The Reference 6SN7 Thread

[HTSkywalker]

A technical question
since 6SN7 tubes have dual triodes for left and right channels, what is the benefit of using 2 x 6SN7 tubes in an amp architecture ?

 

[L0rdGwyn]

A technical question
since 6SN7 tubes have dual triodes for left and right channels, what is the benefit of using 2 x 6SN7 tubes in an amp architecture ?

It depends on the circuit. If the two sections are run in series, now you have two voltage gain stages - double the gain, double the voltage swing into the output stage, more power if the output tube can swing the voltage without clipping.

If the sections are run in parallel, the voltage gain of a single section will be maintained, but the output impedance will be halved. This can be beneficial when driving the input capacitance of the output stage - a low pass RC filter is formed form the output impedance of the input stage and the input capacitance of the following stage. The lower the output impedance of the input stage, the higher the rolloff frequency of the filter, ideally outside the audio band, >20kHz. This is only an issue with those output tubes with relatively high input capacitance, 300B is a good example.

Yet another thing you can do with a dual section triode is run one section as a voltage gain stage in series with the second section as a cathode follower. Similar to the parallel scenario above, you will get the voltage gain of a single triode, but a much lower output impedance than running in parallel, due to the cathode follower buffer. A cathode follower does not add any additional voltage gain, it operates at near unity gain, gain of 1, but will actually be less than 1 in reality. The same benefits of low output impedance apply here. It will be much lower than a parallel section dual triode though, it can be approximated by 1/gm of the tube at its operating point (gm = transconductance).

You can even run both sections as cathode followers in parallel - each section's output impedance approximated by 1/gm as before, now divide that in half to account for both sections in parallel. The disadvantage here is there is no voltage gain! So this is uncommon unless you need a pretty hefty buffer to drive an output stage - it would require an additional voltage gain stage before it.

Yet ANOTHER thing you can do is run both triodes together in a SRPP configuration, which has similar benefits to the "voltage gain stage in series with cathode follower" setup, but I won't get into the details as it starts to get pretty technical.

Was there a specific amplifier you had in mind?

 

[HTSkywalker]

It depends on the circuit. If the two sections are run in series, now you have two voltage gain stages - double the gain, double the voltage swing into the output stage, more power if the output tube can swing the voltage without clipping.

If the sections are run in parallel, the voltage gain of a single section will be maintained, but the output impedance will be halved. This can be beneficial when driving the input capacitance of the output stage - a low pass RC filter is formed form the output impedance of the input stage and the input capacitance of the following stage. The lower the output impedance of the input stage, the higher the rolloff frequency of the filter, ideally outside the audio band, >20kHz. This is only an issue with those output tubes with relatively high input capacitance, 300B is a good example.

Yet another thing you can do with a dual section triode is run one section as a voltage gain stage in series with the second section as a cathode follower. Similar to the parallel scenario above, you will get the voltage gain of a single triode, but a much lower output impedance than running in parallel, due to the cathode follower buffer. A cathode follower does not add any additional voltage gain, it operates at near unity gain, gain of 1, but will actually be less than 1 in reality. The same benefits of low output impedance apply here. It will be much lower than a parallel section dual triode though, it can be approximated by 1/gm of the tube at its operating point (gm = transconductance).

You can even run both sections as cathode followers in parallel - each section's output impedance approximated by 1/gm as before, now divide that in half to account for both sections in parallel. The disadvantage here is there is no voltage gain! So this is uncommon unless you need a pretty hefty buffer to drive an output stage - it would require an additional voltage gain stage before it.

Yet ANOTHER thing you can do is run both triodes together in a SRPP configuration, which has similar benefits to the "voltage gain stage in series with cathode follower" setup, but I won't get into the details as it starts to get pretty technical.

Was there a specific amplifier you had in mind?

That was a nice and technically detailed description of dual 6SN7 architecture, the amp I already ordered is a cathode follower and exactly how you describe it above. Thanks, let me know your opinion on it please:

https://shenzhenaudio.com/products/...upgrade-hi-end-buffer-processor-for-cd-player

I checked with the manufacturer and this is the latest version with few upgraded components over the previous versions, one of which is the previous versions used to send voltage spikes upon switching off through RCAs (using bleeding resistors). As I will put it between a Cambridge CXN and a Mcintosch amp which both I don't want to risk burning.

