Need help in optimizing RF antenna balun on two port network analyzer

[tsunami_all]

What is the best way to optimize lumped-element RF antenna balun two port network analyzer?
S21 and S31 have to be -3 dB.
I have a circuitry that looks more or less like this one.

Current Attempt #1:

• Port 1 of NA connected to Number 1 in the schematic above. Port 2 of NA connected to Number 2 in the schematic above.
• Number 3 in the schematic is terminated with 50 Ohm 0201 size SMD component.
• Tune L to get S21 of -3 dB

In Attempt #1, I managed to get S21 of around -3 dB.
However, when I terminate Number 2 (schematic) with 50 Ohm and port 2 of NA to Number 3 (schematic) to measure S31, I got -9 dB. This is where the problem comes in.

What are the methods that people typically use to optimize lumped-element RF antenna balun two port network analyzer?
Feel free to correct me if attempt #1 is wrong.

Thanks

 

[FvM]

To make a balun, the arms must generate opposite phase shift, e.g. high pass/low pass. Besides ideal -3dB gain, you also need 180° differential phase. Deviation between ideal circuit and actual implementation is often caused by parasitic elements like trace inductance. If you consider this elements in your model, you can achieve better matching. Verification and optional adjustment using NA is nevertheless recommended.

 

[biff44]

you have to be very specific about what you mean by S21, and S31.

Are you just trying to make an in-phase power splitter, with two 50 ohm output impedances?
Are you trying to make a true balun, where the "R" is a 100 ohm differential resistance?
Are you trying to get other than a 50 ohm single ended, or 100 ohm differential output impedance?

 

[tsunami_all]

you have to be very specific about what you mean by S21, and S31.

Are you just trying to make an in-phase power splitter, with two 50 ohm output impedances?
Are you trying to make a true balun, where the "R" is a 100 ohm differential resistance?
Are you trying to get other than a 50 ohm single ended, or 100 ohm differential output impedance?

S21 and S31 are the insertion loss.

1. Phase difference requirement is 180 degree between 2 arms. Magnitude difference requirement is 0 dB between 2 arms.
2. No
3. 50 ohm on each number 1, 2 and 3 (refer to the schematic)

 

[FvM]

The schematic in post #1 surely doesn't achieve 180 degree phase difference. I presume your real schematic is different.

You are only reporting magnitudes, what's the phase?

 

[tsunami_all]

The schematic in post #1 surely doesn't achieve 180 degree phase difference. I presume your real schematic is different.

You are only reporting magnitudes, what's the phase?

Yes, that's not the real schematic. Phase difference is 180 degree in ADS simulation. I haven't tested it on the board and network analyzer yet.

 

[BigBoss]

You have also to check the Phase, not only Magnitude..
In fact, you schematic does not provide 180 Degree because of Capacitor and Resistance at input/output.They will imbalance the circuit.
If you do a well calibration in term of magnitude and phase both, you will see accurately the phase difference.
The best method is to use 4-Port VNA, you can tune it very precisely.
Before testing with VNA, Irecommend you to simulate your board in Momentum with all possible parasitic components.Using ideal Caps. and Inductors will not give you a consistent result.

 

[ktr]

Also, what's the frequency range of the balun? Narrowband and wideband systems and their compensation methods are different.

 

[tsunami_all]

Also, what's the frequency range of the balun? Narrowband and wideband systems and their compensation methods are different.

The frequency range is 2.4 GHz to 2.5 GHz.

 

[volker@muehlhaus]

50 ohm on each number 1, 2 and 3 (refer to the schematic)

Sure? Makes no sense if you post the wrong schematic (divider instead of balun). Is your actual circuit a balun with differential outputs 2 and 3?

 

[tsunami_all]

Sure? Makes no sense if you post the wrong schematic (divider instead of balun). Is your actual circuit a balun with differential outputs 2 and 3?

The schematic I posted here is just to show as a sample. Not exactly looks like that in actual circuit.
Yes, it has differential outputs at 2 and 3.

 

[volker@muehlhaus]

Ok, to avoid trivial mistakes: you designed it for 100 Ohm differential load, so that each port has 50 Ohm to ground?

I designed my lumped element baluns using this calculator:

LC Balun Designer

Radio Frequency Engineering Calculator.

leleivre.com

then used ADS co-simulation (EM + circuit) with actual component models (Modelithics) to verify and fine tune.

For measurement, you can test them back to back with two baluns.

 

[tsunami_all]

Ok, to avoid trivial mistakes: you designed it for 100 Ohm differential load, so that each port has 50 Ohm to ground?

I designed my lumped element baluns using this calculator:

LC Balun Designer

Radio Frequency Engineering Calculator.

leleivre.com

then used ADS co-simulation (EM + circuit) with actual component models (Modelithics) to verify and fine tune.

For measurement, you can test them back to back with two baluns.

Thanks for the suggestion. I will try that out.