Am going to try the P/follower both cascode and single. My bet is the cascode will work better.
Douglas
Douglas
Some comments on the first circuit of the thread after almost four years:
- UL configuration is not optimal even for classic audio tubes, and it was just serendipity that having same amount of a-g2 and a-g1 feedback I got better results on simulations. In reality a pentode configuration sounds better and give better triode curves (more current available where needed). Never tried super pentode, but it could be interesting to try.
- A dissipation of 35W is unrealistic for a TO220 source follower and it is better to stay around 10W of dissipation maximum.
- Better to get local feedback of the driver from the output of the CCS rather than the anode.
- Bootstrapping the load gives more stability and the possibility to have feedback on driver's g1 with a voltage divider from driver's anode or output tube's anode (and drive the output transformer with few tens of Ohms).
- Driver's pmosfet can be a VN0106 (that is what I used almost every time) or pjfets.
- The bottom of the voltage divider that brings feedback to g1 can be connected to ground or to another low impedance source of feedback to combine different feedbacks together.