• E3443.png

    WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.​

Designing for guys who like designing

This time I'm not going to be so easy on you guys, I'd like to hear theory of operation from you :)

This will drive a zero-bias 6V6 PSE (160Vp, 150Vs). Should sound good if the curves are any indication. Feedback comes from the plate rather than speaker output, since I want the I/O isolated to protect anything from the line-operated power supply. ;)

The goal for bias is Vo=0V. Speaking of which, the diode and resistor are represenative of the grid (25mA at 15V), not actual parts.



  • dc coupled cf driver.gif
    dc coupled cf driver.gif
    3 KB · Views: 553
You're welcome :)

The concept with the 220k was to stabilize DC bias. If the output rises say 1V, the rise forms a voltage divider with the bias pot, putting approx. 1/3V on the input. This is inverted and amplified to about -10V at the plate, and maybe -8V after the CF (I haven't calculated gain and distortion for the CF yet), more than countering the original +1V error. That's near what, 20dB of DC NFB?

I could just as well go without it, but I figure it will help stability in all conditions (startup, etc.). Or I could go cap coupled between the input, preamp and CF, but that's not as interesting :cool:


what is the measured dc potential at the input grid? I concerned that the input transformer may short especially if bifilar wound.

Otherwise it's a neat way to achieve two direct coupled stages from a single plus/minus supply, rather than the more easy method of direct coupled cf drive using two different supplies for each of these stages and cap coupling the common cathode to the cathode follower.
Hugh? 3 direct coupled stages? Do you count the input transformer as a direct coupled stage or is there a stage I missed?

With regard to shorting input transformer, I blew up a Sowter TA that way, just advising caution and check with transformer supplier. Brian Sowter blamed me for the problem and I had to pay for another transformer, his are bifilar wound and cannot take any dc offset.
Re: On third thoughts....

dhaen said:
The input transformer's secondary is at the 6V6 control grid voltage.

Actually, if the bias control is set at -150V and the output is at 0V, the two grid leak resistors divide the voltage to -100V.
You'll note that if the grid were at 0V, the plate would have to be pretty close to 0V as well.

I'd love to breadboard this thing but I don't have a 300VCT power supply, hmm....

Maybe you should try say, 12AX7 or 6AU6 for gain, and a sufficiently big triode for the CF, just cap couple the stages so you can get enough signal going. (Neg. supply still needed for the CF of course - unless you use a choke load.....)
Bias to the output tube by varying bias either to CF or output tube grid (when using choke; note this uses a coupling cap between CF and grid :) ).

Yes that is what I am doing using the 6em7. section II is the choke loaded (plus resistance to assist juggling the output potential) direct coupled cf driver for the 212 output stage. The grid of stage II is grounded via a 100k voltage pot which enables me to fine control the bias on the 212.

The circuit is nothing new or special, and it hasn't enough gain (single va stage and no step up transformer) to drive the 212 to the full 40 watts, but it can party at over 20 watts effortlessly.

As I said what I liked about your idea was the possibility to use a single power supply for that part of the circuit. I am using one power supply for section I and the +/- supply for section II. Also I must isolate the two sections, but for that I have some Russian Teflon caps so they shouldn't be too big a loss of quality. The important stage for direct coupling is the driver to power stage. That is as long as safe operating points for sturdy valves are chosen.