Hello, I'm kind of new to tube amp building. Now that I own a pair of type 50 tube from Philco, I'm planning for a type 50 SET amp and deciding the circuit structure I'm going to use.
As far as I know, most (I would say over 90% I've seen) 50 SET builds use interstage transformers to overcome the maximum 10k grid leak resistance limitation. Some also use grid chokes. I personally want to try out some structures without an iron driving 50, so I looked up through the web. The only examples I could find were (1) a direct coupled cathode follower driver, either (Japanese site): http://www.furo-visu.com/s_minor/SG-50-s1.html Schematic: http://www.furo-visu.com/s_minor/minor_gazo/SG-50-s1.gif
or https://www.diyaudio.com/community/threads/type-50-tube-amp.335913/post-5745013
which seemed promising considering that the follower presents an output impedance of less than 1kohms.
Another example was the Loftin-White circuit, which is simple enough. The way I don't like cathode follower is that it requires a C- power supply to swing the output to negative grid voltage (if I don't want to waste power on the 50 cathode bias), which means setting up another voltage rail. And Loftin-White may burn the precious 50 tube if the driver tube's absent or not heating enough to pull down the grid during startup.
But as I learned from other driver circuits, there're plenty of driver tubes like 5842/417A, D3a, EC8010, etc. that will deliver 70Vpp to a 10k load in common cathode. Even parallel 5687 or 6SN7 would do the job just fine. Another possibility might be using other topologies. As I calculated, even the weakest 12AU7 would supply a 70Vpp swing just fine. Swapping the tube to mightier 6CG7/12BH7/6H30, and a 100Vpp output with minimal distortion isn't a dream!
So my question is: Why is nobody driving the 50 tube in these ways? Am I missing something? Would appreciate your opinions.
As far as I know, most (I would say over 90% I've seen) 50 SET builds use interstage transformers to overcome the maximum 10k grid leak resistance limitation. Some also use grid chokes. I personally want to try out some structures without an iron driving 50, so I looked up through the web. The only examples I could find were (1) a direct coupled cathode follower driver, either (Japanese site): http://www.furo-visu.com/s_minor/SG-50-s1.html Schematic: http://www.furo-visu.com/s_minor/minor_gazo/SG-50-s1.gif
or https://www.diyaudio.com/community/threads/type-50-tube-amp.335913/post-5745013
which seemed promising considering that the follower presents an output impedance of less than 1kohms.
Another example was the Loftin-White circuit, which is simple enough. The way I don't like cathode follower is that it requires a C- power supply to swing the output to negative grid voltage (if I don't want to waste power on the 50 cathode bias), which means setting up another voltage rail. And Loftin-White may burn the precious 50 tube if the driver tube's absent or not heating enough to pull down the grid during startup.
But as I learned from other driver circuits, there're plenty of driver tubes like 5842/417A, D3a, EC8010, etc. that will deliver 70Vpp to a 10k load in common cathode. Even parallel 5687 or 6SN7 would do the job just fine. Another possibility might be using other topologies. As I calculated, even the weakest 12AU7 would supply a 70Vpp swing just fine. Swapping the tube to mightier 6CG7/12BH7/6H30, and a 100Vpp output with minimal distortion isn't a dream!
So my question is: Why is nobody driving the 50 tube in these ways? Am I missing something? Would appreciate your opinions.
Last edited:
The does not look so bad as drawn.
One thing I would do is decouple the driver stage plate voltages from the output.
Look to run about 400 volts on the SRPP and 300 volts in the first gain stage. You just want to be sure the driver gets a chance for more ripple reduction and less chance of oscillation.
One thing I would do is decouple the driver stage plate voltages from the output.
Look to run about 400 volts on the SRPP and 300 volts in the first gain stage. You just want to be sure the driver gets a chance for more ripple reduction and less chance of oscillation.
Seems reasonable. Would definitely try more power supply decoupling on gain stages.
So does it mean that nobody's doing so is simply because nobody's ever done it? Rather than possibly existing defects of these circuits.
Maybe I'll try common cathode as well because I have the tube and why not? 😉
So does it mean that nobody's doing so is simply because nobody's ever done it? Rather than possibly existing defects of these circuits.
Maybe I'll try common cathode as well because I have the tube and why not? 😉