 

[mayurs]

It depends on the circuit. If the two sections are run in series, now you have two voltage gain stages - double the gain, double the voltage swing into the output stage, more power if the output tube can swing the voltage without clipping.

If the sections are run in parallel, the voltage gain of a single section will be maintained, but the output impedance will be halved. This can be beneficial when driving the input capacitance of the output stage - a low pass RC filter is formed form the output impedance of the input stage and the input capacitance of the following stage. The lower the output impedance of the input stage, the higher the rolloff frequency of the filter, ideally outside the audio band, >20kHz. This is only an issue with those output tubes with relatively high input capacitance, 300B is a good example.

Yet another thing you can do with a dual section triode is run one section as a voltage gain stage in series with the second section as a cathode follower. Similar to the parallel scenario above, you will get the voltage gain of a single triode, but a much lower output impedance than running in parallel, due to the cathode follower buffer. A cathode follower does not add any additional voltage gain, it operates at near unity gain, gain of 1, but will actually be less than 1 in reality. The same benefits of low output impedance apply here. It will be much lower than a parallel section dual triode though, it can be approximated by 1/gm of the tube at its operating point (gm = transconductance).

You can even run both sections as cathode followers in parallel - each section's output impedance approximated by 1/gm as before, now divide that in half to account for both sections in parallel. The disadvantage here is there is no voltage gain! So this is uncommon unless you need a pretty hefty buffer to drive an output stage - it would require an additional voltage gain stage before it.

Yet ANOTHER thing you can do is run both triodes together in a SRPP configuration, which has similar benefits to the "voltage gain stage in series with cathode follower" setup, but I won't get into the details as it starts to get pretty technical.

Was there a specific amplifier you had in mind?

Very informative. Thank you. Can you please tell what configuration does a Feliks Euforia operate. Thanks in advance. I always wondered why some OTLs like DV, BH Crack only had one driver tube while my Feliks Euforia has a pair of these.

 

[L0rdGwyn]

That was a nice and technically detailed description of dual 6SN7 architecture, the amp I already ordered have is a cathode follower and exactly how you describe it above. Thanks, let me know your opinion on it please:

https://shenzhenaudio.com/products/...upgrade-hi-end-buffer-processor-for-cd-player

I checked with the manufacturer and this is the latest version with few upgraded components over the previous versions, one of which is the previous versions used to send voltage spikes upon switching off through RCAs (using bleeding resistors). As I will put it between a Cambridge CXN and a Mcintosch amp which both I don't want to risk burning.

Well, it's hard to glean much about the design without seeing a schematic, but it sounds like they are running the two 6N8P / 6SN7 as parallel cathode followers. This product is purely a buffer - it will provide an easy load for the input device and a low output impedance to drive the output device. It will also add some small degree of second harmonic distortion. There will be a small amount of voltage gain lost - the ideal cathode follower operates at unity gain, but in reality it might be gain of 0.85 or 0.9.

I will say that unless it is something very out of the ordinary, most gear is designed to have a sufficiently low output impedance a high enough input impedance such that they will pair well with just about anything you connect them to. But who knows! Maybe putting this cathode follower buffer in between the two will have some subjective benefit, it's hard to say, I guess you will have to try it out and see of course if you like the sound of second harmonic distortion, what some people call "tube sound", it might be an upgrade!

 

[L0rdGwyn]

Very informative. Thank you. Can you please tell what configuration does a Feliks Euforia operate. Thanks in advance. I always wondered why some OTLs like DV, BH Crack only had one driver tube while my Feliks Euforia has a pair of these.

Let me find a picture of the circuit, I'm certain a schematic won't be available...

 

[L0rdGwyn]

@mayurs they are running both sections of the 6SN7 as voltage gain stages in parallel (known as a common cathode amplifier). Looks like they are using an active load on the 6SN7 gain stage, likely a transistor-based constant current source, the stage is then direct-coupled to the grid of the 6AS7G output stage. The output is run as a pair of parallel cathode followers. Also looks like they are using some sort of series voltage regulator in the power supply.

So it goes: CCS loaded parallel 6SN7 gain stage > direct coupled to parallel 6AS7G cathode follower output stage.

Given the 6AS7G output stage is in parallel, output impedance is probably something like 30-40ohms.

