@Kay Pirinha - good point on the MJE340. I could get away with something smaller, although does it make that much difference in the cascode position?
Yes, R22, 23 are primarily to reduce gain and lower distortion. I was trying to avoid GNF, but will consider that. Adding the option to my PCB with a jumper so I can compare the sound would be interesting.
I also noticed that I have the output polarity labeled incorrectly. The upper source follower output will be inverted phase from the left triode gate (connected to the positive phase input), so I will probably just swap the SF anode connections to keep the schematic clearer.
@6A3sUMMER - Not a commercial product. Just for my own use. This doesn't seem very complicated to me though. I'm also working on a build based on Allen Wright's RTP3C which is a lot more complex. This is to tide me over until I can finish that project.
Yes, R22, 23 are primarily to reduce gain and lower distortion. I was trying to avoid GNF, but will consider that. Adding the option to my PCB with a jumper so I can compare the sound would be interesting.
I also noticed that I have the output polarity labeled incorrectly. The upper source follower output will be inverted phase from the left triode gate (connected to the positive phase input), so I will probably just swap the SF anode connections to keep the schematic clearer.
@6A3sUMMER - Not a commercial product. Just for my own use. This doesn't seem very complicated to me though. I'm also working on a build based on Allen Wright's RTP3C which is a lot more complex. This is to tide me over until I can finish that project.
The U2/U1 source followers should be a cascode. The variable input impedance due to variable drain voltage capacitance will be audible. I have found that the magnitude of that capacitance is not such an issue, but it varying sure seems to be. So to say, building with a single, and then applying cascode blocks was noticeable.
That said, the whole buffer has struck me as not required; I am driving a crossover right off EL84 triode plates and have no issue. The load it is working into is 10kOhm per phase. Output caps are significantly larger though...
Douglas
That said, the whole buffer has struck me as not required; I am driving a crossover right off EL84 triode plates and have no issue. The load it is working into is 10kOhm per phase. Output caps are significantly larger though...
Douglas
A tube preamp like yours or like Allen Wright's is simple, . . . when compared to the super duper best solid state IC Op Amp that the semiconductor industry can produce.
How many recording studios and production studios use as complex of a vacuum tube preamp as yours?
Is the sound of the recording affected beyond hope because of their preamp's shortcomings?
(I admit, I have not been in a recording studio for a couple of decades).
How many recording studios and production studios use as complex of a vacuum tube preamp as yours?
Is the sound of the recording affected beyond hope because of their preamp's shortcomings?
(I admit, I have not been in a recording studio for a couple of decades).
@Bandersnatch - The drain-to-source capacitance change with voltage is fairly modest with the parts I'm using. Vdss hovers around 107V, so output capacitance will vary by a few pf with typical output voltage swing (although this could be 5-10% change). This assumes that the actual behavior is close to what is shown in the attached graph despite my operating conditions differing from the test conditions. Do you think this is enough to justify the added complexity of cascoding?
My system is bi-amped and I'm actually using a separate op-amp based buffer to drive the woofer amps since these only have single-ended inputs. So this line stage will handle from 170Hz on up which is why I'm using a somewhat smaller coupling cap.
My system is bi-amped and I'm actually using a separate op-amp based buffer to drive the woofer amps since these only have single-ended inputs. So this line stage will handle from 170Hz on up which is why I'm using a somewhat smaller coupling cap.
I have seen curves like that, and I still cascode. Easier still is of course dropping the whole buffer stage on the first build... 🙂
Douglas
Douglas
@Bandersnatch - Thanks for your help. Cascode it is 🙂. Now I have to see if I can make it all fit in my available space.
I assume you are suggesting replacing my bjt current sources with this? Have you had better luck with this design compared to the circuit I posted?
No, just the source follower. Not sure how to do a cascode emitter follower. What I am suggesting is depletion-mode ones MOSFET, and items that act like depletion-mode( the 10M45 ). Then you don't need a voltage reference to lay two on top of each other...
Douglas
Douglas
@Bandersnatch - Douglas, I was planning to take a page out of Allen Wright's playbook and use a circuit similar to what he used in his cascoded cathode follower in the RTP3 (see attached). I have almost finished the PCB layout and everything fits ok.
