Hi All,
I've been reading about Ale's hybrid mu follower on the Bartola Valves website with interest. I'm sure it works very well, but it's complex enough that I would probably prefer to go all tube.
I'd like to try something a bit simpler and have a bunch of DN2540's in my junk box. Merlin Blencowe discusses a hybrid SRPP/mu follower similar to the one below in his tube preamp book, and I think it might be just the ticket. BTW I cannot recommend his book enough, it is up there with the best, a great read.
I'd be very interested to hear from anyone with experience using a circuit similar to that shown below, particularly with respect to distortion, sound and performance at high frequency. In simulation it appears pretty decent in the audio band at least, particularly with the 6SN7 below running at reasonable current.
Appreciate your thoughts.
Greg
Seek
I've been reading about Ale's hybrid mu follower on the Bartola Valves website with interest. I'm sure it works very well, but it's complex enough that I would probably prefer to go all tube.
I'd like to try something a bit simpler and have a bunch of DN2540's in my junk box. Merlin Blencowe discusses a hybrid SRPP/mu follower similar to the one below in his tube preamp book, and I think it might be just the ticket. BTW I cannot recommend his book enough, it is up there with the best, a great read.
I'd be very interested to hear from anyone with experience using a circuit similar to that shown below, particularly with respect to distortion, sound and performance at high frequency. In simulation it appears pretty decent in the audio band at least, particularly with the 6SN7 below running at reasonable current.
Appreciate your thoughts.
Greg
It's a CCS loaded amplifier stage.
It's good, but cascode CCS would be better: higher impedance loaded, lower distortion and gain more close to the tube's mu.
It's good, but cascode CCS would be better: higher impedance loaded, lower distortion and gain more close to the tube's mu.
Yes except the output is taken from the source of the fet so it can supply more current and has a lower source impedance.
Yeah I guess I'm just going to have to cascode it aren't I?! Guess I better look for some JFET's 🙁
Thanks EL506, this is a great summary of the SRPP! Unfortunately some of the technical discussion is above my head, was there an aspect to it you thought was particularly applicable to my question?
Ah yes, so it does. I have to admit I do like the high powered all mosfet topology he discusses (for use in another amp one day perhaps).
I can't speak about the hybrid SRPP but I experimented and measured the all-tube SRPP a lot. The TubeCAD article and everybody says that it is push-pull only for a certain load. When the current on lower triode increases, the current on the upper triode decreases. It is evidently impossible with a series topology, but here comes the load in the picture: the difference will exit through the load.
However, as I measured by an oscilloscope, overdiving the SRPP stage, I get a perfect symmetrical S-curve even without any load. That means its distortion is very low at low output level, and it is increasing with the output level. But the transfer curve shown on an oscilloscope in X-Y mode is similar to an FM detector S-curve. There is a long linear region, and there is "soft" limitation in the positive and negative half period when overdriven. Human hearing likes monotonically decreasing distortion with decreasing sound volume.
Anyway, it worked for me without any load (or with a 100k load, to be precise).
However, as I measured by an oscilloscope, overdiving the SRPP stage, I get a perfect symmetrical S-curve even without any load. That means its distortion is very low at low output level, and it is increasing with the output level. But the transfer curve shown on an oscilloscope in X-Y mode is similar to an FM detector S-curve. There is a long linear region, and there is "soft" limitation in the positive and negative half period when overdriven. Human hearing likes monotonically decreasing distortion with decreasing sound volume.
Anyway, it worked for me without any load (or with a 100k load, to be precise).