 

[HTSkywalker]

Well, it's hard to glean much about the design without seeing a schematic, but it sounds like they are running the two 6N8P / 6SN7 as parallel cathode followers. This product is purely a buffer - it will provide an easy load for the input device and a low output impedance to drive the output device. It will also add some small degree of second harmonic distortion. There will be a small amount of voltage gain lost - the ideal cathode follower operates at unity gain, but in reality it might be gain of 0.85 or 0.9.

I will say that unless it is something very out of the ordinary, most gear is designed to have a sufficiently low output impedance a high enough input impedance such that they will pair well with just about anything you connect them to. But who knows! Maybe putting this cathode follower buffer in between the two will have some subjective benefit, it's hard to say, I guess you will have to try it out and see of course if you like the sound of second harmonic distortion, what some people call "tube sound", it might be an upgrade!

Am already using the Darkvoice as a preamp adding tube warmth to my amp which I really like but it also adds gain and a potentiometer in between, after checking for dedicated cathode follower buffers I ran into the Yaqin with around 10 years of R&D in this specific buffer. Although it's a Chinese amp but reviews are good and given the price I was tempted to try as I have a good bunch of 6SN7.

 

[HTSkywalker]

Attached is the schematic of the Yaqin

 

[L0rdGwyn]

Am already using the Darkvoice as a preamp adding tube warmth to my amp which I really like but it also adds gain and a potentiometer in between, after checking for dedicated cathode follower buffers I ran into the Yaqin with around 10 years of R&D in this specific buffer. Although it's a Chinese amp but reviews are good and given the price I was tempted to try as I have a good bunch of 6SN7.

Well it's worth a shot! Pretty cheap either way.

Attached is the schematic of the Yaqin

Yup, it's a pair of parallel cathode followers with a simple CRC passive power supply.

 

[HTSkywalker]

Well it's worth a shot! Pretty cheap either way.



Yup, it's a pair of parallel cathode followers with a simple CRC passive power supply.

Looks like a clean straight design

 

[Paladin79]

This weekend I will be adding some unusual Melz 1578 solid plate 1958's and Tung Sol 5998's to two of my Incubus amps for a comparison between ZMF Verite's and Focal Utopias. (I already have a pretty good idea of the winner but this is fun anyway.)

I should have a distinguished assortment of listeners and as many as five systems setup, mostly class A, tubes and then a thing I built for my wife that should go unmentioned. This is supposed to be a mini open house for my wife's pottery studio but I should be able to sneak a few of the husbands inside for some serious listening.

 

[Galapac]

This weekend I will be adding some unusual Melz 1578 solid plate 1958's and Tung Sol 5998's to two of my Incubus amps for a comparison between ZMF Verite's and Focal Utopias. (I already have a pretty good idea of the winner but this is fun anyway.)

I should have a distinguished assortment of listeners and as many as five systems setup, mostly class A, tubes and then a thing I built for my wife that should go unmentioned. This is supposed to be a mini open house for my wife's pottery studio but I should be able to sneak a few of the husbands inside for some serious listening.

Verite open or closed? I have both now and love them for different reasons as well as the Utopia.
If you are comparing the VO to Utopia you may find that the VO is a warmer headphone in the upper range but the Utopia is still a detail monster.

 

[Paladin79]

Verite open or closed? I have both now and love them for different reasons as well as the Utopia.
If you are comparing the VO to Utopia you may find that the VO is a warmer headphone in the upper range but the Utopia is still a detail monster.

VO, I rarely do closed and having listened for a few days myself I kind of know the differences. I even went as far as adapting a DanaCable Lazuli Reference for use with the VO in an attempt to level the playing field.

I will get a few other opinions out of fairness including a former music professor who taught trumpet soloists and a couple other friends.

If I have time I will build identical cables for each so those are less of a factor. 8 strand OCC copper most likely.

 

[HTSkywalker]

This weekend I will be adding some unusual Melz 1578 solid plate 1958's and Tung Sol 5998's to two of my Incubus amps for a comparison between ZMF Verite's and Focal Utopias. (I already have a pretty good idea of the winner but this is fun anyway.)

I should have a distinguished assortment of listeners and as many as five systems setup, mostly class A, tubes and then a thing I built for my wife that should go unmentioned. This is supposed to be a mini open house for my wife's pottery studio but I should be able to sneak a few of the husbands inside for some serious listening.

Am sure you will be doing the husbands a favor especially if you serve single malts, probably recruit some proteges