Hello,
I would like to confirm the correct procedure for measuring antenna impedance using a NanoVNA. The chip under test is the ST25R3916.
I came across the post titled "How to Measure Antenna Parameters According to AN4974 with NanoVNA," which was very helpful in understanding the general procedure. However, I would like to request a few additional details to clarify certain aspects of the measurement:
Following is my differential antenna circuit.
Since the NanoVNA supports only single-ended measurements, I am considering the following approach:
I plan to remove the voltage divider blocks (highlighted in red) and disconnect the ground at the center point marked with “?”. This way, the VNA will only see the differential network, without any path to ground.
Could you please confirm if this approach for impedance measurement is correct?
If not, I would appreciate your guidance on how to properly measure the impedance of this differential circuit using a NanoVNA.
Regards
Suraj
1 ACCEPTED SOLUTION
Accepted Solutions
Hello Suraj,
As visible in the picture above, the grounds of the miniVNA and NFC05A1 are not connected.
In this way there will not be any ground loops or something similar.
Care must be taken if the ground of the VNA is connected to the GND of the NFC05A1 (e.g. both connected via USB).
In most cases the ground path is quite high-ohmic in this frequency band (e.g. 10MHz to 30Mhz). and you can proceed measuring single ended.
A good approach can be to measure the board un-powered and afterwords powered in the same condition.
I propose to take a eval board and measure the delta on your side.
Care must be taken, that the driver is in high-Z state during the powered measurement. It could harm your VNA.
The antenna design application note provides guidance for each device, how to execute VNA measurement when the reader is powered.
Below the measurement of the un-powered reader. In this case it is a ST25R3911B-DISCO with a ST25R3914 mounted.
This is the corresponding smith chart in un-powered state:
For comparison the same measurement in powered state.
To not harm the device, the register 0x27 is set to 0xff (dres set to high-z)
To have the same matching condition, the AAT has been set to trim0: 0x21 set to 0x80.
Due to powering the device some parasitic capacitance is changing. Resulting the matching impedance to change from: 18,9 +j3,6Ohm to 19,5 + 3,9Ohm.
But we can also see that in a lower frequency range (1-2MHz) an additional resonance is being introduced.
This resonance may be a consequence of the ground loop. But our main area of interest 10MHz to 30MHz is not affected.
BR Travis
Hello Suraj,
Below you can find a picture how we typically measure the matching impedance. The antenna can be measured similarly, just connected to a different connector and disconnecting the Antenna from matching network by opening the series damping resistor.
When measuring the matching or antenna parameters, the reader should be in its final configuration. E.g. in the housing. The housing and other things close by (other PCBs,...) can change the characteristic of the antenna and there change the matching and performance when not being accounted in the design process.
BR Travis
Hello Suraj,
As visible in the picture above, the grounds of the miniVNA and NFC05A1 are not connected.
In this way there will not be any ground loops or something similar.
Care must be taken if the ground of the VNA is connected to the GND of the NFC05A1 (e.g. both connected via USB).
In most cases the ground path is quite high-ohmic in this frequency band (e.g. 10MHz to 30Mhz). and you can proceed measuring single ended.
A good approach can be to measure the board un-powered and afterwords powered in the same condition.
I propose to take a eval board and measure the delta on your side.
Care must be taken, that the driver is in high-Z state during the powered measurement. It could harm your VNA.
The antenna design application note provides guidance for each device, how to execute VNA measurement when the reader is powered.
Below the measurement of the un-powered reader. In this case it is a ST25R3911B-DISCO with a ST25R3914 mounted.
This is the corresponding smith chart in un-powered state:
For comparison the same measurement in powered state.
To not harm the device, the register 0x27 is set to 0xff (dres set to high-z)
To have the same matching condition, the AAT has been set to trim0: 0x21 set to 0x80.
Due to powering the device some parasitic capacitance is changing. Resulting the matching impedance to change from: 18,9 +j3,6Ohm to 19,5 + 3,9Ohm.
But we can also see that in a lower frequency range (1-2MHz) an additional resonance is being introduced.
This resonance may be a consequence of the ground loop. But our main area of interest 10MHz to 30MHz is not affected.
BR Travis
