class E power amplifier impedance matching

[goatmxj666]

Hello

I designed Class E power Amplifier with CMOS.

While doing impedance matching, I had some questions.

In some papers,
when doing impedance matching, the shunt capacitor existing in the standard class E PA structure is left as it is, and then an LC structure (L-match, T-match etc) is added.

However, in another paper,
the shunt capacitor is removed and the capacitor generated by LC matching is used as a shunt capacitor.

As in the second example, I wonder if it is okay to use the capacitor generated by LC matching as a shunt capacitor.

Also, I wonder if this LC impedance matching method is mainly used in practical RF circuit design.

Thanks!

 

[rd_dinesh]

Hello

I designed Class E power Amplifier with CMOS.

While doing impedance matching, I had some questions.

In some papers,
when doing impedance matching, the shunt capacitor existing in the standard class E PA structure is left as it is, and then an LC structure (L-match, T-match etc) is added.

However, in another paper,
the shunt capacitor is removed and the capacitor generated by LC matching is used as a shunt capacitor.

As in the second example, I wonder if it is okay to use the capacitor generated by LC matching as a shunt capacitor.

Also, I wonder if this LC impedance matching method is mainly used in practical RF circuit design.

Thanks!

Is the shunt capacitor you mentioned here, is the internal capacitance?

 

[goatmxj666]

I mean, there are papers where the Cshunt capacitor is still there and the Cm is added due to impedance matching, like in the structure shown below.

But in another paper, there was a class E power amplifier with no Cshunt capacitor and only a Cm capacitor. Is this case a bad design? Or is the Cm capacitor acting as a shunt capacitor?

A. Mazzanti, L. Larcher, R. Brama and F. Svelto, "Analysis of reliability and power efficiency in cascode class-E PAs," in IEEE Journal of Solid-State Circuits, vol. 41, no. 5, pp. 1222-1229, May 2006, doi: 10.1109/JSSC.2006.872734.

 

[vfone]

Class-E employs a single transistor operated as a switch. The drain voltage waveform is the result of the sum of the DC and RF currents charging the drain-shunt capacitance Cshunt which is in parallel with transistor internal capacitance Cdd.
If the value of Cdd is greater than Cshunt, Cshunt will not be used and the excess output capacitance of the transistor should be tuned out by part of the drain bias RF choke Lck.
Cs and Ls+Lm resonate on the fundamental frequency, and Cm is used for output matching.

 

[goatmxj666]

Class-E employs a single transistor operated as a switch. The drain voltage waveform is the result of the sum of the DC and RF currents charging the drain-shunt capacitance Cshunt which is in parallel with transistor internal capacitance Cdd.
If the value of Cdd is greater than Cshunt, Cshunt will not be used and the excess output capacitance of the transistor should be tuned out by part of the drain bias RF choke Lck.
Cs and Ls+Lm resonate on the fundamental frequency, and Cm is used for output matching.

Thank you very much!
as you mentioned, I found some papers that have analysis about shunt capacitor and parastic capacitance.

I have one more question, as you said

the output capacitance of the transistor should be tuned out by part of the drain bias RF choke.

If I increase the inductance of RF choke, then parastic capacitance will be decreased?

If so, what's the principle of chaning inductance of RF choke for tuning?

Thanks a lot!!!!

 

[vfone]

The RF choke value should be tuned for best compromise for PA efficiency and output power..

 

[D.A.(Tony)Stewart]

3 variable DC gate biases with lots of nonlinear impedance and resonant filter, looks rather unstable.

 

[goatmxj666]

The RF choke value should be tuned for best compromise for PA efficiency and output power..

Okay Thanks!

I understand to reduce current flows to the TR's parastic capacitance, I have to increase the inductance of RF choke.
But This affect to the system efficiency, so I have to tune the RF choke in optimal point that satisfy the balance of efficiency and capacitance.

Is this right?

 

[vfone]

Yes, it is right. Usually the choke value for best efficiency is smaller than the value for max output power.

 

[goatmxj666]

Thank you so much!
Your answer is very helpful.