How to Design a Class C Power Amplifier in ADS?

[Bailey_NTF]

Hello everyone,

I am currently working on designing a Class C power amplifier using ADS (Advanced Design System), but I’ve encountered some challenges and would like to seek advice from experienced users.

Here’s my current situation:

1. I am able to use Load Pull simulations to design Class A and Class AB amplifiers successfully.
2. However, when I attempt to use Load Pull to simulate a Class C amplifier, I cannot achieve the theoretically high efficiency (close to 100%) that Class C amplifiers are known for.
3. I’ve noticed that ADS provides some Design Guide tools, but I’m not sure how to effectively use them for designing a Class C amplifier.

I have a few questions I’d like to ask:

• Could anyone share the specific steps or workflow for designing a Class C amplifier in ADS?
• Are there any particular considerations or tips when simulating a Class C amplifier?
• Does the ADS Design Guide offer any recommendations or templates for Class C amplifiers? If so, how can I use them?

Additionally, I’d like to mention that the reason I’m designing a Class C amplifier is because I’m working on a Doherty amplifier project.

I’d greatly appreciate any help or advice! If you have any related examples or documents to recommend, please feel free to share them. Thank you very much!

 

[BigBoss]

C-Class Power Amplifiers request relavitely Input RF Power. Therefore the Efficiency is a function of Current Flow angle and Input RF Power.
You have try this amplifier under different operating conditions. You should see that the Efficiency will rise up with RF Input Power then drops slightly.
A simple Load-Pull with lower RF Input Power won't give you an good insight for Efficiency.And every Input RF Level will change Optimum Loading Condition.
Designing such amplifier needs an iteration procedure.

 

[Bailey_NTF]

I have already designed a fully functional Class A amplifier operating at 30 GHz. How should I adjust the voltage to convert it into a Class C amplifier?

 

[BigBoss]

I have already designed a fully functional Class A amplifier operating at 30 GHz. How should I adjust the voltage to convert it into a Class C amplifier?

You talk about two different classes. C-Class Amplifiers have negative bias that is achieved by generally auto-biasing techniques.
RF Signal Current flow is less than 180 degree. So they are completely different.

 

[D.A.(Tony)Stewart]

There is a tradeoff for gain, power added eff and conduction angle.
For θ→0° η→100%, but gain→0, making big pain with no gain.

η=100% is a myth at 20 GHz, maybe PAE=80% or 90% with a lot of tricks.

- low loss tangent duroid 5880
- choose Ropt≈2Vdd/Ipk
- use short harmonic stubs to recycle lost energy e.g. λ/8 stubs at 40 GHz (maybe)
- define your required Pout and choose RdsOn, Cds wisely ($ if you can afford > 20 dBm), GaN isn't cheap.
- add a pre-amp if to reduce gain to increase PAE might be small gains.
- tighter tolerances on impedance control ($)

I have limited expertise and no ADS, but I can read possible solutions may include :
- big budget
- CGH40006P
- Vds = 28 V, Vgs = -3
- gold-plated, ultra-smooth microstrip or with parallel PCB iterations of stripline to find optimal performance since tweaks are hard if not impossible.
- pay for controlled Zo TDR in PCB tests to tighter tolerances
- months! of tweaking parasitics, thermal effects, high Ropt and stub misalignment ( or faster if you plan for failures)
- clean room assy
You get an atta boy if you reach 90% PAE and a gold star is 93%.

 

[Bailey_NTF]

您談論的是兩個不同的階級。 C 類放大器具有負偏置，這是透過一般的自動偏壓技術來實現的。
RF訊號電流小於180度。所以它們是完全不同的。

I read this paper, and it uses the same circuit to create both a Class A amp and a Class C amp for a Doherty circuit. If I can't use load-pull to analyze the Class C circuit, how should I determine the optimal impedance for the Class C circuit?

 

[D.A.(Tony)Stewart]

The schematic leaves out alot of FET parameters like Vt and chemistry {Cox, Si, GaN, etc} and the desired gain v. effic. tradeoffs with no output power specs, thus your question is indeterminate.

If you assume the same bias Vgp and Vt of each IGBT,
IGBTs for A and C have an RdsOn ratio A:C = 2:1 yet both use the same Vdd thus Imax for A:C = 1:2 approx neglecting Vt~Vknee

Then start with Ropt,A = (2Vdd-Vt)/Imax
let this = Ropt,C then step higher to compare gain, effic. vs RdsOn

 

[BigBoss]

I read this paper, and it uses the same circuit to create both a Class A amp and a Class C amp for a Doherty circuit. If I can't use load-pull to analyze the Class C circuit, how should I determine the optimal impedance for the Class C circuit?

First, you have to be sure that C-Class Amplifiers is working well from all aspects.
Then, you must define a RF Power Level where the auxiliary C-Class Amplifier start to amplify (simultaneously A-Class Main Amplifier starts to get into hard compression or beyond). After that, you can separate the amplifiers and analyze independently for Optimum Load Impedance.