Tubes4HiFi VTA 12AU7 vs 6SN7

[ceylon]

I have the ST-120. A couple of years ago, I ordered the octal board for it and put it in. I have a set of old russian 6H8C tubes that I put in. I will never go back to the 12AU7 board.

can you please shoot me some photos of the 6N8c u use? i would like to buy some but dont know which oes to choose.

Since the 6H9C is the same as a 6SL7, which is the 8 pin equivalent of the 12AX7,

as a high mu voltage gain tube, its a predecessor to the 12AX7.
but 6SL7 is not the octal equivalent to 12AX7. it specs differently across the board and is not a drop-in substitute.
any such substitution is circuit dependent.

 

[gadget73]

yeah, its ballpark similar, but not a dead ringer. In some circuits it will drop in just fine, others maybe not. Feedback EQ in phono preamps is gain-sensitive so probably not a great idea. Passive EQ it likely will not care though.

If you want octals with the original gain though, you need two 6SF5. Its as close as it gets in an octal, but its a single triode. I don't think there are any octal twin triodes higher than the 6SL7.

I'd be more curious what the operating parameters are on that circuit if its making tubes go microphonic. I've had microphonic tubes but I've never had anything that took good quiet tubes and made them microphonic.

 

[ceylon]

yeah, its ballpark similar, but not a dead ringer. In some circuits it will drop in just fine, others maybe not. Feedback EQ in phono preamps is gain-sensitive so probably not a great idea. Passive EQ it likely will not care though.

If you want octals with the original gain though, you need two 6SF5. Its as close as it gets in an octal, but its a single triode. I don't think there are any octal twin triodes higher than the 6SL7.

I'd be more curious what the operating parameters are on that circuit if its making tubes go microphonic. I've had microphonic tubes but I've never had anything that took good quiet tubes and made them microphonic.

2C52 is an octal twin triode of 100mu, but i dont know the rest of the specs.

 

[Mr Boochie]

I've had microphonic tubes but I've never had anything that took good quiet tubes and made them microphonic.

I have it happen more than you would think. When leaky coupling caps are allowing significant VDC onto the signal grid (>1-2VDC).

Early signs include noise and microphonics.

 

[gadget73]

2C52 is an octal twin triode of 100mu, but i dont know the rest of the specs.

12.6v to light it would be the biggest issue in most things

Higher transconductance but not horribly so. 1400 vs 1250, so it would want a pinch more bias for the same idle current.

 

[Mr Boochie]

Another nice 9-pin mini, medium-mu, tube is the 5687WA. I like mil-spec, TungSols, and Raytheons.

 

[gadget73]

I have it happen more than you would think. When leaky coupling caps are allowing significant VDC onto the signal grid (>1-2VDC).

Early signs include noise and microphonics.

Hm, possible. I usually change or at least check the caps in my stuff and make sure the voltage and current is reasonable, maybe that's why I haven't had this happen. The noisy tubes I've had came to me that way.

 

[Mr Boochie]

Hm, possible. I usually change or at least check the caps in my stuff and make sure the voltage and current is reasonable, maybe that's why I haven't had this happen. The noisy tubes I've had came to me that way.

Also, in the 1990s through the early 2000s, CJ amps really pushed their gain stages hard (like 300VDC on the plates). Those tubes often went noisy or microphonic. A friend went through several sets of mil spec or fancy 6922/E88CC tubes until he figured it out.

Some 1990s-era guitar amps had the same issue. You just couldn't really get around that hot setup without changing the overall sound too much.

It was a trend at the time. I guess to compete with hot solid-state units.

BTW... the friend sold the CJ and bought tubed Mac stuff. No more eating tubes.

 

[chris35745]

Is there room to just use socket adaptors? I have a 6SN7 to 12AU7 socket adaptor I've used. You can get 6SL7 to 12AX7 adaptors. No need to make modifications.

I do have socket adapters for 6SN7/6SL7 to 12AU7/12AX7. I already rebuilt my phono pre because I thought there was a problem with either the tubes or caps/resistors. I built a brand new board with all American parts. Turned out it was an RCA jack on the back of my phono pre giving problems. I still have the original board from my PH14. I decided to get another chassis for it and use the original board and use the adapters and use 6SL7s instead of 12AX7. Looking at some Russian 6H9Cs instead. The original PH14 chassis is to short to handle 6SL7s and adapters without drilling holes in the top or leaving the top off. Bought a nice piece of Morado wood to use as my chassis. Just need to buy the transformer from Dan.

