Copyright only means that you cannot publish another's document. Instead, you can create your own similar document and publish that. It can still contain similar information.
Last edited:
jaz_12,
Thanks for including the complete schematic in your Post # 20.
I am not sure exactly what you want this circuit to do. Did you want a stereo preamp that is non inverting from input to output?
It does appear to be a two channel non-inverting preamp. But if you have two signals, such as Left and Right, the common plate resistor of the input triodes will cause the channel signals to blend (it will not have good channel separation). You will have to use separate plate load resistors for the input tubes if you want channel separation.
The circuit is a long tailed pair of cathode coupled stages like a phase splitter, but is modified in that the un-driven side is an SRPP stage. The load that the two cathode coupled stages see is not symmetrical, because of one plate load is a resistor (the common plate resistor in Post # 20 that causes channel blending), and the other plate load is a higher impedance: the triode of the top half of the SRPP.
Also, there is signal to one grid on one side of the cathode coupled tubes, and the other side grid receives the negative feedback signal. Since this means both grids move, there is no grounded grid stage (a naming problem for Mr. ____ ).
You indicated a similar circuit that used 12AU7 tubes. The 12AU7 tubes have a maximum filament to cathode voltage rating of 180V.
But the 6N1P has a maximum filament to cathode voltage rating of only 100V. I suspect that there is going to be problem when using 6N1P tubes (unless you use one of the 6N1P tubes as the SRPP top tube in both channels, and you use a separate filament supply that is elevated so that the 100V rating is not violated.
Thanks for including the complete schematic in your Post # 20.
I am not sure exactly what you want this circuit to do. Did you want a stereo preamp that is non inverting from input to output?
It does appear to be a two channel non-inverting preamp. But if you have two signals, such as Left and Right, the common plate resistor of the input triodes will cause the channel signals to blend (it will not have good channel separation). You will have to use separate plate load resistors for the input tubes if you want channel separation.
The circuit is a long tailed pair of cathode coupled stages like a phase splitter, but is modified in that the un-driven side is an SRPP stage. The load that the two cathode coupled stages see is not symmetrical, because of one plate load is a resistor (the common plate resistor in Post # 20 that causes channel blending), and the other plate load is a higher impedance: the triode of the top half of the SRPP.
Also, there is signal to one grid on one side of the cathode coupled tubes, and the other side grid receives the negative feedback signal. Since this means both grids move, there is no grounded grid stage (a naming problem for Mr. ____ ).
You indicated a similar circuit that used 12AU7 tubes. The 12AU7 tubes have a maximum filament to cathode voltage rating of 180V.
But the 6N1P has a maximum filament to cathode voltage rating of only 100V. I suspect that there is going to be problem when using 6N1P tubes (unless you use one of the 6N1P tubes as the SRPP top tube in both channels, and you use a separate filament supply that is elevated so that the 100V rating is not violated.
Another way to look at how your circuit works is as follows: The first tube is wired as a cathode follower that is DC coupled to the next stage. But instead of driving the grid of the next stage, it drives the cathode. The tube on top in the second stage is acting as a pseudo current source. Whenever you have stacked tubes like this where the top tube has a significantly elevated cathode voltage, you need to worry about max heater to cathode rating for the tube as 6A3sUMMER points out. The 47K resistor on the output provides a feedback signal, in conjunction with the 20K resistor below it, sets the amount of feedback provided.
Some things you might consider to improve things:
Some things you might consider to improve things:
- Replace the tube-based pseudo current source (top tube in the second stage) with a silicon current source.
- I'm not sure you need a stabilized filament supply. I think you can run the filaments on AC and be fine, if you use good construction techniques (but you still have the problem of needing to elevate the filament voltage for the upper triode in the second stage, unless you go the route of a silicon current source).
- More personal preference, but I don't care to use bridge rectifiers. I'd use discrete diodes for rectification as that gives you more choices for diode types.
