Hi, I'm interested in putting a new amp together. I've been accumulating old Soviet tubes via sellers in Ukraine and it's time to heat some of them up. In addition to the more common "guitar amp" tubes, I've picked up some oddballs like the GU-50, G-807 and its pal the 6P7S, many 6N1P, 6N2P, 6N6P noval tubes, as well as the octal preamp 6N8S (6SN7) and the 6N9S (6SL7) tubes. I would like to learn more about designing Single-Ended amplifiers and gain a more thorough understanding of preamp/driver stages.
So, based on some ideas about the GU-50 as a "Poor Man's 300b" I've hacked up a popular schematic for a 300b SET amplifier, and jammed the GU-50 in there:
Creating an Operating Point of 400v, 83mA with a 5K load calls for a 680ohm Cathode resistor, resulting in -56v btw Grid and Cathode according to this Vtaidy chart:
And while I can't test these tubes myself, one batch came with a hand-written note indicating the tubes had been tested at 400v on the plate, -60v on G1, with G2 and G3 connected to the plate, each tube had a value from 80-100mA written on the cap, so I feel confident powering the power tubes in this way.
So, how do I determine the required drive signal for this power tube, and how do I increase the line level signal from my audio source (DAC or Phono pre) enough to do that? The schematic shown uses 2 gain stages, both sides of a 6SN7. Could I use the 6N1P in the same circuit? I I can easily find data on the 6SN7 as a resistance-coupled amplifier, but the data for the 6N1P is a little harder to come by. Any advice on how to proceed is welcome, thanks for taking a look.
So, based on some ideas about the GU-50 as a "Poor Man's 300b" I've hacked up a popular schematic for a 300b SET amplifier, and jammed the GU-50 in there:
Creating an Operating Point of 400v, 83mA with a 5K load calls for a 680ohm Cathode resistor, resulting in -56v btw Grid and Cathode according to this Vtaidy chart:
And while I can't test these tubes myself, one batch came with a hand-written note indicating the tubes had been tested at 400v on the plate, -60v on G1, with G2 and G3 connected to the plate, each tube had a value from 80-100mA written on the cap, so I feel confident powering the power tubes in this way.
So, how do I determine the required drive signal for this power tube, and how do I increase the line level signal from my audio source (DAC or Phono pre) enough to do that? The schematic shown uses 2 gain stages, both sides of a 6SN7. Could I use the 6N1P in the same circuit? I I can easily find data on the 6SN7 as a resistance-coupled amplifier, but the data for the 6N1P is a little harder to come by. Any advice on how to proceed is welcome, thanks for taking a look.
Hi!
Regards,
Jose
- Yes, you can use 6N1P instead of 6SN7, but you will have to play with the 1st and 2nd stage biasing to make sure the second stage's control grid is always negative relative to the cathode. This is crucial here because the second stage is DC-coupled with the first one. Please note you changed the tube to 6N1P but the pin numbers are 6SN7. ASttached 6N1P and 6N8S datasheets for reference.
- The Vtaidy chart is confusing, it shows screen grid at 250V, but Vq at 400V, that make no sense for triode operation. I assume that value is ignored. Also, if it is assuming fixed bias, then it has quite a bit of discrepancy with the second picture of the real measurement.
- The swing needed will be close to the bias value, in your case 56V peak (112Vpp).
Regards,
Jose
Hey, thanks for taking a look! I get most of what you're saying, I'll address your points in reverse order:
I assume the DAC puts out a typical line level signal, my phono pre is a bit hotter, I had to put a 100k pot on the output to avoid distortion with certain amps. All to say, I think I'm starting with a 1vrms signal, and it needs to be amplified to 56vrms?
Yes, the vtaidy charts are confusing, but they help me visualize load lines and the relationship btw mA, V, and the K resistor. You can see I ticked the box for Triode, which disables the Screen field and loads the appropriate chart.
As to biasing those preamp stages, here's where I need help. You're saying the amplified signal from the first stage must be lower than the voltage of the cathode of the 2nd stage? Or just that the Va of the 1st stage can't exceed the Vk of the 2nd stage? How do I reckon the value of that cathode? I tried to use the Vtaidy chart for that 2nd stage, with the 280v supply and the plate measuring 210v, that puts the OP at 1.3mA. Using my trusty Ohms Law calculator, 1.3mA * 27k (the K resistor) = 35v at the Cathode? I think that's wrong. A different version of this schematic shows 78v at the 2nd stage Cathode.
w
Last edited:
No problems!
1) Input signal: I have done some simulations, and with a 6N1P tube you get a relatively high gain. 1V RMS is too much, from what I see you will get 40V RMS (56V peak) with 55mV RMS at the input. The maximum of two 6N1Ps is around 900, in the circuit I simulated was around 730. Even a 6SN7will give you something between 300 and 400.
2) Yes, Va of the first stage cannot go higher than Vk of the second stage. Moreover, the second stage has a bypassed cathode, meaning you can only have a relatively small AC signal in Va to keep the second stage in pure class A.
I use a free tool called LTSpice for simulations. It does not replace the real hardware, but it gives you a decent prediction of how things will work, especially for simpler circuits without transformers. You can also do the bias calculations using the tube curves, that's a bit more difficult.
