Hi, I built this 6SN7 preamp a couple of years ago and would now like to lower its gain. From what I've been reading on here this preamp appears to have gain of around 20db. What would be the best way to lower its gain overall?
Circuit is as below...
Thanks in advance...🙂
Glen.
Circuit is as below...
Thanks in advance...🙂
Glen.
Attachments
You could try a feedback loop from the output capacitor to the input tube grid. Just experiment with various value resistors.
You can try this topology:
http://kuva.termiitti.com/image/17219.gif
This makes possible to adjust the gain from -1 dB to some 20 dB.
With gain below 10 dB the output impedance is below 1 kohms and distortion below 0,1 % when RL is some 10 kohms and output level up to 5 Vrms.
The good feature of this topology is that the input impedance is constant unlike when the feedback loop is brought to grid and different kind of source impedances used.
An externally hosted image should be here but it was not working when we last tested it.
http://kuva.termiitti.com/image/17219.gif
This makes possible to adjust the gain from -1 dB to some 20 dB.
With gain below 10 dB the output impedance is below 1 kohms and distortion below 0,1 % when RL is some 10 kohms and output level up to 5 Vrms.
The good feature of this topology is that the input impedance is constant unlike when the feedback loop is brought to grid and different kind of source impedances used.
adjust the 820 ohm resistor on the second triode to 1k - 2k for hard set gain.
if you keep the control, try a 250k for the input also.
if you keep the control, try a 250k for the input also.
Last edited:
No. It is bias resistor for the second triode and has no effect to total gain.
Since the last triode is a cathode follower, its gain is already near unity.
The gain comes entirely from 1st triode.
With existing topology, just use attenuator at the input.
Now somebody can say that this ruins the S/N-ratio, but it is not the case since the line level is so high that it can be attenuated remarakably before the S/N becomes an issue.
Your thread doesn't make any sense to me at all. I built Franks 6sn7 line stage and do not see your point of needing less gain. When built the same as the schematic shows it is very adjustable. I even added a cap to parallel the cathode resistor in the input stage thus increasing the gain I believe. When used with a good quality pot on the input one can adjust this line stage and get ANY amount of signal out as needed.
If your not happy with the line stage then scrap it and purchase one stereo pot or two mono pots and do without a line stage.
Quote:
You can try this topology:
http://kuva.termiitti.com/image/17219.gif
http://kuva.termiitti.com/image/17219.gif
This makes possible to adjust the gain from -1 dB to some 20 dB.
With gain below 10 dB the output impedance is below 1 kohms and distortion below 0,1 % when RL is some 10 kohms and output level up to 5 Vrms.
The good feature of this topology is that the input impedance is constant unlike when the feedback loop is brought to grid and different kind of source impedances used.
Sorry but not a believer in this type of topology. A control on the output generally sucks in my opinion. Never heard one that sounded worth turning on.
If your not happy with the line stage then scrap it and purchase one stereo pot or two mono pots and do without a line stage.
Quote:
You can try this topology:
http://kuva.termiitti.com/image/17219.gif
http://kuva.termiitti.com/image/17219.gif
This makes possible to adjust the gain from -1 dB to some 20 dB.
With gain below 10 dB the output impedance is below 1 kohms and distortion below 0,1 % when RL is some 10 kohms and output level up to 5 Vrms.
The good feature of this topology is that the input impedance is constant unlike when the feedback loop is brought to grid and different kind of source impedances used.
Sorry but not a believer in this type of topology. A control on the output generally sucks in my opinion. Never heard one that sounded worth turning on.
Last edited:
The reason is maybe that this does not generate distortion like Frank's topology.
Some people hear somewhat distorted signal sounding great, especially those who evaluate amplifiers by listening only.
quote:
The reason is maybe that this does not generate distortion like Frank's topology.
I would be more than willing to review your distortion figures for both line stages.
The reason is maybe that this does not generate distortion like Frank's topology.
I would be more than willing to review your distortion figures for both line stages.
I have only distortion figures of my version, not those of Frank's.
Mine are as follows:
Sorry that my documentation is in finnish only, but I guess that we will manage.
Let's look the first table at the top of the page. There the gain of the stage is set to 2 (6 dB ). There are distortion values given with two different load resistances namely 22 kohms and 10 kohms.
Next table below is the case when the gain is set to 3 (9,5 dB).
And in the last table the LM317 CCS was substituted with 10 kohms cathode resistor and the anode resistors matched for optimum linearity.
I have done several circuits with ordinary common cathode triode and know that the distortion figures are some 3 to 5 times worse.
However, I suggest that we share this project so that you will build Frank's topology and publish the distortion figures here. Then we can compare. OK ?
Mine are as follows:
An externally hosted image should be here but it was not working when we last tested it.
Sorry that my documentation is in finnish only, but I guess that we will manage.
