ST25R3911B - antenna parameters measured X matching components

[JPehl.1]

Hi.

We have a custom board which includes NFC and it is based on ST25R95 transceiver and it has a double layer (top and bottom layer) antenna/coil of 10.5mm x 26mm, 5 turns at each side and 0.3mm width for traces, 1/2oz copper thickness. The reading distance for NFC tags was not sufficient for our project so we decided to switch the transceiver to a ST25R3911B and keep using the same antenna. Use the same antenna because we don't have more space to increase its size, and now the objective is to reach a higher distance reading for NFC tags.

We sent a bare board without components assembled to a university and they measured the antenna parameters of it with a ZNL 6 network analyzer, these parameters are required to calculate the matching componentes on "ST25R Antenna Matching Tool". v3.1.0.0 was used.

**********************************************************

NFC ANTENNA/COIL PARAMETER MEASUREMENTS FOR PCB-SL2-G-BLE

NETWORK ANALYZER: Rohde Schwarz ZNL 6

**********************************************************

Rsdc = 856 mR (@ 1MHz)

Lant = 1730 nH (@ 1MHz)

fres = 55.68 MHz

RP@fres = 25.1 KR

**********************************************************

CALCULATIONS ACCORDING TO ST AN4974 PAGES 17 AND 18

**********************************************************

Cant = 1 / ( (2*pi*55680000)^2 * (1730*10^-9) ) = 4.72 pF

K = sqrt(55.68/13.56) = 2.02637

RP@fwork = 2.02637 * 25.1 KR = 50.86 KR

RPdc = ((2*pi*13560000*(1730*10^-9))^2) / (856*10^-3) = 25.38 KR

RPant = (50.86*25.38)/(50.86+25.38) = 16.93 KR (16930 ohms)

Qmax = (16930) / (2*pi*13560000*(1730*10^-9)) = 114.86

**********************************************************

I have 3 options for matching circuit inductors values, all them from Murata series LQW18CNR

- 270nH x 160mR (max)

- 330nH x 250mR (max)

- 390nH x 280mR (max)

The best combination I have seen is with 330nH inductor:

 

As can be seen, my calculation of "Qmax" = 114.86 does match the "Q" value of 115 shown in the picture above.

A more complete report of the tests with the network analyzer are to be received on next week, do you want to check it?

I have 2 doubts.

[DOUBT 1] Do I need to change the parameters shown below or keep default values? Remembering that the aim is to reach a higher reading distance for the the tags. I have few knowledge about RF design so I would like to know if it is recommended to keep the default values for these fields, or to change them.

 

[DOUBT 2] As can be seen in the first picture, there are 4 capacitors which their values can be changed by the user, I mean the 10pF capacitors (yellow fields) and the 150pF capacitors (red fields), they are related to the pair of RF input pins RFI1 and RFI2. 10pF and 150pF are the default values shown by the tool, but I have seen on development board X-NUCLEO-NFC05A1 that this board uses the same 10pF capacitors, but uses a pair of 82pF instead of a pair of 150pF. Which value is best/recommended to use as a starting point? 82pF or 150pF? 

 

OBS: As our board is very space restricted, we will not use AAT, automatic antenna tunning, there is no enough space.

OBS: I know few about RF design, I need to know what is best/recommeded as a starting point for my next custom board.

OBS: We want to reach a reading distance of 5 to 7cm in open field.

Regards,

Jeferson.

 

 

 

 

 LATE EDIT:

I was able to move C22 closer to the transceiver

[Travis Palmer]

Hello JPehl,

I try to answer all questions related to your very detailed post. Please let me know, if i have overseen one.

The 330nH selection looks good for me. Anyhow you should not see too much difference when targeting the same matching impedance for all inductors.

The GUI calculates several parameters. For displaying them a rounding is done. For me your calculated antenna q of 114.86 fits quite well to displayed Q of 115.

The target matching impedance is influencing the current consumption of the reader. You can try to higher or lower it, depending on your first trails. (if you have enough read range, you can choose a higher matching impedance to have less power consumption).

Your target q-factor should bee chosen according to your application / tag you want to read. If you only want to read ISO15693 tags you can increase the Q-factor to ~45. If you want to read type A and B tags with 106kbps, you can increase the q-factor to ~35. If you also want to read HBR (bit rates up to 848kbps) then you should stick to a q of 17.

Critical capacitors for the ST25R3911B are C35 and C34 (VSP_RF decoupling caps). They should be as close as possible to VSP_RF (=positive supply) and VSN_RF (= negative supply).

Please let me know, if your read range expectations are fulfilled.

br Travis

[JPehl.1]

Hi Travis.

