Telefunken ECC83/12AX7

[DaveP]

I think that audiophools buy these expensive tubes because they are basically scared.  They have so little actual experience of making gear that they throw money at it to make themselves feel confident.

Lets look at it in more detail:
An ECC83/12AX7 has a certain spec that is dictated by the dimensions of the parts and the distance between the grids.  If the gain is in spec, then the distances are identical.  The gm is due to the effectiveness of the cathode and emission paste on the heater coils.

There is a good chance that the emission paste (triple alkaline earth carbonate) was purer in the 30's, 40's and 50's.  I know from my career in chemistry that knowledge and practice of inorganic chemistry was far better then, than it is today.  Who knows what the Chinese and Russians do on the small scale of tube production in those countries?

Microphonics is just to do with the rigidity and resonance of the tube components and there will obviously be some tubes that were made by women on bad days.

Regarding Noise:
I have attached the test data produced by the sadly departed Steve Bench, a guy who really knew tubes.  The point to note is that one of the Chinese tubes had very low noise.  I find it very hard to believe that some tubes sound warmer or some other such description.  If that were true, then some tube manufacturer would have measured it and marketed it back in the day when there was money to be made.  If warmer meant lacking top end then it would have shown up in a frequency response test.  If it meant more or less distortion harmonics then it would have shown up in test data.  In fact, tube ads I've seen only ever emphasised  their power, lifetime, reliability and lack of hum and microphonics.

DaveP

 

[emrr]

I think that audiophools buy these expensive tubes because they are basically scared.  They have so little actual experience of making gear that they throw money at it to make themselves feel confident.

This is the predominant driver.

Steve's comments about heater voltage are interesting, and match some of what was done purposefully in some equipment. 

I'll refer back to others in the know who've tried to make the 417A work, and discovered they all seem to go noisy after 6 months despite their apparent promise. 

12AU7 is interesting, this matches Ian's comments. 

 

[Gene Pink]

... Look for the triangle on the bottom.

Triangle? These have a diamond, if that means anything to anyone.

Gene

 

[mjrippe]

Triangle? These have a diamond, if that means anything to anyone.

Gene

I think CJ misspoke there.  A diamond imprint on the bottom is one sign of a genuine Telefunken.

 

[CJ]

it's two triangles without the common wall. 

 

[Gene Pink]

I think CJ misspoke there.  A diamond imprint on the bottom is one sign of a genuine Telefunken.

Makes sense. Thanks to all, for the education.

I got this Harmon Kardon sometime in the mid 80's, don't remember where or how. It tested good at the time, according to the masking tape I put on it back then. Ever try to get 30 year old masking tape off a faceplate?  I started the "softening process" yesterday, shouldn't take much more than a month to clean it off.

Gene

 

[DaveP]

I for one agree that paying crazy money for a tube is foolish as is many things but, I also know I hear a difference,heck my wife does, in the affects of different tubes in a piece of gear. If that means that someone didn't build this unit perfectly to spec, or the tubes weren't biased  to make every single tube in the world sound exactly the same in it, then that's understandable. Not sure I've ever come across anything like that?

I'll be perfectly honest here, apart from microphonics and hum and noise, I have never heard the difference between tubes in the last 40 years, they amplify period.

Some tubes have more intermodulation distortion than others as described in RDH4 with 6J7's versus 6SJ7's.  I think that tubes with lower gm might have less IM distortion, this may be what people are hearing and it should be measurable.
I tested a dozen EF85's last year and found that the Mullards had higher gain than the Siemens, this can shift operating points and may influence the sound.

Since I tested some Thiele- Small parameters this summer, I realised that speakers can change with altitude and humidity.  Although altitude does not change on a daily basis, humidity does.  That variable would have to be accounted for in any meaningful test.  It may also be important to get a good nights sleep before conducting such a test, we may be looking at such tiny differences.  Like most things, it's wise to follow the money.  If people can jack up the price  just by using colourful prose, they will.  Buyer beware.

DaveP

 

[ruffrecords]

So all tubes should sound alike in a properly designed piece of gear? Is this a theory or happening ?

Depends what you mean by properly. All tubes vary due to manufacturing tolerances. Tubes with the same spec from different manufacturers have different internal constructions. So the raw performance of different tubes of the same type does vary.

Audiophiles abhor negative feedback so most of the circuits they use do not have it. In that case you can and will hear differences between tubes because they are the most inconsistent part of the signal chain. That is why Mr. Black invented negative feedback and it is a good job he did because semiconductors without it sound terrible.

