RF amplifier - design suggestion

[yawijaya]

Hi all,

I am designing a RF amplifier. I would like to get some suggestion input about what type i should use and possibly design reference. Here's the requirement:

1. output power 2W to 50 ohms load
2. Frequency range 50MHz-120MHz
3. Low 2nd harmonic
4. high efficiency

I made one before but the efficiency was only 20 percent or so using class E RF amplifier. SO i'm now open to any suggestion.

Thanks!!!

 

[vfone]

You didn't specify the power class of the amplifier, or at least the modulation type of the amplified signal.

 

[yawijaya]

well, that's basically what my question was. I am looking for suggestion on what power class i should use. I've tried class E before, but they are only good between 50M-70Mhz.

 

[vfone]

The best efficiency power class is Class C, but this class is very nonlinear and can be used only for CW or FM.

Class AB is pretty linear. On this power class tuning the bias point and load-pull the output match can get minimum of the harmonics and best efficiency.

 

[rezaee]

if u want high efficiency then u must select switching mode amplifier same class E OR F OR D. which u want to design? amplifier or power amplifier?
which class E circuit topology you use for your designing? do you simulate your schemitic or no?

 

[yawijaya]

I've designed it using HEPA PLUS (class E power amp software) which simulates the schematic and worked very well ~70% efficiency, but the circuit sucked in reality. I am designing power amplifier. I also forgot to mention that the circuit needs to be able to give up to 200-300mA output current (aside form the 2W at 50ohm load)

 

[rezaee]

with which type of elements you simulate your designed circuit in HEPA PLUS?? i means is you use ideal elements or use realistic elements model?
also i am trying to design a class E power amplifier but i use ADS software ofcourse my design is for IC i think if you use realistic model for simulation your result dont be very differ.
which topology you use for your fully circiut schematic?

 

[yawijaya]

it was realistic elements model...

 

[rezaee]

Hmmm... 70% to 20% !
which class E circuit topology you use?

 

[yawijaya]

i'm not sure by what you mean with topology...but here's the schematic...

 

[rezaee]

i see your schematic diagram of course somethings isnt clear in your designed circuit schematic but i think its better to use matched circuit instead of transformer and also use match filter between deriver stage and amplifier stage and check your elements values with using good fitting analytical equations and also notice the breakdown voltage of your transistor and the other way is to check your circuit point to point with scope for example do you see desired output at the adc(square wave)....

 

[vfone]

Class E is hard for practical implementation for someone with no experience.

Class AB it will work right away, allowing a lot of tuning of initial design.

Now depends what you are looking for. To learn, or to get quickly a final product.

 

[FvM]

Your results simply show the limitations of designing a RF amplifier by just using a software. Assuming, that the said circuit
simulation is correct so far, some real circuit parameter is obviously different. But which one? Experience is needed to identify
it and only empirical measurements can verify the design.

 

[yawijaya]

Class E is hard for practical implementation for someone with no experience.

Class AB it will work right away, allowing a lot of tuning of initial design.

Now depends what you are looking for. To learn, or to get quickly a final product.

Ha, well, ideally i want to learn and get a final product quickly...but i guess i have to choose one of them...I guess i'll lean toward quick final product...so you say class AB?
but i also read before that class AB does not perform well for low signal application...

 

[vfone]

Surprisingly Class-AB offers wider dynamic range than any other power class, including Class-A. This because saturation comes from different additional sources.
In Class-AB tuning the bias point (quiescent) for optimum conduction angle leads to better efficiency and linearity.