Let's look the first table at the top of the page. There the gain of the stage is set to 2 (6 dB ). There are distortion values given with two different load resistances namely 22 kohms and 10 kohms.
Next table below is the case when the gain is set to 3 (9,5 dB).
And in the last table the LM317 CCS was substituted with 10 kohms cathode resistor and the anode resistors matched for optimum linearity.
I have done several circuits with ordinary common cathode triode and know that the distortion figures are some 3 to 5 times worse.
However, I suggest that we share this project so that you will build Frank's topology and publish the distortion figures here. Then we can compare. OK ?
I found some of my old test results for 6N1P used in common cathode circuit.
Used values are:
Ra = 56 kohms, Rk = 1 kohms, Rg1 = 680 kohms, Ub = 350 V, Ua = 160 V, Rg2 = 150 kohms.
Results are:
Uo = 10 V, THD = 0,47 %
Uo = 3,0 V, THD = 0,17 %
G = 24,2 dB
If cathode is by-passed:
Uo = 10 V, THD = 0,72 %
Uo = 3,0 V, THD = 0,24 %
G = 29,4 dB
When this kind of circuit is connected to cathode follower like in case of Frank's circuit, the THD is slightly lower since the Rload (Rg2) becomes much higher than 150 kohms I used, but I think that the results are in right size range.
An externally hosted image should be here but it was not working when we last tested it.
Used values are:
Ra = 56 kohms, Rk = 1 kohms, Rg1 = 680 kohms, Ub = 350 V, Ua = 160 V, Rg2 = 150 kohms.
Results are:
Uo = 10 V, THD = 0,47 %
Uo = 3,0 V, THD = 0,17 %
G = 24,2 dB
If cathode is by-passed:
Uo = 10 V, THD = 0,72 %
Uo = 3,0 V, THD = 0,24 %
G = 29,4 dB
When this kind of circuit is connected to cathode follower like in case of Frank's circuit, the THD is slightly lower since the Rload (Rg2) becomes much higher than 150 kohms I used, but I think that the results are in right size range.
Last edited:
sorry about post #5 was not thinking clearly.
if you increase the cathode resistor on the first stage to say 2k will drop your gain
if you increase the plate resistor, your gain goes down exponentially. However, you might have to change the 220k resistor on the grid of the next stage for phase/frequency response (in extreme cases).
but here is the formulas for the first stage:
key abbreviations:
Av= audio voltage gain
Rin= Input impedence
Rout= output impedence
f1= freq response bandwidth product 1
f2= freq response bandwidth product 2
Rg = the grid resistor
Rp = the plate resistor
Rk = the cathode resistor
Rl = the load resistance, or the input resistance of the next stage
Ra = the total load resistance, which is Rp in parallel with the input resistance of the next stage, Rl. If there is no Rl, Ra = Rp.
ra = the internal plate resistance of the tube
mu = the mu of the tube
formulas:
Av = (mu * Rp)/(Rp + ra + (mu + 1)*Rk)
Rin = Rg
Rout = [(ra + (mu + 1)*Rk) * Rp] / [(ra + (mu + 1)*Rk) + Rp]
f1 = 1/(2*pi*Ci*Rg) - highpass breakpoint due to Ci/Rg
f2 = 1/(2*pi*Co*(Rout + Rl)) - highpass breakpoint due to Co/Rout/Rl
if you increase the cathode resistor on the first stage to say 2k will drop your gain
if you increase the plate resistor, your gain goes down exponentially. However, you might have to change the 220k resistor on the grid of the next stage for phase/frequency response (in extreme cases).
but here is the formulas for the first stage:
key abbreviations:
Av= audio voltage gain
Rin= Input impedence
Rout= output impedence
f1= freq response bandwidth product 1
f2= freq response bandwidth product 2
Rg = the grid resistor
Rp = the plate resistor
Rk = the cathode resistor
Rl = the load resistance, or the input resistance of the next stage
Ra = the total load resistance, which is Rp in parallel with the input resistance of the next stage, Rl. If there is no Rl, Ra = Rp.
ra = the internal plate resistance of the tube
mu = the mu of the tube
formulas:
Av = (mu * Rp)/(Rp + ra + (mu + 1)*Rk)
Rin = Rg
Rout = [(ra + (mu + 1)*Rk) * Rp] / [(ra + (mu + 1)*Rk) + Rp]
f1 = 1/(2*pi*Ci*Rg) - highpass breakpoint due to Ci/Rg
f2 = 1/(2*pi*Co*(Rout + Rl)) - highpass breakpoint due to Co/Rout/Rl
Last edited:
quote:
The reason is maybe that this does not generate distortion like Frank's topology. Some people hear somewhat distorted signal sounding great, especially those who evaluate amplifiers by listening only.
