# Doherty Power Amplifier

1. ## Doherty Power Amplifier

How to determine the impedance at combining node (Rcn) of doherty sub-amplifiers? Is it a standard for all DPA designs? Or depends on the design? Thanks in advanced

•

2. ## Re: Doherty Power Amplifier

Originally Posted by Mabrok
How to determine the impedance at combining node (Rcn) of doherty sub-amplifiers? Is it a standard for all DPA designs? Or depends on the design? Thanks in advanced
This is the standard Doherty impedance that you see.
Rp=Rl' / (1 + I_carrier'/I_peaking)

Where I_carrier' is the current coming out of the quarter wavelength line in series with the carrier amplifier and I_aux is the current from the auxiliary amplifier. Rl' = Z_t^2/2R_L

This is the standard for current combining based Doherty PA. You can also have a voltage combining based Doherty which has different expressions.

•

3. ## Re: Doherty Power Amplifier

Originally Posted by vivekroy
This is the standard Doherty impedance that you see.
Rp=Rl' / (1 + I_carrier'/I_peaking)

Where I_carrier' is the current coming out of the quarter wavelength line in series with the carrier amplifier and I_aux is the current from the auxiliary amplifier. Rl' = Z_t^2/2R_L

This is the standard for current combining based Doherty PA. You can also have a voltage combining based Doherty which has different expressions.
Thank you. These equations can be used to design real DPA (to be fabricated)?

•

4. ## Re: Doherty Power Amplifier

Originally Posted by Mabrok
Thank you. These equations can be used to design real DPA (to be fabricated)?
It serves as a starting point. It really depends on your frequency of operation.

A nice 'textbook' on the design of Doherty power amplifiers was published in 2018. https://www.oreilly.com/library/view...9780128098752/

Several different implementations exist, some examples of which are:

1. Using transformers: Look at the work by Ercan et al (Here is one of their papers from their research group: https://ieeexplore.ieee.org/document/7059254). This can also be done at lower frequencies and I think their first work was at 2.4GHz

2. Using transmission lines: Look at the paper by Hua Wang's group in ISSCC 2017. Personally, I found this approach to be much better at mmwave

3. Making Doherty combiners using a more mathematical approach: Look at the work done by Mustafa Ozen's group https://ieeexplore.ieee.org/document/8103016

In ISSCC 2019, they even proposed a digital Doherty which was a very nice publication.

1 members found this post helpful.

5. ## Re: Doherty Power Amplifier

Originally Posted by vivekroy
It serves as a starting point. It really depends on your frequency of operation.

A nice 'textbook' on the design of Doherty power amplifiers was published in 2018. https://www.oreilly.com/library/view...9780128098752/

Several different implementations exist, some examples of which are:

1. Using transformers: Look at the work by Ercan et al (Here is one of their papers from their research group: https://ieeexplore.ieee.org/document/7059254). This can also be done at lower frequencies and I think their first work was at 2.4GHz

2. Using transmission lines: Look at the paper by Hua Wang's group in ISSCC 2017. Personally, I found this approach to be much better at mmwave

3. Making Doherty combiners using a more mathematical approach: Look at the work done by Mustafa Ozen's group https://ieeexplore.ieee.org/document/8103016

In ISSCC 2019, they even proposed a digital Doherty which was a very nice publication.
Thanks for the information. Do you have the idea about how to get I_carrier' & I_peaking through the simulation using ADS?

•

6. ## Re: Doherty Power Amplifier

Both these currents depend upon your load impedance. I normally designed Doherty by first designing the main amplifier, and then the peaking amplifier for optimum performance. Then I used to try and design the output combining network given those currents such that the amplifiers see the correct impedance at different power points.

Doherty is an extremely challenging class of amplifiers. There is no design procedure and people have different takes on it. Papers don't discuss the design procedure, just the results and any novelty in the combining network.

1 members found this post helpful.

--[[ ]]--