Design questions around ST25R3911B

[Nicolas Goy]

Posted on June 19, 2018 at 19:10

I need to integrate an NFC reader for medical device pairing. I already designed a tag for the sensor which works with my phone.

The next step for me is to add an NFC reader to the portable monitor. The biggest antenna I can fit into the back of the monitor is about 15cmx10cm and as it is quite big, I picked the ST25R3911B to drive it because it has good power. I need a read range of about  20-30cm. Apparently, the best standard to achieve such range is ISO15693. My current tag uses ISO14443/B but I can easily swap the IC with something like the MLR24LR.

The sensor has a zigbee chip, and I only need the NFC to exchange the encryption key (about 128bytes with all metadata), so datarate is not important.

Some info:

- Tag is 45mmx35mm, using TI RF430CL330, but if ISO15693 is preferred, I can swap the IC and redesign the PCB.

- The max antenna I can fit inside my monitor (NFC reader) is 15cmx10cm.

- I need to pass 128bytes of data. (1kbit) Read only (only written by the sensor microcontroller using SPI)

- Read distance should be minimum 20cm, 30 or even 50cm preferred.

I have a few questions:

- Is ISO15693 the best choice for my application?

- I picked the ST25R3911B, is it good, or another chip?

- If MLR24LR the right choice for ISO15693 for the tag?

- I downloaded the antenna matching tool, I understand the antenna parameters inputs, but I wonder about the 'matching inputs', there are presets called 'VHBR, EVMCo and General purpose', but I am not sure I understand how those impacts the system.

- In AN4974 the antenna is attached to the ground at the middle, is this necessary for differential design (using both RFI0/1 and RFO0/1 connections, design of figure 3 in DS11793 page 14)?

#nfc

[Travis Palmer · Accepted solution]

Posted on June 26, 2018 at 09:05

Dear Nicolas,

I think ISO15693 would be the right choice for your read range requirements.In order to read the ISO15693 tags you will need a powerful and sensitive reader - which means the ST25R3911B fits perfectly.

I did a little proof of concept. I connected a 14x10cm 2 turn wire wound antenna to a ST25R3911B-DISCO board and measured the read range with several tags. The antenna inductance is around 1.7µH. The antenna has no middle GND connection.

Thisare the inputs i did in the ST25R Antenna Matching Tool:

Besides the measured antenna parameters i have chosen my EMC filter settings and target matching impedance and target antenna Q.

The presets define matching inputs for a special purpose. VHBR has data rates up to 6.5mbit and therefore requires very low Q-Factor. EMVCo only requires 106kbps therefore a Q-Factor around 25 is suite able. ISO15693 even has lower data rates. Therefore we can further increase the Q-Factor.

The matching impedance defines the power transmitted towards the antenna.

Since we want quite some output power i have chosen 10 Ohm matching impedance. Since ISO15693 is quite relaxed on rise and fall times i have chosen a target reader Q-Factor of

This was then my reader setup:

This are the followingresults i got:

Tag type read range [cm]

LRi2K (ID1 8.5 x 5.4 cm - picture) 24

ST black coin (3cm diameter - picture) 21

ST25DV_Discovery_ANT_C5 (4.0x2.2 cm) 20

I did some tweaking on the receiver settings (AM only, 1st level gain set to boost). also due to the high output power it was necessary to bypass the internal Vsp_RF regulator (connect pin 9 to VDD). V_RFI was adjusted to 2.7Vpp. For the EMC filter i have chosen low DCR, high Q coils.

Also you have to be aware that you will need a verygood PCB design and good decoupling to other noise sources (like DC/DC, ...)

With further improvements (adjusting the Q-Factor and matching impedance) you should be able to reach at least 25cm read range with the ST25DV ANT C5. Placing a pick up coil close to the tag will tell you if your reader is sensitivity or power limited and tell you in which direction you have to optimize.

BR Travis

[Nicolas Goy]

Posted on June 26, 2018 at 13:44

Woaw, this was an insanely helpful answer, I did not expect someone to create a test setup. Thanks a lot.