Negative feedback (NFB) reduces the variations of all parameters  of all  active components in a circuit by the amount of feedback. So if you have 20dB of NFB all the variations in tube parameters will be reduced to one tenth their original values. Unless the tubes are badly biased, and /or coupling capacitors are wrongly sized,  with that amount of NFB  at normal line levels you will need platinum plated golden ears to hear any difference between tubes.

Power tubes is a whole different ball game. These are often operating close to their limits which amplifies (sic) differences in their characteristics.

Cheers

Ian

 

[rmburrow]

I don't think  said that. I have spent a lot of time looking at different brands and versions of 12AX7 trying to find ones that are consistently quiet. I am lucky that my Lindos test set includes a loudspeaker so you can listen to the noise. The other thing you can do is check for microphonics by gently tapping the tube. You can often hear the grid vibrating in the worst examples and you can even hear pinging noises when you turn the tube off as the metal parts cool and contract.

Bottom line is I have found only one 12AX7 which has consistently low noise AND low microphonics and that is the 12AX7WB by Sovtek. There are plenty of other low noise ones but most of them have poor microphonics like the 12AX7LPS for example. For tube mic pres it has always been necessary to select tubes for lowest noise but I have found 90% of the Sovteks are good enough
If I find time I will look if I can find all the noisy (brand new) ones I tested and send them to you.

So you are right, there are plenty of poor current production 12AX7 tubes but the NOS industrial versions I have always found to be even worse.  One problem with buying them is you don't really know their history but it is certain that the many I have tested have always been worse than the Sovteks.I would be happy to test them out. Any information that improves our chances of obtaining better tubes must be a good thing.

Cheers

Ian

Tried any NOS mid 1960's RCA 7025's or 5751's?

 

[ruffrecords]

This makes sense to me and would explain why I didn't hear much of a difference in tubes with this TL Audio V1 valve box I have/ Had.??

It was just a straight input to output,no gain, solid state box with a tube stage.The difference between the stock Edicron or Sovtek was minor to a 59 Bugle Boy. When the box was pushed with level differently there was a more noticeable difference but that brings in all kinds of other things obviously.


Thanks for the explanation!

(Un)fortunately, there are lots of designs around that deliberately operate a tube out of its comfort zone in order to entice some kind of mystical 'tube tone' out of it and in doing this you can get all sorts of odd distortion effects, some pleasing and some not. Since they involve operating the tube under conditions it was never designed to operate under, it is not surprising that different tubes can sound radically different but that is not a proper design in my book.

Cheers

Ian

 

[ruffrecords]

Tried any NOS mid 1960's RCA 7025's or 5751's?

Yes, and without exception they have been noisy.

The 12AX7 and its variants is a very good preamp tube. It actually has a very low inherent distortion level. Used as intended, in a low level preamp, the distortion  it produces is really tiny and pales into insignificance compared with the distortion produce by the later higher level stages. With NFB its contribution to the overall tone of the preamp is insignificant. Its major contribution is to the noise of the preamp because it is invariably the first stage. That is where you really notice the differences between brands and between NOS and new. I have found the best of the so called NOS tubes I have bought to be much noisier and more  microphone than the best new production tubes.

In devices like guitar amps, 12AX7s are often used at levels they were never intended for, so tube variations do tend to show up more. Tube rolling is fine for guitar amps but should be unnecessary in proper preamp designs.

Cheers

Ian

 

[Recording Engineer]

I'm digging what you're saying Ian, but there's something about the this proper-thing I keep getting stuck on... I thought it was mostly all the IMproper part that really has gotten people getting crazy into this whole tube thing... Really, I suppose I'm not just talking tubes either.

 

[bluebird]

something about the this proper-thing I keep getting stuck on...

This whole forum was started on "this proper-thing". Its basis was to take some of the voodoo out of recording equipment, so super expensive gear could be actually attainable for an average recording engineer who was willing to pick up a soldering iron and do some reading.

We often use the word "audiophool" around here because there is little test evidence to support the claims of that community when it comes to what they hear. I will say that it is an art form in itself, It takes a lot of imagination to come up with some of those descriptions! Its definitely a form of poetry, but not what this place is about.

Since tubes wear and change over time, unlike transistors, I think its really hard to definitively say an NOS GE 12AX7 sounds like this or that. I have boxes of old preamp tubes and rarely if ever, get the same reading from my tube tester of the very same tube and brand. So that means two GE 12AX7's from the 60's are not going to measure the same current draw when APPLYING THE SAME BIAS AND PLATE VOLTAGE. 

