Do different make 12AX7 tubes sound different? and why?

[Groover]

We had so much fun with the "different types of caps thread" I thought I'd ask about tubes. I'm really interested in all types, but for the sake of simplicity I thought I'd limit it to 12AX7s.

A tube must meet certain criteria to be a 12AX7, correct? So what aspects of the design execution are variable from one maker to another, and how do those aspects create tonal differences?

This should be fun!

[Wilder Amplification]

We had so much fun with the "different types of caps thread" I thought I'd ask about tubes. I'm really interested in all types, but for the sake of simplicity I thought I'd limit it to 12AX7s.

A tube must meet certain criteria to be a 12AX7, correct? So what aspects of the design execution are variable from one maker to another, and how do those aspects create tonal differences?

This should be fun!

Not a difference worth mentioning. Cork sniffing at its best.

[Groover]

I do hear a lot of people "tube rolling" in guitar amps as well as in the audiophile world.

But the more I hear answers like Wilder's, and R.G.'s in the caps thread the happier I am. Less stuff to get a brain tumor over.

[JoeM]

There's some audible difference among various 12AX7s. Probably not a difference I would hear with a band at volume, but it's there. Some I like more than others. It seems to me at least, that they work differently in different amps too, for example a tube that I might really like in my 5F6A clone as V1, I may not like as much in my SFSR.

Same thing with 6L6s.

[Wilder Amplification]

Probably not a difference I would hear with a band at volume,

Which is exactly why I said "not a difference worth mentioning".

My personal opinion as to why people "tube roll" is because most players don't know enough about amps to mod them and as such want to feel that they have SOME sort of control of their tone that we techs have, and tube rolling gives them this mental satisfaction that they're somehow making drastic tonal changes via tube rolling because it's all they understand how to do.

Just my $0.02. Maybe someday players will realize that their own two hands have more control of their tone than any of the components do, but just like in the hot rod community people would rather blame the parts than blame themselves as the source of a problem.

[JoeM]

Which is exactly why I said "not a difference worth mentioning".

I do think it's "worth mentioning" however. Many players do seek the best tone from their amps, and selecting tubes is certainly one of the things easily do-able. But changing tubes isnt going to make a bad amp sound good (or vice-versa). (Well, not within reason of course, assuming quality tubes)

Sadly even the 'best' tubes I have dont make me sound like Larry Carlton, so I appreciate your comment that there's more tone (or less!) in your hands.

[Steve Conner]

Did you see this thread?
http://music-electronics-forum.com/t17410/

I tested one of those Techtube dual triodes against an original Mullard, and it really did sound different. I made sound clips so you can hear for yourself.

But you can see differences in the characteristic curves that would explain the different tone, and Techtube marketed it as an "ECC813" or something, implying that they thought it was stretching the definition of ECC83 a bit far. Sadly Techtube went out of business.

I believe this is an extreme example, because the Techtube was an oddball design made from TV electron gun parts. With any two 12AX7/ECC83s of "normal" construction the difference seems to be less. I found a couple of old Telefunkens that sound different, but I think they're just old and worn out.

[R.G.]

Yeah - there are differences. The size of the difference depends on how close the internal design of the physical elements is and the care with which it's manufactured. And then there's the "personal opinion multiplier", the internal urgency inside the listener's head that evaluates how different the sound is.

Tubes are mechanical parts. How they amplify depends on the materials and physical positions of the parts. Change, for instance, the spacing of the grid wires, and you change the gain. Make the spacing inconsistent from the bottom of the grid-wire spiral from the bottom of the grid to the top and you make the gain different from the bottom of the grid to the top, as well as changing the grid-to-plate capacitance and its effect on frequency. As well as changing how many of the electrons flowing through hit the grid wires, which changes the grid leakage, and the grid shutoff voltage, and...

Then there's manufacturing variation. If every single tube is dead identical to the last, and they're all pumped down to vacuums in which no more than two gas ion molecules are present inside the glass, they will be very consistent. If some tubes have lots of gas and others little, they'll be very inconsistent. If some of them have good vacuum and others have old bubble gum wrappers inside, it gets worse.

Once the manufacturing machines got set up and maintained well, the tubes got pretty consistent. I remember reading about the US military wanting tubes from the very biggest makers of tubes, not because bigger is better, but because in quantity there is consistency. I also read that at GE (I think it was) the tubes got worse after everybody was off for a holiday, and then better and more consistent a day or so later when all the machines and workers were back in the groove.

Then there's the circuit they get put in. Tube circuits use very modest feedback factors, which reflects that tubes were pretty consistent. But the feedback did vary. The more feedback you use, the less the underlying part can color the sound; opamps are the extreme example. Opamps were first invented and implemented with tubes, the Philbrick opamp being the example. 16" naval cannons on US battleships were aimed using Philbrick opamps. In the Philbrick, the tube quality mattered not much at all, because the feedback washed it all out.

In guitar amps, some positions use the tube with lots of AC feedback, and some with little. Those positions with little feedback let the underlying linearity and frequency response of the individual tube show through. Those with lots of feedback hide it.

