What do you think about that simulation?

Dear All,

I am using a software of keysight which call ADS and I am simulating a balun with impedence matching. This is the circuit: (It shows in image ).

It is a balun which links a differential mode to single ended and after there is a impedance matching.In one side there is normal impedance 50 Ohm instead at the other side there is the impedance of antenna which is not adapted

Now, I wanna obtain the S-Parameters, that is S_(11) and S_(12) similar to datasheet of antenna. (See datasheet antenna the graph where stay a lot of curves in different colours).

S_(21)-------->1.5 GHz corresponds -25 dB

I did two simulation.
One -is when at left side is linked two resistance of 50 Ohm
S_(21) ------>1.5 GHz corresponds -21 dB
Two-is when at left side is linked one resistance of 50 Ohm
S_(21) ------>1.5 GHz corresponds -18 dB
Now i think that the first way approaching to datasheet graph but it s not exactly.

What do you think about that? Is it correct my simulation?

LC baluns like the shown circuit are designed for small band only, e.g. 915 MHz or 2.45 GHz. Also the antenna impedance is surely valid for a specific operation frequency only. It makes no sense to perform wide band S parameter simulations of the circuit. You are mentioning CC13xx family, if you refer to the multiband devices they have separate transceiver ports for sub-GHz and 2.45 GHz.

Firstly, you need to specify the operation frequency.
Secondly, the antenna impedance in the simulation circuit is far from 50 ohms and need a specifically designed matching network. What you presently have looks more like a 50 ohms matched harmonic filter.

-Balun is not symmetrical and unusual ( i don't know how it works and what the Phase Relationships are maintained )
-Single Ended simulation set-up is wrong because if you divide the circuit by , you have also to divide the Port Impedance by so each port should have 25 Ohm internal impedance.
-Insertion Loss from 50 Ohm to Antenna is too much even with ideal components,There might be some design errors ( Balun, Filter or both )
-Ripple is to high due to design errors remarked above..

Balun component values are NOT consistent for any frequency.

The values must maintain these equations.

L=X/(2*PI*F₀) AND C=1/(2*PI*F₀*X)

Here X=SQRT ( R_IN*R_OUT)

-Balun is not symmetrical and unusual ( i don't know how it works and what the Phase Relationships are maintained )

Click to expand...

The LC balun has been apparently copied from a TI application note. It works in real circuit, but people have often problems to reproduce the behavior in a simulation, several previous Edaboard threads about the topic. The design method is e.g. sketched in AN068. One problem of the simulation attempts is a wrongly assumed of PA impedance. It's neither real nor floating. My personal experience with a 915 MHz reference design is that the component values should be adjusted for symmetrical output swing and best suppression of even harmonics.

-Single Ended simulation set-up is wrong because if you divide the circuit by , you have also to divide the Port Impedance by so each port should have 25 Ohm internal impedance.

Click to expand...

The biggest mistake is to assume a floating PA impedance, the real PA can be better described by two out-of-phase single ended outputs.

-Insertion Loss from 50 Ohm to Antenna is too much even with ideal components,There might be some design errors ( Balun, Filter or both )

Click to expand...

It's mainly caused by the unmatched antenna (10-j100) without respective matching network. Can be corrected by recalculating the double pi network, of course for a single frequency only.

Thanks for your fast answer. To analize the balun diferrretial mode can I converte the differential mode in single ended mode?
How can I do?
thanks in advance

- - - Updated - - -

Well thanks @Fvm for your attachments. I found Rdiff= 44 Ohm while Ldiff=j*15 Ohm. Now I put on the left side 50 Ohm and after in series 22Ohm (Rdiff/2) and Ldiff/(1.375nH). After I took the top branches of the balun without the inductance between two branches. After I put the impedance matching closed in to not adapted impedance antenna(Z= 9.8927-j*87.811).
Is it correct? what do you think?

The textual description isn't clear to me. You also didn't clarify about the operation frequency, 915 MHz is just a guess of mine.