tube's mechanical design, so again it can vary from tube to tube.

I'd sum up by saying that tubes probably don't have a frequency response but they still manage to sound different anyway.

[trem]

It's funny.....
Scientists can tell us the mass of the black hole, at the center of our galaxy (thousands of light years away) but there "never" seems to be any Definitive information about tube guitar amps.
Perhaps "we" stopped teaching this stuff too soon. It is basically 1930's technology, obsolete in many peoples minds, but mysteries still seem to abound.
I should have been born sooner.
Of course, if I had been, I might be dead now.
best

[Chuck H]

I should have been born sooner.
Of course, if I had been, I might be dead now.

Sigged!

[Alex/Tubewonder]

A tube by itself does not sound at all. It's one of components in the circuit and it's the circuit that produces what we in the end receive as sound.
As for some points raised earlier:
- if your triode leaks grid current then it has become gassy. You may like the sound but it will not be a constant, the tube will grow more gassy by the hour.
- If two different tubes swapped in the circuit end up at different operating plate voltage & current then the circuit will behave differently. Adjust the cathode resistors between swaps so each tube operates at same conditions. You'll have great difficulties detecting a difference in sound.
- If you are just a listener and I act as tube swap monkey with you not knowing if at any instance I swapped tubes or not- you will experience a phenomenon known as "wishful hearing".
- power tubes, I just found out EL34 produce a different sound than KT66 when dropped on a hardwood floor. Both sounds had a profoundly negative effect on my mood. I will not repeat this test with my NOS EL156.....

[Chuck H]

I agree that just swapping tubes doesn't do a thing for the scientific process. And I admit to never having adjusted the circuits for each tube. But your post is just too cut and dry for me to subscribe. I'm telling you now, compare a Sovtek 12ax7WA to a Shuguang 12ax7 (selected for low microphonics). Adjust for bias or anything else you need to to achieve the same operating conditions. They will sound notably different. Until you do this test you can't accurately say anything to the contrary.

I think the issue is that currently made 12ax7's are all different tubes. Some which only resemble actual 12ax7 specs. km6xz has mentioned this regarding current production in both preamp and power tubes. His discoveries were made while testing different tubes to vender as a new business venture. So, in this light, saying all 12ax7's sound the same simply by their virtue of being 12ax7's is like saying that 6l6's sound exactly like el34's or 6V6's like el84's, with all things being equal. Perhaps one reason different 12ax7's DO sound different is that they are all slightly different tubes.