RF Amplifier Performance Characteristic

[@YY]

By my understanding, RF amplifier works better with clean signal (without harmonics or noise). The power amplifier gain will be degraded if the signal consist of harmonics or noise (correct me if i were wrong).

Recently I tested my newly design RF amplifier, I input the filtered (clean) signal to the RF amplifier and measured at the spectrum analyzer and power meter too. I found that the power amplifier has higher gain with the not-filter signal (around 5dB better).:shock:

Would like know how should i proceed for the investigation? is this problem due to the resonant of the component that i selected?:?:

 

[BigBoss]

The one of the most important specification of an PA is Linearity. If you drive the amplifier to where the amplifier works at nonlinear region, it will create harmonics.

 

[cariban]

How high is the harmonics? It is hard to believe that a little harmonics in the input signal would impact the gain by 5 dB. Maybe you forgot to consider the extra loss introduced by the LP filter?

 

[@YY]

Bigboss, i have checked the linearity. The PA is working in the linear region. However, would like to know is there any special knowledge about PA design? Such as the grounding, is ground guard preferable for design?
what design would be better for PA. I tried to design in microstrip before and i think co-planar seems to give better performance (which I always thought the co-planar might be better).

- - - Updated - - -

cariban, yes, it's strange. I would need further investigation, i have confirmed that the filter is only 0.5dB loss.
Just wonder if there is harmonics, i thought it might degrade the performance of the filter while clean signal will have better performance.
is there any further info regarding the PA performance to harmonics?

 

[BigBoss]

Bigboss, i have checked the linearity. The PA is working in the linear region. However, would like to know is there any special knowledge about PA design? Such as the grounding, is ground guard preferable for design?
what design would be better for PA. I tried to design in microstrip before and i think co-planar seems to give better performance (which I always thought the co-planar might be better).

How do you know that ?? If an amplifier produces harmonics, so there are some unwanted errors..
-Bias point is wrong
-Load condition is too low or too high than desired..
-Stability issues..
etc.

Can you post the schematic with bias point and noted measurement values..
If it's been designed in ADS or MWOFFICE, you can also post the project file here..