Current in production 6072/ 12AY7 tubes

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
As a point of interest re: EH 6072 tube noise, I once tested 250 of them. They fell into a standard bell-shaped curve. 4 were unimaginably quiet and almost without microphonics - unless I really smacked them. A total of 14 were "superior grade" in every way. 10 went straight to the trash can. Some of those were so microphonic, I could hardly cough in the testing room without hearing a high frequency "ping" from the tube. At least half were true "Bogies", they met expectable standards for general application. In the end, I decide that about 20 were what I'd call "microphone quality". I have no idea if the current productions are similar, but "there you have it".
 
I hear you, I will test them in circuit. I have just bought a GE6072 5 star and EH 12AY7, now I need a JJ and a TAD for a shootout. I am curious about the results.
Did you do this test? Knowing results would be very interesting if GE 6072s were new or in good enough condition. I did some limited tests where tubes were mostly Tesla and EH, the rest were mics where conditions are very different from amps.
 
As a point of interest re: EH 6072 tube noise, I once tested 250 of them. They fell into a standard bell-shaped curve. 4 were unimaginably quiet and almost without microphonics - unless I really smacked them. A total of 14 were "superior grade" in every way. 10 went straight to the trash can. Some of those were so microphonic, I could hardly cough in the testing room without hearing a high frequency "ping" from the tube. At least half were true "Bogies", they met expectable standards for general application. In the end, I decide that about 20 were what I'd call "microphone quality". I have no idea if the current productions are similar, but "there you have it".
Were "microphone quality" EH 6072 ever used in mics and compared to nos? 250 tubes is a lot, you might want to check Ian's white paper about tubes.
EH EF86 doesn't work that well in mics, ime not even their EF806G costing 50eur sounds as good as nos in this application. I'm buying nos for this reason, several tubes here or there aren't that expensive from good sellers that will test them.
All three are doing fine in preamps as V1, i got only one a little noisy EF86 pair from 2011, Ia was also lower than in tubes they were replacing... They were mechanically inferior to other EH EF86 used before, so maybe rejects from larger batch or monday morning in EH factory.
 
Last edited:
Did you do this test? Knowing results would be very interesting if GE 6072s were new or in good enough condition. I did some limited tests where tubes were mostly Tesla and EH, the rest were mics where conditions are very different from amps.
Yes, I have done some completely unscientific tests.

In my Champ in V1 position all tubes played flawlessly. I had the GE, TAD and EH on test, JJ was not available at the time. There were differences but they were not that huge.

In a Hampton Pre in V1 position the differences were bigger. The GE had the least noise, EH was ok, the TAD tube was the taillight. There were also differences in sound, not dramatic but audible. I didn't like the two EH tubes as much.

In a microphone, a modified tbone SCT800 with ELA M251 style circuit the differences were significant, the GE was good usable, I was lucky with my purchase. This tube is still in use, all fine so far. With the EHs and TAD I had no luck. Noise was too high, the TAD had microphonics problems in this application, the EHs didn't sound great either.

Of course, all this is not meaningful due to the small number of test subjects and partly also very subjective.

These experiences were the reason why I chose other types of tubes for (not only) microphones now.

What counts for me now are the availability and the price, because I want to select. This is not so easy with expensive NOS darlings and even the current tubes from EH, TAD and JJ cost partly already 30€ a piece.

My favorites are now "uncool" NOS pentodes wired as triodes. Here you get really a lot for the money.
 
Last edited:
What about 6386 for compressors even JJs are expensive. WESTERN ELECTRIC could make good ones I'd bet
That would be nice to have a 6386 from current production that actually works.

But I doubt if this is really realistic, because the possible market is numerically very limited to revive such a tube.

I would love to be wrong...
 
I first heard about long plates having lower distortion from audiophiles on DiyAudio or some other forum, your tests were of course much better. EH 6CG7 has pretty long plates compared to other LPs like RCA, will check your paper about types tested..
If memory serves, the most common short plate type was by a Japanese maker (I forget the manufacturer bit I remember being surprised that Japan made tubes. As you would expect they were extremely similar).
Do you think gain in twin line amp would change more than ~5dB if ECC83 was replaced with 12AY7?
Hard to say. The characteristics of the two tubes are quite different and there is dc feedback from the output SRPP stage to the cathode of the ECC83 so you would need to adjust its cathode resistor to maintain this. Other than that, the open loop gain would be reduced because of the lower mu of the 12AY7. Furthermore, the gain setting resistor increases the stage gain of the ECC83 stage as the closed loop gain is increased in order to reduce variations in loop gain and hence improve stability. lAs it stands, there is about 20dB of negative feedback at the highest gain setting (40dB) which would be reduced by about 6dB with the 12AY7 so I would expect distortion and output impedance to approximately double. 20dB of NFB will reduce circuit component variations ten fold so closed loop gains normally vary by no more than 1dB. With less NFB this will be greater but I would be surprised if it is more than 2dB.
Cheers