And we already know what the "personal opinion multiplier" does.

[Wilder Amplification]

But changing tubes isnt going to make a bad amp sound good (or vice-versa). (Well, not within reason of course, assuming quality tubes)

And that is exactly the misnomer that the "different preamp tubes sound different" statement has created amongst the guitar community. Yes there is a subtle difference of course...there's going to be. The problem comes in when you tell average players who aren't techs that don't understand this stuff a statement such as "Different tubes sound different" and they take this to the bank a bit too literally and think that if they have an amp circuit that wasn't voiced well to begin with (i.e. the DSL) that different tubes are going to make this night and day difference that just turns a turd amp into a holy grail amp. It is for this reason that I stated "Not a difference worth mentioning".

Some people just don't understand that you can polish a turd all you want but in the end, it's still a turd.

[soundmasterg]

That was an excellent post R.G.

I was going to just say that since a tube is a valve, the differences between them is basically a more or less efficient valve depending on mechanical defects in their construction. But that is rather simple compared to your explanation. I know from my own experiments that as you increase the gain in the circuit, you get less and less audible difference among tube swaps assuming that microphonics isn't a concern. As far as 12AX7's are concerned, I generally prefer European old stock 12AX7/ECC83's to American ones, and modern 12AX7 types can sound good in some cases but generally don't last as long as vintage ones.

Greg

[the great waldo]

And that is exactly the misnomer that the "different preamp tubes sound different" statement has created amongst the guitar community. Yes there is a subtle difference of course...there's going to be. The problem comes in when you tell average players who aren't techs that don't understand this stuff a statement such as "Different tubes sound different" and they take this to the bank a bit too literally and think that if they have an amp circuit that wasn't voiced well to begin with (i.e. the DSL) that different tubes are going to make this night and day difference that just turns a turd amp into a holy grail amp. It is for this reason that I stated "Not a difference worth mentioning".

Some people just don't understand that you can polish a turd all you want but in the end, it's still a turd.

I tried about 20 different preamp tubes in a 65 Fender deluxe reverb that belongs to one of my workmates (he was playing his guitar thorugh the amp while I was swapping just one of the preamp tubes with all the tubes that I had collected over the last few years) while most of the tubes sounded different to some degree one yellow script mullard really did make a big difference (improvement) to the amps sound. Needless to say I let him keep the tube and I did wonder what would have happened if it would have been possible to swap out the other tubes. I am sure any well designed amp will sound good with decent quality modern tubes. I did have an aquantance in London that worked for an electronics whloesaler and he told me that about 30% of the tubes that mullard produced were disposed of as being not good enough for their quality control. I can´t imagine any modern companys´accountants allowing that today.

Cheers
Andrew

[rf7]

about 30% of the tubes that mullard produced were disposed of as being not good enough for their quality control.

From what I've heard that is par for the course in tube manufacturing. And I've heard of worse rejection ratios for tube manufacturing. That's just part of why they are relatively expensive.

[boots]

From what I've heard that is par for the course in tube manufacturing. And I've heard of worse rejection ratios for tube manufacturing. That's just part of why they are relatively expensive.

What do you suppose happens to all the rejected tubes? Do you think those are the ones that are "re-branded" by smaller companies? I understand that re-branded tubes vary greatly in quality - some of them are actually decent tubes (made by major manufacturers) - and of course some of them are duds. I have a couple ancient-looking 12AX7 (no suffix) branded by "Rad Tel". The glass envelopes are a little mis-shapen, but boy, they are some of the best sounding 12AX7's in my collection. Anybody else ever heard of Rad Tel?
-Boots

[boots]

I think RG is right on. I have taste-tested quite a few 12AX7s as well as 6V6s and 6BQ5s, and there are definitely differences in the sound, at least to me. And the personal opinion factor is a very real thing as well, but I try to be objective in my listening tests. Sometimes I "want" to like a highly regarded tube because everybody says it is great, but it just doesn't flip my switch. And vise-versa - some of my favorite 12AX7s are some oddball rebranded ones.

It's shouldn't be surprising that different tubes sound different. The differences in tone we hear are due in part to frequency response, and frequency response is affected by differences in R, L, and C both inside and outside the tube. We spend a lot of time and effort voicing our amps by changing out caps in the circuitry outside the tubes, but we can't do much to alter the internal capacitances inside the tube. The internal capacitances vary from one manufacturer to the next, and from one tube of the same brand to the next.

Although I'm not an engineer, I assume that distortion character is at least partially affected by interelectrode spacing, which varies from tube to tube as well. I think it all boils down to the fact that, unlike solid-state devices, there are unavoidable physical differences from one tube to the next that make tubes a little like snowflakes - they are all a little different.

So to me, that seems a reasonable explanation why tubes sound different. The difference may indeed be subtle - if you are playing a wall of Marshall stacks in a stadium, it's probably pretty tough to tell any difference. But to geeks like us, those subtle differences make for a wonderful, twisted, addicting hobby.