I originally posted about this in this thread:
http://www.audio-talk.co.uk/phpBB2/view ... sc&start=0
after I'd run a frequency response test, and discovered that this may actually qualify as hi-fi. The discussion started to diverge somewhat from frequency response, and so I decided to start an new thread. And, here we are.
I had a few 6M11 'Compactron' valves on hand from a previous project in which I'd lost interest. These have a pentode section and two triode sections. Plus, they're tiny, and adorably cute. My plan was to use the pentode as a preamp stage, and the two triodes as a push-pull output. The triodes have a maximum plate dissipation of 2.25 watts each, so this is not going to burst anyone's eardrums. On the other hand, I thought it might make a fun little desktop amp for the computer.
The project is still in the prototype/breadboard stage as the photo shows.
This is the original schematic before adding refinements:
A push-pull amp normally requires some sort of phase inverter, but since the pentode was already doing preamp duty, there weren't many options left. I could have used an interstage transformer, but that's a rather dated technique requiring an expensive transformer. Another option is to ground the grid of the second triode and use cathode coupling: essentially a long-tailed pair, or differential amp. That was the method I chose, sort of. However, this would have required a high resistance in the cathode (to ensure reasonably constant current) which would not be practical for biasing. The present arrangement could be called a 'short tailed pair' which doesn't guarantee constant cathode current. So, instead of grounding the second triode's grid, I connected it to a voltage divider which gets a signal from the first triode's anode. This compensates for the rather variable cathode current and applies sufficient signal to the grid to ensure the second triode shares the load.
After the original circuit was reasonably debugged, I added a tone control:
This was my own concoction after searching the web and not finding exactly what I wanted. It's not perfect by any means, but it works well enough for what I want. The frequency response is shown here:
The amount of boost/cut is not huge, but it is a bit of a compromise. More boost/cut means more insertion loss. As it is, it gives enough range to keep me happy.
The circuit as it is right now:
Aside from the tone control, I've added some components to balance the bias of the output stage. Initial tests show that it works well, but I'm not really sure that something as low power as this really needs anything this complex, and I may remove it.
Oh yes, the 'A' and 'C' in the schematic refer to the taper for the potentiometers. 'A' is audio taper. 'C' is reverse audio taper. So, the tone control requires a non-standard pot. If a C-taper is not available, a standard audio taper can be used, but wired backwards so that bass increases as the control is turned clockwise. Or, a 20k standard linear taper pot can be used (and deleting the 33k fixed resistor), but the adjustment is not quite linear.
Oh... and the schematic shows supply voltage to be between 180 and 220 volts, but I'm actually running at 250V as we speak.
Feel free to post your comments, questions, ridicule...