Buddy, I have tried my best in terms of layout of VSP-RF / VSN-RF pins, thats the best I can do I think, the capacitors are very close to the positive pin, being the 10nF closer than the 2.2uF one. You can check:

 

And the board have a ground plane, here is the ST25R3911B area:

 

Do you see any problem about the antenna parameters values that was measured? The parallel resistance is 25.1 KR @ 55.68MHz... That's not so high value?

 

We don't need high baud-rate transfer with the tags (HBR), that's not required or critical, so I'm thinking to change Q factor from 17 (original value) to 35 (double), then the best combination for matching that I have seen is:

47pF, 100pF and 1.5 ohms

 

You did not answered my orginal post regarding the [DOUBT 2], that is, about the 82pF/150pF capacitor that goes connected to RFIx pins... Please check original post again. Whats the best as starting point? 82 or 150pF?

Best Regards,

Jeferson.

[Travis Palmer]

​Hello Jeferson,

The above mentioned data looks good. Also the VSP_RF capacitor is connected via 3 vias to VSN_RF. This should be fine.

Regarding the Doubt 2:

The capacitive voltage divider is used to divide the antenna voltage ( rouchly between 20Vpp and 60Vpp) to 3V max (better 2.7V) on the RFI pin inputs.

Since different readers use different antennas and matching components (=> different Q-Factor), the antenna voltage will be different and therefore the capacitive voltage divider has to be different.

Your next steps should be:

Producing and assembly of the PCB

populating the matching network and do some fine tuning

measure the RFI input voltage (direct command "Measure Amplitude")

Adjust the capacitive voltage divider to have around 2.7Vpp on the RFI input using the measure Amplitude command.

BR Travis

[JPehl.1]

Hi Travis.

• What you mean with "direct command Measure Amplitude" ???, is the STR25R3911B able to measure the vpp voltage on RFI pins and then read the values via SPI?
• Can we measure the voltage with the scope? single ended relative to GND on each RFIx pins? Or differential between RFI1 and RFI2?
• I have to say for you that our next 5 prototype boards will come 100% assembled from China, so I have to choose an initial value for capacitor of voltage divider, and I will choose 82pF (exactly the same value of development board X-NUCLEO-NFC05A1), then I can adjust the divisor later in order to get a maximum of 2.7Vpp... is it really 2.7vpp? or it is recommended a maximum of 2.7V of positive peak (relative to GND) due the supply be 3.3V? How about negative voltages at RFI pins?
• Can the capacitors of the whole NFC RF circuit be all rated for 50V? Or REALLY need to be rated for 100V?

 

Regards,

Jeferson

[Travis Palmer · Accepted solution]

Hello Jeferson,

1.) Yes, the ST25R391x can measure the voltage on the antenna via a direct command. Please find details about the direct commands in chapter 1.2.13 of the datasheet DS11793 - Rev 6 and details about the measure amplitude command on page 32/116.

2.) Yes, the RFI amplitude can also be measured vs. GND using a scope. It is strongly recommended to use a GND spring, otherwise the NFC field will couple in and produce a wrong measurement. Typically you then see different voltage for RFI1 and RFI2. Therefore using the D3 command.

3.) I would recommend to use 10:150pF, especially if you have planned to increase the Q-Factor. 10:150 will cause the voltage on the RFI to be smaller. A too small voltage may cause a lack of sensitivity. You will still be able to read tags in close distance. A voltage above 3V on the other side could damage the device.

The RFI pin is biased by the receiver at 1.5V. If the AC signal is below 3Vpp then the voltage will not become negative.

4.) In you configuration it should be ok to have 50V rated capacitors. I recommend to check the voltages once you receive the board. Looking for example at C47 and C48, the voltage at the antenna can reach up to 100V until the rating of the caps is exceeded.

br Travis

[JPehl.1]

Hi Travis, thanks buddy.

Could you explain (shortly) a little bit more about the "GND spring" for the scope? What need to be done?

So, I have now rated ALL the capacitors for 100V and modified from 82pF to 150pF in order to decrease voltage at RFIx pins, as you said. A capacitive divisor is the inverse of a resistive one in terms of voltage relationship then...

Regards,

Jeferson.

[Travis Palmer]

Hi Jeferson,

please have a look at Keysight document N2840-97000.pdf page 9.

For legal reasons i cannot copy the content here.

Using 100V rated caps for the matching network you are on the absolute safe side.

br Travis

[JPehl.1]

Travis.

I don't have any question about this topic for now.

Please, I need an answer on the following topic, so we can start the production of our prototype boards:

https://community.st.com/s/question/0D73W000000VSvv/detail?s1oid=00Db0000000YtG6&s1nid=0DB0X000000DYbd&emkind=chatterCommentNotification&s1uid=0053W000000xydl&emtm=1606818212578&fromEmail=1&s1ext=0