So depending on the loading of the circuit they are in, they may have different distortion characteristics because one is stronger than the other. So two new JJ or Sovtek tubes are going to be different as well.
You can describe the difference you hear in any poetic way you want... But its just poetry.

 

[ruffrecords]

I'm digging what you're saying Ian, but there's something about the this proper-thing I keep getting stuck on... I thought it was mostly all the IMproper part that really has gotten people getting crazy into this whole tube thing... Really, I suppose I'm not just talking tubes either.

You are both right and wrong. Certainly there are a lot of people making devices that make tubes deliberately make extraordinary levels of distortion (Thermionic Culture for instance and the Roots console). Many of these supposedly tube products have mostly semiconductor signal paths with the odd tube thrown in literally 'for effect'.  But to my mind these are in the same class as the Aphex Aural Exciter which was transistors made to do the same thing; they are effects. So in that sense you are right.

However, the real beauty in tubes comes in more subtle packages like the REDD 47, the V74, the V76, the UA610 and so on. Here the designers were trying to get the best out of the available tubes but in doing so they imparted some kind of magic to the sound. Whether it was the incredible headroom, the iron at the input and the output,  the class A operation, the gentle rise of second harmonic distortion with increasing level or a combination of them all, they created something has has never been bettered. That is what I call proper design.


As you probably know, Holger Classen built a 12 channel tube mixer based on my designs (which was inspired by the V76) and it went head to head with an SSL4K. As Jensenman said in his post:

Gents, this desk sounds even better than it looks. Holger had it sitting next to a SSL4k in tiptop shape, recording a jazz band. The SSL owner went dead pale after switching tracks from his desk to the Krässemaschine. I yet have to hear a better sounding desk.

Cheers

Ian

 

[Recording Engineer]

This whole forum was started on "this proper-thing".

I'm actually agreeing, but I'm actually even more confused with the above.

 

[bluebird]

That's all right, Most people don't understand what I say, but I keep trying!

 

[emrr]

A great test to make you crazy is hooking up a 12AX7 in this sort of cathode follower combo:

https://www.ampbooks.com/mobile/classic-circuits/fender-bassman-voltage-amp/

Note what the voltage relationships are supposed to be, then measure the voltages at various points.  Then do the same thing with 6 more 12AX7's.  They are usually all over the place based on the current draw of the various halves. 


Most people haven't heard a lot of different tube circuits.  They mistake the overall circuit sound including the transformers for the tube sound.  I've found the transformers in vintage preamps are the majority of the signature.  The amount of feedback from a practical point affects the qualities of distortion, and that adds another noticeable layer, but we don't try to operate in that territory; it still doesn't sound good, outside of the occasional special effect usage.    In most cases you could swap the transformers from one circuit type to another and the majority of the sound would move with them.  Not all, but most. 

 

[Recording Engineer]

Ian,

That absolutely does help clarify. Thank you. I suppose just as with anything, even "proper" is hard to define with absolute certainty.

For example, while the SSL 4K was designed by a team of very exceptional designers with what I assume was a rather large R&D budget, I know plenty of people wouldn't consider that signal-chain proper.

By the way, I used to have a SSL 6048 and loved it... Still, now days I'd rather a Krassemaschine! I'd consider that more proper!

 

[DaveP]

Most people haven't heard a lot of different tube circuits.  They mistake the overall circuit sound including the transformers for the tube sound.  I've found the transformers in vintage preamps are the majority of the signature.  The amount of feedback from a practical point affects the qualities of distortion, and that adds another noticeable layer, but we don't try to operate in that territory; it still doesn't sound good, outside of the occasional special effect usage.    In most cases you could swap the transformers from one circuit type to another and the majority of the sound would move with them.  Not all, but most.

Plus 1

DaveP

 

[ruffrecords]

I will have to take the experts opinions on this.

At least it gives me more insight in to why I hear a difference when swapping out tubes. I have a modified hammond amp that is now a push pull type vari mu limiter.....kinda DOC ish.... and, changing the 6sn7s around makes a very noticeable difference.  I did actually settle on two different brand 6SN7s for the outputs so maybe I just tested them by ear until I found two that behaved alike and brought out the best sound to me.... definitely didn't test anything beyond that.....

Very interesting stuff.......

Thanks so much for the insights!!!!!

Vari-mu is an interesting case because there is no opportunity to place NFB around the most important stage, the vari-mu itself and many products of this type have no NFB at all. It is also another case where the tube is used over a much wider range of operating values than normal. Both these factors mean there is much more chance of hearing tube differences. Probably one of the few examples where 'proper design' is not possible.

Cheers

Ian