LNA power gain in RFIC 0.18um processor
make accurate impedance match
Select a loading inductor with a higher Q value can improve power gain.
Or use cascode type to increase isolation.
Last, increase current also can improve some performance
put 2 stage
Hi all just one comment to all the answers.
Conjugate impedance match at LNA is not the thing to do. An LNA should be optimized towards noise figure. Its important that the first stage is the best in point of noise figure.
I don't understand the comment by cbirdman to
Please explain, do yo mean that one shouldn't go for a conjugate match at the output ?
Regards,
StoppTidigare
to StoppTidigare:
Yes ,I agree with you !
The noise figure of LNA is crucial for the receiver ,but meawhile the mixer
contribution to the system noise figure depent on the LNA gain value,so
the LNA gain is too important for the receivers.
I think high Q inductor ,have low loss,because theoretically capacitor and
inducor do not consume power ,but in practice have parasitical resistor ,it
consume power.
About matching ,I think that is necessary for LNA design ,even for RF
Hi,
There are two types of matching.
a) for NF
b) for Power
The difference in power gain will be normall around 0.5-1dB. This may not be really significant in many applications. So it is better to go for noise matching always.
Use ceramic single layer High Q capacitors
Low loss inductors
Madhukar
Because the LC tank parallel resistance in resonant frequency is propotional to Q value of inductor.
to Madhukar
1)For noise matching ,how to noise matching in Spectra RF or ADS .
2)When the circuits is for the noise matching ,the S11 may be larger -10dB, which is the request commonly.
thank you
sjyatou
[quote="sjyatou"]to Madhukar
1)For noise matching ,how to noise matching in Spectra RF or ADS .
2)When the circuits is for the noise matching ,the S11 may be larger -10dB, which is the request commonly.
thank you
Hi!
Using ADS is very simple to do that.
1)To get the input matching network for NFmin:
a) You just have to extract NFmin of the device vs Vbias for the working frequency.
b) Next with NFopt reflection coefficient uses smith chart utility from the ADS palette to generate the input matching network automatically.
c) Then choose the right input matching network to generate NFopt reflection coefficient to get NFmin.
d) Build ADS... to put the input impedance matching on schematic... and... you are done!
If you have any doubts... reply...
2) -10 dB it's okay! If you have a filter or an antenna behind the LNA in the receiver chain and coefficient input reflection, S11; with more than -10 dB can generate distortion on filter characteristics.
Normally you have tradeoff between NF, Power and S11. So choosing the right device size and bias is important.