Ian
 
I select lot of 6072/12ay7 for mics, hifi/studio equipments, the old GE's are still superior in sound for mics, the EH/GT russian 12ay7 sounds bit dark/"pale" good for taming bright capsules(still a bit lifeless) JJ is thin and bright.If no budget for GE better mod to ecc81.
 
Do you think gain in twin line amp would change more than ~5dB if ECC83 was replaced with 12AY7?
Well, I "think" this is the math: the Mu of a ECC83/12AX7 is 100. The Mu of a 12AY7 is 44. If it were 50, then the 12AY7 would have half the voltage gain of an ECC83/12AX7 tube. BUT, because we're talking voltage gain, then having only a bit less than half the Mu would mean you'd be down a bit more than 6dB. Actual circuit conditions will, of course, come into play and shift your results a tad (or more), one way or the other.
 
Well, I "think" this is the math: the Mu of a ECC83/12AX7 is 100. The Mu of a 12AY7 is 44. If it were 50, then the 12AY7 would have half the voltage gain of an ECC83/12AX7 tube. BUT, because we're talking voltage gain, then having only a bit less than half the Mu would mean you'd be down a bit more than 6dB. Actual circuit conditions will, of course, come into play and shift your results a tad (or more), one way or the other.
Putting a medium mu tube in the circuit designed for a hi mu like the 12AX7 will move the plate voltage around to a place where you could run into premature lopsided clipping. Noise is hard to predict, but in this scenario it will probably be higher because the plate resistor is higher. If the grid is in a hi z state like a 1meg R the tube will be much noisier than if it a low z or shorted to ground.

Large plate tubes have lower frequency mechanical resonances which may accentuate that resonant frequency. Some tubes because of their design are more resonant. 6DJ8s have a lot of hi frequncy audible microphonics. They sing like a flock of birds.
 
Yes, I have done some completely unscientific tests.

In my Champ in V1 position all tubes played flawlessly. I had the GE, TAD and EH on test, JJ was not available at the time. There were differences but they were not that huge.

In a Hampton Pre in V1 position the differences were bigger. The GE had the least noise, EH was ok, the TAD tube was the taillight. There were also differences in sound, not dramatic but audible. I didn't like the two EH tubes as much.

In a microphone, a modified tbone SCT800 with ELA M251 style circuit the differences were significant, the GE was good usable, I was lucky with my purchase. This tube is still in use, all fine so far. With the EHs and TAD I had no luck. Noise was too high, the TAD had microphonics problems in this application, the EHs didn't sound great either.

Of course, all this is not meaningful due to the small number of test subjects and partly also very subjective.

These experiences were the reason why I chose other types of tubes for (not only) microphones now.

What counts for me now are the availability and the price, because I want to select. This is not so easy with expensive NOS darlings and even the current tubes from EH, TAD and JJ cost partly already 30€ a piece.

My favorites are now "uncool" NOS pentodes wired as triodes. Here you get really a lot for the money.
In my unscientific tests nos triodes in mics sounded much better than new production, same with pentodes i use more often.
Some experienced mic builders mentioned new production pentodes don't take very high grid resistors used in mics, i have to check this one more time to be sure.
In preamps or compressors it depends ime, it is worth checking slightly used but tested nos from good sellers because this tubes last a long time, especially in hot biased outputs.
Next thing to try is nos at inputs but i'm not willing to pay much more for slight differences, if the rest like microphonics or noise is ok. "Weird" pentodes wired as triodes are certainly something to try in diy due to low prices and possibly high quality.
 
If memory serves, the most common short plate type was by a Japanese maker (I forget the manufacturer bit I remember being surprised that Japan made tubes. As you would expect they were extremely similar).

