Biasing C class in Doherty amplifier

[Young_Electronic_00]

Hello everyone,

I want to consult with you my doubts which appeared during design of Doherty amplifier. I'm designing simple configuration consisting of AB and C amplifiers. Schematic of single amplifier which I used for a carrier and an auxiliary amplifiers looks as follows:

So, for such a circuit, I calculated IV curve to choose biasing (presented below). The pinch-off voltage of the transistor is -0.65 V.

I don't have bigger problem with AB class (Vds = 3.75 V, and Ids = 26 mA what corresponds to Ugs = -0.54 V), however I'm totally confused with C class. I looked through several publications, books, and I've noticed different information how to choose biasing point. On this moment, I'm testing two of them. In first one I choose Vds = 5V, and Vgs = -0.95 V what is signifcantly under the pinch-off voltage (this corresponds to point C on the graph), and the second one: Vds = 2.5 V, and Vgs = -0.7 V (this is almost B class but for different load-line). So I'm confused, because after design of single amplifiers, and use them in Doherty configuration I should see Doherty region, right ? Something like this should be observed:

I see improvment of linearity (comparing to B or AB classes) however, I don't observed such a region...


So, how the biasing of a class C amplifier should be done ? How should the operating point for the HEMT transistor be selected?

Regards,

E

 

[vfone]

What transistor is Q1?

 

[BigBoss]

C-Class Amplifiers need pretty high RF Driving Levels. Maybe you make a mistake in this regard..
Does C-Class start to conduct just after AB-Class reach its \[ P_{1dB} \] or \[ P_{3dB} \] or whatsoever ?

 

[vivekroy]

That curve is terribly hard to obtain especially at mmwave frequencies. You need to compare your class AB with Doherty to see how the back-off efficiency.
As for biasing the class C, go for an adaptive bias (detect the input power and increase the bias voltage). For Doherty, I have found adaptive bias to work much better than a fixed bias for the class C stage. Can you maybe show how you are doing the power combining?