LNA design problem about s2p VS spice model

[bless_fooling]

how to design lna from s2p file

Have anyone use this LNA MMIC BGA615L7 to deisgn LNA for GPS Front-end??
Give some comments about it?
or U had ever used other type component to design--tell me the type?

I can't find the spice parameters of Infineon 's internet . it only supplys the s2p
does anybody has the spice model? upload it please.

thanks in advance

 

[JOhnTNZ]

lna s2p file

MA/com does some good LNAs.
S2p usually used for RF analysis.
Doesnt look like hard part to use - pay attention to layout guidelines and part selection

 

[khouly]

bga615l7

S2P file is good for LNA , it give u all the information u need.
so use it with a linear simulator "S Paramter"

khouly

 

[bless_fooling]

lna s2p

I think the package info will be the best ,just the spice model informance.

I use s2p and spice model to simulate the S11,S21, S22, find that the spice model performance will a litter worse than s2p ,
I think the spice model will more close to the actual design on PCB board.
But I still have a doubt taht whether use s2p to simulate the K factor or use spice model to simulate??

I meet a problem that with s2p K<1
with spice model K>1

 

[RealAEL]

how does lna look like

For simulations using SPICE parameter models, even linear simulations, the device will only operate correctly if the circuit is biased. You need to add DC source etc to get the require bias conditions for the device before letting loose the AC/Linear/HB analysis.

The s-parameter model is derived from measurement at a specific bias and in that case the bias conditions are implied therefore DC sources etc do not need to be added to that circuit.

 

[GVVIN]

bga615l7

yes as RealAEL said S2p simulation is for a particular bias. You can open .s2p file in a notepad and read the value of k directly. It seems your device is not unconditionally stable at all frequencies for that purticular bias , you need to make it stable by matching circuits.

Regards
GV

 

[mouloud08]

Hi,

To bless,

As they have said and explained to you; To compare the two devices, you have to bias the spice model according to s2p biasing point. i.e, The S2p parametres are given for a specific biasing quinscent point (e.g VdsQ and IdsQ for FET, you can read the file with notepad) but for spice model and even for active libreray available in ADS or others the transistors are given without biasing...

Mouloud

 

[bless_fooling]

Thnaks all of yours good ideas
I achieved the S11 about -14dB ,S22 about -16dB ,S21 about 18.5dB NF about 0.88dB, with GPS band frequency
I have solved it.

 

[bless_fooling]

I also design the LNA with BGA615L7, the performance i think can accept.

 

[akrcal]

Hello,

The nonlinear model u have chosen for spice will need DC bias voltages also. So, you will have to bias the nonlinear model with ideal DC block and DC short which are given is ADS library. While using s2p file the parameters are evaluated at particular bias point. So, the s-parameter files are only valid for that particular bias while nonlinear model is more generic. I think this will clarify your doubt

Akr

 

[yours]

To say simply and generally:
S2p is used in the samll signal analysing while spice model is used in the big signal analysing.
Above-narrated is only my own opinion.