Hard to say. The characteristics of the two tubes are quite different and there is dc feedback from the output SRPP stage to the cathode of the ECC83 so you would need to adjust its cathode resistor to maintain this. Other than that, the open loop gain would be reduced because of the lower mu of the 12AY7. Furthermore, the gain setting resistor increases the stage gain of the ECC83 stage as the closed loop gain is increased in order to reduce variations in loop gain and hence improve stability. lAs it stands, there is about 20dB of negative feedback at the highest gain setting (40dB) which would be reduced by about 6dB with the 12AY7 so I would expect distortion and output impedance to approximately double. 20dB of NFB will reduce circuit component variations ten fold so closed loop gains normally vary by no more than 1dB. With less NFB this will be greater but I would be surprised if it is more than 2dB.
Cheers

Ian
Below are both papers about 6CG7 for others who didn't see them yet.
So changing tube with dc feedback between stages won't do any good without redesign, the best i can do is check older threads and RDH for understanding how NFB is applied here.
TLA would be nice circuit with for the local forum where diy beginners asked for something with NFB + more distortion than One Bottle 6SN7 preamp designed by NY Dave in 2006 (pretty clean no nfb design built by several users). I will check appropriate thread for more information about pcb because euroboard would be too hard for them... Thank you
 

Attachments

  • mufollowerdistortion.pdf
    112.8 KB · Views: 0
  • 6CG7mufollowerdistortion.pdf
    222.3 KB · Views: 0
we charge 35 bucks for EHX 12ax7a's nowadays because of the Ukraine thing which makes going to NOS a little more inviting,
Or preamp V1 and mic EH 5751 or JJ ECC83 tubes tested by Manely for $100!
What do you think about claims that ECC83 sounds bad in pro audio? Wrong gain staging or too much gain for V1 are two reasons i could find, maybe bad production with new tubes but they say the same about nos. 5751 looks like easy fix.
 
What do you think about claims that ECC83 sounds bad in pro audio? Wrong gain staging or too much gain for V1 are two reasons i could find...
I've never heard a 12AX7 in a microphone circuit that sounded as good as 12AT7 or 12AY (6072). I experimented quite a bit, and could always get better results, often without making significant changes to plate/grid/cathode resistors. Admittedly, I have not heard every 12AX7-style mic there is, let alone swapped each type's tube to see what could be accomplished. But, like I said, I could always get a better sound from a microphone by using something other than a 12AX7. And, yes, I know many excellent sounding pieces of audio gear that use 12AX7s. They just aren't microphones.
 
I've read that there's a certain coloration/tonality that the 12ax7 design brings to the table. From a thread on the gear page (in the context of guitar amps, of course):

The 12ay7 is like playin' your guitar,only louder.
The 12ax7 is like playin' through an amp (colored).

The above is kind of simplistic, but I tend to agree :)
 
One thing you have to t
I've read that there's a certain coloration/tonality that the 12ax7 design brings to the table. From a thread on the gear page (in the context of guitar amps, of course):

The 12ay7 is like playin' your guitar,only louder.
The 12ax7 is like playin' through an amp (colored).

The above is kind of simplistic, but I tend to agree :)
Things you all have to take into account is the circuit for each tube.. 12AX7 -Highest gain audio tube, Plate resistor typ 220K, very low current approx 1ma, reduced hf response. 12AY7 - Medium gain audio tube, Plate resistor 20K to 50k, wider bandwidth, higher current approx 4 ma.

Assuming you are not just replacing and AX with an AY, and each with their proper circuit values, the AY has rolloff much higher f, and thus phase shift inside the audio band is reduced. And they are not typically used that often as far as I know. The AU7, for instance, is used where you need some power, like driving a reverb spring or as a low impedance phase splitter to drive the output tubes.

The thing that makes these kind of discussions difficult is the tube is only 1 element of the circuit. With a 50 to 100 other parts in the circuit there really isn't an absolutely better tube for every application.

Each is made for a specific purpose. If you look at the spec sheets the AY is a low microphonic design. Not only is it built to vibrate less internally, its lower gain will amplify its self noise, including microphonics, less. Maybe that leads to a sense of transparency.

Like in golf, a 9 iron is no "better" than 5 iron or a driver. Each is designed to be optimum for their intended use. You may be better at swinging a 9 iron than a driver and feel more comfortable but that's operator error. There's an analogy in there, I'm sure of it.
I don't play guitar professionally so I don't get caught up in the AX controversy. As a hifi tube it's usually used as a phono pre.
 
Or preamp V1 and mic EH 5751 or JJ ECC83 tubes tested by Manely for $100!
What do you think about claims that ECC83 sounds bad in pro audio? Wrong gain staging or too much gain for V1 are two reasons i could find, maybe bad production with new tubes but they say the same about nos. 5751 looks like easy fix.
I have no idea why people think this. I tested quite a few 12AX7 tubes many years ago and I found their intrinsic distortion is extremely low. Any design that makes it sound bad is a problem with the designer not the tube.

Cheers

Ian
 
Back
Top