This design is a refinement of one I designed and built about 15 years ago. In most respects the front end is not that unusual, being a derivative of a Krell front end. The most unusual feature is the use of separate feedback loops for the voltage gain and output stages. If there is interest I can post additional details and a circuit description.
The electronics have been designed and simulated, and the PCBs will shortly be arriving. I'll be posting some photos of the mechanicals: no acceptable chassis was available, so I had to fabricate one from scratch.
Jeff
The electronics have been designed and simulated, and the PCBs will shortly be arriving. I'll be posting some photos of the mechanicals: no acceptable chassis was available, so I had to fabricate one from scratch.
Jeff
The schematic is hard to follow as it crosses multiple pages. Did you verify with a prototype?
I have concern about your DC servo. It is through positive input. An input capacitor will contribute an extra pole and it could oscillate.
I have concern about your DC servo. It is through positive input. An input capacitor will contribute an extra pole and it could oscillate.
Regarding a prototype: yes. That design I built over 15 years ago, and it has been operating flawlessly since then. This design is a refinement thereof, where the only significant changes are to the VAS and diamond buffer stages. I have simulated with both DC and AC coupling to the input and have encountered no oscillation problems.
Thanks for the notes on the servo. I'll check to make sure that the pole added by an input capacitor remains in the left half plane.
Anibal,
No, they are correct. Q7/Q8 form a constant current source between the +70V rail and ground, Similarly, Q9/Q10 form a CC source from ground to the -70V rail. In both cases the current flow is from the most positive to the least positive potential.
No, they are correct. Q7/Q8 form a constant current source between the +70V rail and ground, Similarly, Q9/Q10 form a CC source from ground to the -70V rail. In both cases the current flow is from the most positive to the least positive potential.
OK, my bad, I just took a quick look and I immediately thought on conventional constant current sources and neither of them are conventional.
I cannot deeply analyze your design, just let me tell you on the positive side R34 limits the output impedance of the source and in the negative side you are connecting as output the emitter Q10 + 150 ohms...
In a conventional design the purpose of those current sources is provide a high impedance to improve common mode rejection ratio and also improve power supply ripple rejection, your design works as it is; but it seems to me it does not provide high impedances.
I cannot deeply analyze your design, just let me tell you on the positive side R34 limits the output impedance of the source and in the negative side you are connecting as output the emitter Q10 + 150 ohms...
In a conventional design the purpose of those current sources is provide a high impedance to improve common mode rejection ratio and also improve power supply ripple rejection, your design works as it is; but it seems to me it does not provide high impedances.