There are good tutorials for triode stages at Merlin Blencowe's website: http://www.valvewizard.co.uk/gainstage.html
This is the way the LTSpice simulator looks like:
Thanks. I really need to learn how to use LTSpice. Did not realize the 6N1P had such different characteristics.
You do have 6N8S (6Н8С) which is similar to the 6SN7. Less gain, but you can use the values in the original schematic.
For what is worth, I don't think this is a good schematic. Distortion is quite high after the first two stages. Also 280V is quite a low voltage for a driver. Are you set on this schematic or are you looking for other options? What voltage you got in the source?
What about removing the cathode bypass caps or on the second stage adding a series resistor with the bypass cap.
I'm not set on anything, just trying to learn how to design my own circuits. Probably need to figure out how to use LTSpice to do that. I don't need the power of a big push-pull amp and am curious to explore the world of SE amps. I'm currently in possession of a variety of parts, I was planning on using a Hammond 274BX PT that makes 750v@200mA, 193J choke and a pair of 18w OPTs from Shenzhen. Beyond that, I don't have any restrictions. If a single tube can give me enough voltage gain to drive the GU-50, I'll go that way.
I would think about gain structure.
If you have too much gain, every little bit of input noise will be amplified as much as the audio signal. You can then turn down the signal level with a potentiometer, but that makes the signal to noise ratio worse, as the pot contributes its own noise.
I would design from the output back to the input. Start with the output tube.
Figure out the load line for your GU-50 triode.
Decide how much signal voltage is needed to drive it to full power, with maybe a few dB extra for sources with unusually low signal level (like some DSD files, for instance).
Let's say 65V peak of audio signal will drive your GU-50 to clipping.
OK, now pick the driver tube (2nd stage). This tube will be driving the grid of the GU-50-triode, which has some capacitance, and probably will sound better driven from a low source resistance. I'd look at a low-mu, small power tube like a triode-wired 6V6GT (or the Russian equivalent 6P6S). That will have a gain of about 8x when triode wired, and output resistance of about 2k ohms. (6N1P has low enough rp, but has almost 4x as much gain.)
Figure the 6V6-Triode will have a grid-cathode voltage of about -10V, so can accept up to a 10V peak signal. That means it can swing as high as 80V peak into the GU-50 grid. That's good. It gives the 6V6 a little headroom.
Now, let's say you want to use one of your 6N1P tubes as the input tube (1st stage). You could leave its cathode unbypassed so it will have a gain of about 20x (that's a guesstimate). That means it will take 500mV peak audio signal to make 10V peak into the 6V6-Triode grid, which will amplify that to 80V peak signal into the GU-50-Triode grid.
Now figure out how to couple those three stages together.
If you decide to DC-couple the first two stages, you will quickly see how many times you need to re-figure your operating points, drawing loadlines, trying to juggle things to get close.
If you AC couple the first two stages (CR), you will see why so many circuits are done that way. It's so much easier to get your desired results.
Try using those voltages with 6N1P > 6V6GT-Triode > GU-50-Triode and see if you can figure out the DC voltages and parts values needed to get these tubes to behave as described above. It will be a good exercise in how to design a three-stage SE amplifier.
If you have too much gain, every little bit of input noise will be amplified as much as the audio signal. You can then turn down the signal level with a potentiometer, but that makes the signal to noise ratio worse, as the pot contributes its own noise.
I would design from the output back to the input. Start with the output tube.
Figure out the load line for your GU-50 triode.
Decide how much signal voltage is needed to drive it to full power, with maybe a few dB extra for sources with unusually low signal level (like some DSD files, for instance).
Let's say 65V peak of audio signal will drive your GU-50 to clipping.
OK, now pick the driver tube (2nd stage). This tube will be driving the grid of the GU-50-triode, which has some capacitance, and probably will sound better driven from a low source resistance. I'd look at a low-mu, small power tube like a triode-wired 6V6GT (or the Russian equivalent 6P6S). That will have a gain of about 8x when triode wired, and output resistance of about 2k ohms. (6N1P has low enough rp, but has almost 4x as much gain.)
Figure the 6V6-Triode will have a grid-cathode voltage of about -10V, so can accept up to a 10V peak signal. That means it can swing as high as 80V peak into the GU-50 grid. That's good. It gives the 6V6 a little headroom.
Now, let's say you want to use one of your 6N1P tubes as the input tube (1st stage). You could leave its cathode unbypassed so it will have a gain of about 20x (that's a guesstimate). That means it will take 500mV peak audio signal to make 10V peak into the 6V6-Triode grid, which will amplify that to 80V peak signal into the GU-50-Triode grid.
Now figure out how to couple those three stages together.
If you decide to DC-couple the first two stages, you will quickly see how many times you need to re-figure your operating points, drawing loadlines, trying to juggle things to get close.
If you AC couple the first two stages (CR), you will see why so many circuits are done that way. It's so much easier to get your desired results.
Try using those voltages with 6N1P > 6V6GT-Triode > GU-50-Triode and see if you can figure out the DC voltages and parts values needed to get these tubes to behave as described above. It will be a good exercise in how to design a three-stage SE amplifier.