I have two questions in addition, first, I didn't quite understood what you meant by 'V_RFI was adjusted to 2.7Vpp' adjusting the peek to peek RF voltage is done in hardware or software? Second, for PCB design, is there anything else you recommend beside what is in AN5043 (page 15) for power filtering, especially with VSP_RF bypassed?

Thanks

[Travis Palmer]

Posted on June 26, 2018 at 15:31

Dear Nicolas,

The V_RFI is the antenna voltage applied to receiver. This voltage is divided down by C401:C402 and C411:C412 (UM2042, page 13). The V_RFI should not exceed 3Vpp. Due to component and PCB tolerances 2.7Vpp is recommended. Depending on the Q-Factor the antenna voltage is typically between 20Vpp and 70Vpp - higher Q gives a higher antenna voltage.

Besides the AN5043 there is no further guideline for PCB layout. Filter component manufacturer have special web tools (for example SimSurfing) which can help you to design special power supply filter. You can foresee T and PI filter footprints in the VDD supply of your PCB layout. I would still foresee a 0R jumper (like the J303) between VDD and VSP_RF.

BR Travis

[Nicolas Goy]

Posted on June 26, 2018 at 15:50

Ok I get it, with the capacitor voltage divider you placed, 10 and 150pF 3Vpp would be an antenna voltage around 48V, right? I guess I have a last question, how do you measure the antenna voltage? I get how you calculate the Q factor with the results from the VNA, but not the voltage.

[Travis Palmer]

Posted on June 27, 2018 at 08:28

Hi Nicolas,

Either you measure the voltage on the antenna using an oscilloscope - be aware that there will be a ground loop done by the scope probe GND connection which will pick up the magnetic field and give a wrong result. Using a 'scope probe ground spring' can greatly improve your measurement result.

The second approach is to do it by try and error.

I ended up with something like 10pF : (220pF + 33pF). Which means something like ~ 68Vpp on the antenna.

This voltage may be too big for using AAT (25Vpp trim pin limit). Therefore you either place a capacitive voltage divider instead of each trim capacitor or you disconnect AAT. 

BR Travis

PS: The VNA should only be used when the board is not powered. Otherwise it could be damaged.

[Nicolas Goy]

Posted on June 27, 2018 at 23:14

I should have precised, under what condition to do the voltage measure. Should just disconnect TRIM and RF_IN, turn the board on, and measure the resulting voltage?

I'll try to make a prototype without AAT.

Thanks.

[Travis Palmer]

Posted on June 28, 2018 at 09:16

Hi Nicolas,

The reader has to be in normal operation condition. 

If you want to use AAT, then AAT has to be connected.

RF_IN is a high ohmic input and does not have to be disconnected.

The reader can be in polling mode or emitting a constant field.

BR Travis

[Nicolas Goy]

Posted on June 28, 2018 at 15:40

I'm not going to use AAT as, as you said, the voltage might be higher than the 25Vpp accepted by the AAT pins.

Thanks for the support.

[Nicolas Goy]

Posted on July 15, 2018 at 22:19

Hello again.

I was able to build a working board using a ST25R3911B, but I have few more questions.

For reference, my antenna parameters are:

Inductance 1340nH

DC resistance 800mOhm

Parallel resistance 8230Ohm

Self resonance 37.42Mhz

I used the following matching components:

EMC inductor 217nH/700mOhm

11750kHz EMC filter

10Ohm target matching Z

40 Target antenna Q

As I know I would get high antenna voltage, I did not connect AAT on my test board.

The antenna calculator gave a 10/150pf voltage divider for RFI which I followed for the test board.

After assembly, the board works, but I have 90Vpp antenna voltage, which leads to 5.6V on RFI pins. I was surprised to see it was working while being above the 3V range.

If my calculations are correct, I need to replace the 150pf by a 350pf capacitor to bring down the RFI voltage to 2.5V.

Also I want to use the AAT now that I know my antenna voltage. What target voltage (capacitor divider ratio) do you recommend on AAT TRIM pins? Also, my main question, does the voltage on all TRIM pins need to be the same?

Thanks a lot for your support.