 

[SS904]

Decided to revisit the heat concerns that were brought up regarding the LM334 CCS at the PI. I found some appropriate heat sinks and had some time this past weekend to put the amps back on the bench. I wanted to see if I could visualize heat related instability in the process.

I imaging the problem should show as an instability or amplitude imbalance between the two outputs.

I connected my scope to the output of each coupling cap and monitored for more than 30 mins with it running in to an 8 Ohm load and with a 1KHz signal injected. Pretty well rock solid for the entire time. I decided to tempt fate (I have some LM334s on hand in case fate won) and hit the CCS regulator with my small heat gun in short to moderate bursts trying to mimic the temps it may have been exposed to if I had put it on top of the PCB where it would have been almost touching the tube socket.

I fully appreciate my test methods may not have been detailed enough, but I feel I could not clearly replicate the issue looking at the output of the PI on the scope. I thought later that I may have noticed something if I had really expanded the vertical display and got a good look at the waveform peaks. CH1 is the CCS side and maybe we are seeing a few mV less. Still though, there was no change even with the heat gun.

I will say that it does not get very warm under the chassis. Maybe after several hours of use, but in the decent amount of time it was on the bench, things just got a little warm to the touch. All the heat is topside. Maybe a good move to mount the LM334 on the bottom regardless. I did end up installing the heat sinks. Can't hurt!

 

[Lavane]

I mounted them on the bottom of the boards in my M-125 mono blocks as well. Also the coupling caps, mica's, and that film cap.
Its good to know the LM334's are that stable with temperature changes. I've even read that temperature changes can cause slight channel imbalances. I never noticed any with my ST-70. I put the LM334 under the chassis for the same reason you did, heat. But I also figured it would also keep both at the same temperature. But it seems that isn't really an issue.

 

[spantou01]

I have applied that fix to a project I had using the LM334. It works.

I have a few questions for you. Have you installed the Miller modules ($30 each when I last looked) or did you added the parts 15 cents worth of parts to your LM334? Are you also doing ~8.4 ma on the two 12au7 triodes like the VTA-70 PCB? If not what resistor values have you used in the two tempco stabilizer resistors?

I am currently populating two VTA-70 boards; a 12AU7 and 6SN7 each. I thought I should include the CCS temperature coefficient modification while I have a clean build, but I’m not sure how derive the right values of the two resistors needed. The standard set resistors (R27 & 28) for the LM334 in these two boards are 10 Ohm and 8.2 Ohm respectively resulting in a 10.5 V elevation of the cathodes.


The fellow from the link (ieLogical M-125 saga) that @nerdorama provided shows the single R28 is replaced by the two parallel resistors with values 22 and 121 Ohms for his M125. He then shows the cathodes sitting a 12.5 V. Is this the standard for M125s, of did his resistor choices change the CCS current?

Has anyone else did this modification for their VTA-70s, and what resistor values have you used? Instead of the 1N4001 I plan to use a 1N457 diode recommended in the LM334 application notes.

 

[spantou01]

Decided to revisit the heat concerns that were brought up regarding the LM334 CCS at the PI. I found some appropriate heat sinks and had some time this past weekend to put the amps back on the bench. I wanted to see if I could visualize heat related instability in the process.

I imaging the problem should show as an instability or amplitude imbalance between the two outputs.

I connected my scope to the output of each coupling cap and monitored for more than 30 mins with it running in to an 8 Ohm load and with a 1KHz signal injected. Pretty well rock solid for the entire time. I decided to tempt fate (I have some LM334s on hand in case fate won) and hit the CCS regulator with my small heat gun in short to moderate bursts trying to mimic the temps it may have been exposed to if I had put it on top of the PCB where it would have been almost touching the tube socket.

I fully appreciate my test methods may not have been detailed enough, but I feel I could not clearly replicate the issue looking at the output of the PI on the scope. I thought later that I may have noticed something if I had really expanded the vertical display and got a good look at the waveform peaks. CH1 is the CCS side and maybe we are seeing a few mV less. Still though, there was no change even with the heat gun.

I will say that it does not get very warm under the chassis. Maybe after several hours of use, but in the decent amount of time it was on the bench, things just got a little warm to the touch. All the heat is topside. Maybe a good move to mount the LM334 on the bottom regardless. I did end up installing the heat sinks. Can't hurt!

I read your test with interest. It is good to see, as @Lavane stated, that your ‘scope did not reveal any visible aberrations on the phase inverter traces at 1kHz when the LM334 temperature was changed. But we know that the LM334 CCS current changes with temperature - the spec sheet says 0.33%/°C, so we know the operating conditions of the PI tube will change significantly from cold to hot.

