More 6SJ7 Discussion (Variable Screen Voltage)

[brewdude]

So... with some of the recent threads having interest in pentode preamp tubes and there operating conditions I thought I would offer start a new thread to offer some of my experience. I must first state that my amps are not conventional designs (although there is nothing revolutionary). Most of the designs parameters have been achieved by trying to incorporate some of the lessons in Merlin Blencowe's books.

The pertinent part of this amp's peculiarities is the use of a VVR to control the voltage of the 6SJ7's screen grid. When I first built this amp I used the tube data sheet to first draw adjusted load lines for the chosen voltage of 50v to be around the center of the VVR's travel such that I could raise or lower the voltage by 20-25 voltes--making screen voltage adjustable between about 25v-75v. I used the VVR with adjusted lower limiting resistor value to reach the desired minimum and used resistors in series after to set the maximum limit.

I chose a plate load that put the load line through the knee at screen voltage of 50v and then calculated a cathode resistor value.

This amp is not using the pentode as V1. It has a parallel 12AX7 gain stage in V1 that feeds an early Fender style TB tone stack and gain control before the pentode. There is a MV after the 6SJ7 and before a cathodyne PI using a parallel 12AU7 which feeds a pair of 6V6's. Also of note is that this amp employs a choke input power supply.

I realize that there is little interest in this combination of features, but feel it is necessary to disclose the oddities of this amp incase it effects the observations and subsequent conclusions.

Also, I have made many adjustments to this amp from its original design, but I believe the following schematic is the correct representation of the final design.

Please take a look ask a question, give a comment or criticism. I welcome it all.

Edit: looking through my notes I was reminded that I originally used 75v at the screens to draw the load lines, not 50v and it was variable from about 50-100v
Rp = 56k (163v), Rk = 680R (2v) at Vg2 @ 75v
Then through experimentation ended up at the conditions described further up and found in the attached schematic.

[tubeswell]

Nice schematic.

The 12AU7 cathodyne bias voltage generated by the 3k3 Rb should enable a bigger input signal without distortion, but the 3k3 Rb adds significantly to the 22k load resistor at the cathode, which could affect the output balance. Maybe you could try a fixed bias input and have a more balanced load? (You will be getting heaps of interesting textures from the 6SJ7 anyway). Over to you.

Another thought is you could just have a single 12AU7 triode* for the cathodyne and use the other 12AU7 triode for a virtual earth input driver stage in front of the cathodyne. A virtual earth stage would have low input sensitivity - which would help capture the pentode sound better without adding extra distortion, which could then feed the cathodyne. YMMV

* there will be more than enough current from this to easily drive the 6V6s

[brewdude]

Thanks for the input tubeswell

Nice schematic.

Thanks, I like colors. When I updated my Mac OS a while back, the program I was using to draw schematics wouldn't work. I bought ElectricDesign from the Apple App Store. It seems to do what I need. This was my first drawing on this program. I've made a few others since. Its getting easier to figure out how to do what I want... I always wonder what others think of the format/layout of the drawings. For me, the addition of colors makes it much easier to follow and understand.

The 12AU7 cathodyne bias voltage generated by the 3k3 Rb should enable a bigger input signal without distortion, but the 3k3 Rb adds significantly to the 22k load resistor at the cathode, which could affect the output balance. Maybe you could try a fixed bias input and have a more balanced load? (You will be getting heaps of interesting textures from the 6SJ7 anyway). Over to you.

I have thought about adjusting the other resistor to 27k in an effort to balance the cathodyne, but wonder how much of a difference and would it make, and would it be for the better. In the end I went path of least resistance and did not try a change.

Is there an advantage to a fixed bias input?
Would it behave correctly when power scaled with a VVR?

Another thought is you could just have a single 12AU7 triode* for the cathodyne and use the other 12AU7 triode for a virtual earth input driver stage in front of the cathodyne. A virtual earth stage would have low input sensitivity - which would help capture the pentode sound better without adding extra distortion, which could then feed the cathodyne. YMMV

I'm not hip to the virtual earth input driver. Does low input sensitivity mean that it will handle input levels that can vary over a wide range? Or, it could handle the high gain of the pentodes output?

My understanding looking at the 12AU7 data sheet was that if relatively center biased, it would take a very large signal voltage to cutoff or clipping. Did I misunderstand?

* there will be more than enough current from this to easily drive the 6V6s

I was also under the impression that although the 12AU7 cathodyne wouldn't supply any gain, that it would provide plenty of current with a single triode, and figured with a parallel pair of triodes it should provide double the current.

I love this amp. It can achieve many colors/textures/tones, yet I find the controls very forgiving. It can be very dynamic and touch sensitive or somewhat compressed and fuzzy with a few front panel adjustments. It can be set to emphasize pre amp overdrive or power amp overdrive, and I can get the tone I want at many volumes.