Dear Sir,
I want to derive the Third order Input Intercept Point for my low noise amplifier (LNA), which is working at 2.5 GHz.. How to derive the third order input intercept point of my LNA, what I have to follow, where I will get the detailed expression for deriving LNA and how to start the deriving expression for LNA. Can you advice me.
Herewith I attached the example LNA diagram.
Deriving the IP3 from theory will be unnecessarily time consuming and the final results may be inaccurate. It will be easier to determine your preamp IP3 by harmonic balance simulation. You will also need to confirm the simulated IP3 by measurement. This will require two signal sources combined to provide a two tone input to the preamp. The preamp output spectrum third order products then allow the IP3 to be determined.
I will try to derive IIP3, if anybody knows documents about the IIP3 derivation, how to derive basic expression for IIP3 from basics. Please can you attach the help document for deriving IIP3 for LNA.
I would expect that this is potentially a very difficult thing to determine and will very much depend on what you mean by 'from basics'. I have not looked into doing this myself but I would expect the key thing would be to get an accurate model for your devices, whether you are going to use simulation or from just an analysis of the models and the circuit topology and values.
The model non linear characteristics will need to be a very close match for your actual devices. I guess what you are then interested in is the 1st and 3rd order derivatives of the power transfer function. I would expect you can use/derive the voltage transfer functions but you will then need to specify the load impedance which is not shown in the diagram (50 ohms?).
As an aside, I presume that the output 772pH L and series 247ff cap are an output matching network - they do appear somewhat extreme at approx. 12 ohm shunt and 258 ohms series - I would expect that this is quite a narrow band design.
Apologies if this is all stuff you are already aware of but I haven't looked into this before - I generally use vendor models in ADS and use high efficiency/linearity PA techniques to maximise the IP3 in LNAs - comes down to a compromise between noise figure, input return loss and IP3.
I am presuming that literature searches regarding deriving IP3 are not successful.
- Badges
- Science topic
- Similar topics
- Political Science
- International History and Politics
- Globalization