If we were really interested in distortion only then it is very possible we would scrap the tubes in favor of 100% SS circuits. You site a difference in the distortion between your circuit and Franks but don't have anything to back up your claims. Distortion is a funny animal. We can totally buy into the 001% theory and convince ourselves that distortion is all that matters. Granted distortion does play a small part in our quest for a good sounding piece of equipment and or circuit. It has to sound good also. It has been my experience that not all that measures low in distortion sounds good. Personally I do not like any line stage with a control on the output. I believe the low frequency response suffers as a result. If this topology was superior then I would make the assumption that most circuits would utilize this.
To me your control on the output is like driving down the street with the accelerator pushed to the floor and your left foot hard on the brake pedal. It makes more sense to my limited thinking to adjust the signal being amplified instead of letting the circuit take what is being fed to it and then trying to adjust after all the amplification has been done.
A lot of what we do takes into consideration of what our ears hear. Got a favorite cap? People are convinced that they can and do hear difference between types and brands of capacitors. I wouldn't criticize people for selecting what sounds pleasing to their ears.
The reason is maybe that this does not generate distortion like Frank's topology. Some people hear somewhat distorted signal sounding great, especially those who evaluate amplifiers by listening only.
If we were really interested in distortion only then it is very possible we would scrap the tubes in favor of 100% SS circuits. You site a difference in the distortion between your circuit and Franks but don't have anything to back up your claims. Distortion is a funny animal. We can totally buy into the 001% theory and convince ourselves that distortion is all that matters. Granted distortion does play a small part in our quest for a good sounding piece of equipment and or circuit. It has to sound good also. It has been my experience that not all that measures low in distortion sounds good. Personally I do not like any line stage with a control on the output. I believe the low frequency response suffers as a result. If this topology was superior then I would make the assumption that most circuits would utilize this.
To me your control on the output is like driving down the street with the accelerator pushed to the floor and your left foot hard on the brake pedal. It makes more sense to my limited thinking to adjust the signal being amplified instead of letting the circuit take what is being fed to it and then trying to adjust after all the amplification has been done.
A lot of what we do takes into consideration of what our ears hear. Got a favorite cap? People are convinced that they can and do hear difference between types and brands of capacitors. I wouldn't criticize people for selecting what sounds pleasing to their ears.
It's actually HF response that suffers- and the frequency response changes with control setting. Worse yet, depending on the next device in the chain, this topology can increase noise.
Yes, bad idea.
There is no "control" at the output.
The variable resistor I have drawn at the output only describes that the NFB and thus the gain can be set by two resistors that the variable resistors forms.
So it is really not any volume control ! The volume control - if required - should ofcourse be situated at the input.
I think you have understood my circuit diagram wrong. It's idea was not to show an example of actual line stage. It was inteded to show one way to make a non-inverting amplifier stage with the possibility to adjust the gain low without typical negative result like when the NFB is fed to the grid of 1st triode.
My experience is that not very many amplifiers that measures high in distortion sound good, but most amplifiers with low distortion and other good test results sound good.
Low distortion ( = good linearity) is only one important feature of a good amplifier. There are several more.
Can you enlighten me a bit ? I am not sure what you mean.
I have claimed that "my" circuit has lower distortion than Frank's. and also given test results of my circuit. So what is your point ?
There is no control at the output, but this should be already clear.
No. NFB improves both. With gain 3 (9,5 dB) the hf roll-off was 600 kHz.
These things need not to be based on what you or I believe, these are basic thing that can (and must) be tested while developing amplifiers.
I have not said that this topology is superior, only given some facts about it. But nothing prevents you to build one. Then you also know what it is instead of what you now believe or imagine.
Neither do I.
My way to do is first measure that all main parts of the performance is OK.
(distortion, frequency response, hum and noise, transient response,etc. etc. ) Only after that I begin listening tests.
Some people only listen, not test or measure anything. That is their way to do amplifiers, but I would not call such process as developing.
If this had a volume pot after the circuit your criticism would be valid. Instead, it is adjustable negative feedback. This can have problems, but replace the pot by a pair of resistors and you have a fairly sensible circuit. Some (wrongly) call this configuration a "grounded grid" preamp. It is actually a cathode-coupled circuit, or an alternative version of LTP.6BG6GA said:
It might have a relatively high proportion of 3rd-order components, as the approximate balance will reduce 2nd-order. There is the potential for LF problems with this circuit, but the component values appear to have been chosen to avoid this. Tweakers would need to resist the temptation to simply increase the value of the 22nF input capacitor, without making corresponding adjustments elsewhere.
You can try any of these:
1. Increase cathode resistor in the first stage;
2. Attenuate input signal
3. Make anode resistor of the first stage as a seres connection of 2 resistors and take the signal from from their connection point.
1. Increase cathode resistor in the first stage;
2. Attenuate input signal
3. Make anode resistor of the first stage as a seres connection of 2 resistors and take the signal from from their connection point.
- Status
- Not open for further replies.