There is a capacitor on that junction. If it is Sufficiently Large, the audio voltage here will be Very Small, and the crosstalk may be acceptable.
Do math. I can't read the values. Assume that is a 10uFd cap and the plate resistance is around 10K. The impedance of 10uFd @ 20Hz is 1K; @ >200Hz <100r. Therefore the signal at the common plate node is -20dB @ 20Hz, >-40dB @ >200Hz. Not a lot of crosstalk, especially in the critical midrange. However signal at the common plate node is further attenuated by the plate resistance of the other triode, probably another 20dB. Making >60dB @ >200Hz. Not something to boast about but certainly better than most stereo sources or listening systems.
EDIT- looks like 47uFd. Add 14dB. >70dB over most of the audio band.
And a comment. You (OP) want to build this because it is supposed to be good. The Designer surely breadboarded it and tried multiple variations before putting His Name on it and selling to to the world. Do you think you know more than he does? (Then why adopt his plan?) Do you think he cheaped-out by not using another 1-buck R-C network? (In a $500+ product?)
But sure, it is your project. Do what ever you want. The changes to separate the plate feeds should be obvious.
Last edited:
PRR,
Yes, I did not notice the capacitor. That takes care of that amplifier.
There are some 6N1P data sheets out there that says: 1. The 6N1P has 7500 uMhos and 4400 Ohms of plate resistance.
2. But If you look at that very same data sheet, and draw lines on the tube curves, you will find that the 6N1P has about 4400 uMhos, and about 7500 Ohms plate resistance. Exactly the opposite of #1. Interestingly, since the numbers are the same but swapped, that causes the u be the same for both the correct and incorrect spec (u = 33). u = Gm x rp
Having used the 6N1P, my experience has told me that #2 is closer to the truth. Pick the tubes you want to use by the curves not only by the numbers.
Yes, I did not notice the capacitor. That takes care of that amplifier.
There are some 6N1P data sheets out there that says: 1. The 6N1P has 7500 uMhos and 4400 Ohms of plate resistance.
2. But If you look at that very same data sheet, and draw lines on the tube curves, you will find that the 6N1P has about 4400 uMhos, and about 7500 Ohms plate resistance. Exactly the opposite of #1. Interestingly, since the numbers are the same but swapped, that causes the u be the same for both the correct and incorrect spec (u = 33). u = Gm x rp
Having used the 6N1P, my experience has told me that #2 is closer to the truth. Pick the tubes you want to use by the curves not only by the numbers.
Stereo preamp.
It was first post question. Now is clear.
It makes sense.
Mr. Broskie explained it very well. I was looking for information on this forum and on the Internet and the name " Grounded Grid preamp" is incorrect but used. Sorry. I should add "Grounded Grid by Mr. _____".
6N1P-EW was used because a few years ago I bought it. According to the original Russian datasheet Ukh= 120V. In other documents, I found 100V+5%.
In my prototype Ukh is around 100V.
Yes. For 6N1P
=======
Very thank you
NO offense. But do you think I know more? What is the reason to ask you on this forum if it's truth ;-)
For me is only fun. I do not want to build a new brilliant preamp and seel it. Or write a book.
I'm very thankful for all help and with humility I read all critic and advices/tips.
I was going to say--- Gm and rp are VERY variable with current. The same tube can have rp of 20K or 200K just by going from high current to low current.
Gm also changes, the other way (low current, low Gm).
And as you say: Mu tends to vary very little. (You can almost measure Mu with a ruler. The essential Mu can be figured from dimensions. In real tubes, end-effects and other non-ideal effects make Mu fall-off at low and very high current, but it changes little in the "normal" range of the tube.) Mu is the only "Constant" of a tube. Gm and rp do vary widely.
So they could show "good" numbers just by picking an extreme current.
BUT--- I found the Svetlana sheet and after plotting a 4.4K line.... I agree with you, it will never be 4.4K in the negative grid range. Somebody got the numbers swapped.
Have fun!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.