6C33 single ended class A2 (cathode network question)

[jane]

I am doing a 6C33 single ended class A2 amplifier.
The circuit is simple: Choke loaded cathode follower driver with bootstrapped input/gain stage.



My main concern is the cathode network for the 6C33. How could I ensure that the network is stiff enough for class A2?
I suppose that the extra class A2 current (grid current) in the positive period of the signal will charge the 470u capacitor, and as a result, the cathode potential will rise..

My question is: what should I put inside the box named "120v source" to keep the cathode voltage steady?




Jan E.

[M Gregg]

Well there is all sorts..LOL

just for fun link

Fixed Autobias?

Or CVS ..or mosfet amplified zener (mozener) simple and works well..

There are many ways of using it...

Then there is the Baby Huey ideas...


Regards
M. Gregg

[DF96]

Grid current should only be a small proportion of total cathode current, so in itself it won't create a bias problem. The real problem is that you have two conflicting aims: bias stability with valve ageing, and bias stability with signal level. The first is solved with CCS bias. The second is solved with constant voltage bias. These are opposite solutions! The cathode resistor is a compromise between these two extremes. I think you just need to put a fat capacitor across the resistor.

The reason for the problem is that all valves suffer from second-order distortion. This does not just create harmonics and IMD, but also a DC shift - extra current. Voltage bias ignores this, as it should. CCS bias reacts by increasing voltage which means the valve runs cooler during brief zero signal periods, so you either accept this or bias the valve hotter to start with. A cathode resistor is a compromise. This is one of those rare cases where the simplest solution is almost the best solution. The ideal solution is a servo which only measures the valve current during zero-signal moments.

[jane]

Imbalance in cathode current is equal to (second harmonic) distortion.
Looking at the Ia/Va curves for 6C33 idling at 180mA into a 1k25 load shows max/min current in class A1 to be ~340mA/60mA.


For a steady sine wave the average cathode current will be pulled up from the 180mA idle. 340ma/60mA should give an average of 200mA cathode current but due to dc-feedback in the cathode resistor this will be limited.
Spice simulation of Ik and Vk in class A1 (ig=0), 1k sine input:


Same circuit in class A2 (ig=20mA):



In class A1 Vk will climb 4-5V and in class A2 (ig=20mA) Vk will end at +7-8V according to Spice.
I am not sure if the cathode voltage is stiff enough?

Jan E.

[M Gregg]

OK I know this sounds strange..

What happens if you put a zener across the cathode capacitor in the simulation?


Regards
M. Gregg

[flatheadmurre]

As M.gregg says a zener will keep the voltage in place when leaving Class A.

Using a Mosfet instead of resistor you get a resistance in the milliohm region and don´t need a big cap.

Cheapest solution i found sofar for constant katodvoltage is the Mozener.

[DF96]

As long as your music only has occasional signal peaks a bypassed resistor will be fine. If you like listening to full volume sine waves or clipping then you need a proper servo.

Putting a zener in may stop the valve adjusting itself to the DC output from the driver stage.