I guess what you found was that at 1 kHz you can’t see the effects of temperature change on a scope. The big question is, can you hear it? Can you hear it in complex musical passages with a gazillion fundamentals, overtones, etc. that are perceived as music with spatial cues? Based on hundreds, if not thousands of VTA amplifiers out there with happy owners I conclude the answer is no, most listeners can’t near the unwanted variations in music, caused by changes in tube operating conditions due to variations in CCS current due to temperature changes. But I’m still going to try to apply the tempco fix to my VTA board.

I also agree with your installation of heatsinks. The spec sheet says the max current for a LM334 is 10 ma IIRC. So, at 8.5 ma that the VTA board are running it we are too close to max for my comfort. But again, all those happy VTA owners out there don’t seem to have problems with failing CCS.

 

[SS904]

I have to be honest. I am trying to understand what I am supposed to hear! I come from the guitar and bass amp world and distortion in that realm normally refers to clipping type distortion. My amps, the VTAs (70 and MK3s) and my home built SET and SEUL examples provide wide clean soundstages and balanced sound top to bottom. There are differences between all of them, but they all sound very good to me and I enjoy listening to music on them. I'm often wondering what people hear as distortion in HiFi....

 

[ceylon]

I do have socket adapters for 6SN7/6SL7 to 12AU7/12AX7. I already rebuilt my phono pre because I thought there was a problem with either the tubes or caps/resistors. I built a brand new board with all American parts. Turned out it was an RCA jack on the back of my phono pre giving problems. I still have the original board from my PH14. I decided to get another chassis for it and use the original board and use the adapters and use 6SL7s instead of 12AX7. Looking at some Russian 6H9Cs instead. The original PH14 chassis is to short to handle 6SL7s and adapters without drilling holes in the top or leaving the top off. Bought a nice piece of Morado wood to use as my chassis. Just need to buy the transformer from Dan.

will you please post a few photos of the 6N8C youve been using in your ST-120? thanks

 

[Lavane]

I read your test with interest. It is good to see, as @Lavane stated, that your ‘scope did not reveal any visible aberrations on the phase inverter traces at 1kHz when the LM334 temperature was changed. But we know that the LM334 CCS current changes with temperature - the spec sheet says 0.33%/°C, so we know the operating conditions of the PI tube will change significantly from cold to hot.

I guess what you found was that at 1 kHz you can’t see the effects of temperature change on a scope. The big question is, can you hear it? Can you hear it in complex musical passages with a gazillion fundamentals, overtones, etc. that are perceived as music with spatial cues? Based on hundreds, if not thousands of VTA amplifiers out there with happy owners I conclude the answer is no, most listeners can’t near the unwanted variations in music, caused by changes in tube operating conditions due to variations in CCS current due to temperature changes. But I’m still going to try to apply the tempco fix to my VTA board.

I also agree with your installation of heatsinks. The spec sheet says the max current for a LM334 is 10 ma IIRC. So, at 8.5 ma that the VTA board are running it we are too close to max for my comfort. But again, all those happy VTA owners out there don’t seem to have problems with failing CCS.

I haven't seen a post yet on any audio site about the LM334's failing in the VTA amps. Its mostly rectifier tube failures.

 

[Lavane]

will you please post a few photos of the 6N9C youve been using in your ST-120? thanks

The 6N9C is a sub for 6SL7's. Thats a lot more gain than is needed in a ST-120. A 6SN7 or the Russian 6H8C would he better.

 

[ceylon]

The 6N9C is a sub for 6SL7's. Thats a lot more gain than is needed in a ST-120. A 6SN7 or the Russian 6H8C would he better.

typo.

 

[gadget73]

guess what you found was that at 1 kHz you can’t see the effects of temperature change on a scope. The big question is, can you hear it? Can you hear it in complex musical passages with a gazillion fundamentals, overtones, etc. that are perceived as music with spatial cues? Based on hundreds, if not thousands of VTA amplifiers out there with happy owners I conclude the answer is no, most listeners can’t near the unwanted variations in music, caused by changes in tube operating conditions due to variations in CCS current due to temperature changes. But I’m still going to try to apply the tempco fix to my VTA board.

unless its just an absurd shift in operating conditions I doubt its audible, my concern would honestly be more about what the tube is doing. If it ends up shifting to a point where its running at excessive current, that will be bad for tube life.

 

[Lavane]

I haven't had any issues so far. My VTA ST-70 has been using the same driver tubes for years. Two of them since I built it around 2010 or 2011. The third one for about 7 or 8 years now. They still test strong.