CC1125: S-parameters of RF pins

Thank you. S-parameters from that thread are measured from 500 MHz to 3 GHz, but I need data for frequencies from 400 to 900 Mhz.
I also need S11 of PA out in TX and RX mode.

You wrote that you want to simulate in ADS. But why do you want to do so?

The s-parameters in the post I linked to is what we have.

I design two-band (433 & 868) device based on CC1125 transceiver with external power amplifier, two-band PCB antenna and switchable filters in RX path. I can do this only with ADS.
For accurate design and simulation i need s-parameters of LNA in RX mode and PA in both modes (TX and RX). Or, at least Smith charts for these pins.

I analyzed "0243.LNA_P_RX.S1P" (from thread from your link) for 868 MHz point, and complex impedance of this pin (according to this s-parameter model) at this frequency is 16.8-j*80.3. But recommended by the datasheet source impedance for LNA in single ended mode for this frequency is 30+j*30 Ohms (that corresponds to 30-j*30 impedance of this pin).
Which real complex impedance have LNA_P and LNA_N pins at 433 and 868 MHZ?
Also, which complex impedance have PA pin in TX mode? In RX mode (acts as small capacitance? or others?)?
Which complex impedance have TRX_SW pin when is off (in RX mode, acts as small capacitance?)?
Or, may be, there are data on the capacitance and inductance of these pins?

That is basically one of the reasons we are reluctant to give out s-parameter files.

In the text books they often use complex conjugated match. First of all, this may only work for a LNA. The PA in TX goes close to rail-rail and is far from small signal meaning that s-parameters are not valid. Text books also match PA either for max gain or max output power. In the real world the matching also has to take into consideration harmonics and power consumption. And when the LNA and PA share pins the impedance has to be a tradeoff between what the TX and RX wants.

That is why we give out the optimal impedance, the impedance the RF pins wants to see looking into the external network. I would advice to import the layout of the reference design in ADS and simulate this with a high impedance (1 MHz) termination of the either the LNA or PA pin(s) dependent on which one you are not using. For the TRX pin use the s-parameter files supplied.

Combining 433 MHz and 868 MHz in one design is challenging. The LNA require 0 V DC bias. Could be that a wideband balun could be used since I assume you don't want a large amount of switches in the design.

For converting differential LNA inputs i use wideband transformer ETC1-1-13 that already provides ability to apply any required DC level to differential side. After it all of the RX path is single ended.
I didn't see the S1P file for TRX pin in that thread, sorry, thanx, now i see it.
One question left - what is the behavior of PA pin in RX mode? What is the complex impedance of it while receiving?

The PA is in power down while the device is in RX meaning that the PA input is high impedance.

As I wrote in a previous post: 'The s-parameters in the post I linked to is what we have.'