Can I match single line for matching differetial lines?

[tony_lth]

I meet a match problem.
A receiver has a diff line input, and it has a S1P (one port S parameters) file which based on 100 ohms.
And between the antenna and receiver is a un-balanced to balanced filter.
The impedance of the antenna and the unbalanced port of the filter is 50 ohms.
The impedance of the balanced port is 150 ohms.
I need to design the match circuits between the balanced port and the diff line input of the receiver.

My question is:
1. The impedance of the matching circuits should be normalized to 100 ohms, right?
2. Should I match to minimal NF, because it's a receiving system?
Although I can match with S-parameters, it is a minimal VSWR matching, not a minimal NF matching.
3. And how can I test with standard 50 ohms VNA systems? Such as VNA input and output cable, they are both 50 ohm. So I only can test at the un-balanced port of filter, right?

Thanks.

 

[E Kafeman]

Hi Tony,

How to do the matching depends a bit on what measurement tools you have, frequency and bandwidth.
Say that the radio is something with bandwidth and frequency range type Bluetooth, not to wide and not to extreme high frequency, and you have no differential probe for VNA.

A halfwave balun for the VNA is simple to make for that frequency range. Calibrate it together with VNA by soldering a smd resistor at tips of the balun (200 Ohm).
Now can you measure balanced impedance in a limited frequency range.
If you know or have possibilities to measure to find impedance for minimal noise is that the best, if not, measure receiver input impedance with VNA.
As I know that you have AnTune software, I recommend that you use that, directly connected to VNA. Microwave Office or similar is else a possible alternative but VNA still save curves as it was calibrated for 50 Ohm so that must be corrected in software.

In AnTune, set system impedance at 200 Ohm and save measured RX impedance as impedance reference.
Next measure impedance at almost the same spot at PCB, but in the other direction, against the balun with antenna connected.
AnTune will calculate a unbalanced matching network but it is rather easy to recalculate as a balanced network.
Implement that impedance network in one direction, and verify by measuring that AnTune shows expected low VSWR.
Antune shows total resulting VSWR and at VNA Smith chart should impedance curves for respective direction be horizontal mirrors of each other.

A alternative to measure RX impedance is if you have a S11 file for optimal impedance measured at 100 Ohm, then recalculate it for a 200 Ohm system and import the file and set it as reference impedance.

My 2450 MHz balun:

 

[BigBoss]

There won't be any matching between antenna and balun filter.50 Ohm to 50 Ohm..
There should be a matching circuit between receiver circuit and your balun filter.You should consider 75Ohm (150 Ohm diff.) to 50 Ohm (100 Ohm diff.) matching because half of circuit then you have to combine this circuit as diff.-to-diff.
For optimum matching, you must know Noise Parameters as NFmin, Rn and Gammaopt otherwise conjugate matching will be proper.If you know those parameters, in this case you repeat the procedure which is for balun filter to receiver for minimum noise figure.

 

[tony_lth]

Hi, Kafeman and BigBoss,
Thanks for your reply.
Yes, I have AnTune and found it's easy to do Antenna match with AnTune. But for diff match, I need more exercises. I will let you know if I meet some problems.