Thorsten 6SN7 Line stage.

[Cassiel]

I need to lower the gain of a 6SN7 preamp (GC+CF), it has 26dB of gain and I want something around 12dB.

Thorsten linestage seems perfect for my needs but I've read negative comments about its HF response. I have a 100K motorised pot so I can't change that but maybe I could modified the feedback loop to lower the value of the 100K resistor at the 6SN7's grid.

Is it good as it is? Any ideas?

 

[SpreadSpectrum]

I need to lower the gain of a 6SN7 preamp (GC+CF), it has 26dB of gain and I want something around 12dB.

Thorsten linestage seems perfect for my needs but I've read negative comments about its HF response. I have a 100K motorised pot so I can't change that but maybe I could modified the feedback loop to lower the value of the 100K resistor at the 6SN7's grid.

Is it good as it is? Any ideas?

It is a couple of paralleled 6SN7 sections with plate to grid feedback. The trouble with plate to grid feedback when implemented like this is it has a somewhat low input impedance. It needs to be driven by a low output impedance or by a constant high output impedance(and you factor that into the 100k resistor). The volume pot is neither. I would put something between the pot and the feedback network, otherwise your feedback will vary depending on the volume position.

The amount of feedback is determined by the ratio of the sum of the 100k resistor and the output impedance of the pot - and the 240k resistor. You could always build it and see if that variance bothers you or you could design it out by buffering the pot.

 

[Cassiel]

you could design it out by buffering the pot.

Something like this?
BTW, T. 6sn7 line stage has a gain of 6 dB. I need a max. gain of 12 dB but less is better.

 

[SpreadSpectrum]

That looks just a low pass filter between volume and the stage, I was thinking a low impedance active buffer, maybe a source follower would work well and not require cutting another hole...

Well, you can change the feedback resistors to get whatever gain you want up to the gain without feedback. That's not a problem. If you are having trouble with the math, it is practically the same as an inverting op-amp stage. There are plenty of web resources dedicated to thorough analysis of this feedback configuration. Of particular concern is that you want to make the feedback network low enough impedance to drive the miller capacitance of the stage but high enough to not place an undue burden on the driving stage.

 

[Cassiel]

Thanks for your reply, very helpful.

 

[tinitus]

Boozhound Laboratories

 

[mashaffer]

6SN7 is pretty linear is it not? Why not skip the feedback and wire the second triode as a bootstrapped CF? Or if you are only going to be using tube power amps then do the parallel triodes but again skip the FB.

 

[Cassiel]

Or if you are only going to be using tube power amps then do the parallel triodes but again skip the FB.

Too much gain (even with low plate resistor and unbypassed cathode) and too high Zout. I might try a plate follower if this topology doesn't work well.

 

[mashaffer]

Too much gain... Maybe you need a passive attenuator and a 1:2 input transformer.

 

[mashaffer]

I wonder if a triode strapped small power pentode might have some advantage for you (lower mu). Might be worth searching for some outside the box low mu options.