ST25R3916(B) in Card Emulation Mode: MCU_CLK, EXT_LM, advanced waveform shaping

[JWong.10]

Hello,

Does sourcing the MCU clock from the ST25R3916(B) improve the RX sensitivity in card emulation mode?

Is the strongest external load modulation obtained by setting the drain resistor (R301 in the X-NUCLEO-06A1/08A1) to zero?

AN5768 states that the ST25R3916B has "advanced shaping of NFC-A, B, V, F modulation waveforms". Does this improve the card emulation mode as well?

[Travis Palmer]

Hello JWong,

MCU_CLK is independent to the card mode feature of the ST25R3916(B).

It provides different clocks (13.56, 6.78 and 3.39 MHz as well as low frequency clock) via the MCU_CLK pin.

The ST25R3916(B) provides two mechanisms to generate LMA. One is using the internal driver resistance. Here the chip automatically switches between pt_res (unmodulated state resistance) and ptm_res (modulated state resistance). The LMA is given by the delta between pt_res and ptm_res.

It can also drive the EXT_LM pin to generate a load modulation signal via external transistor and resistor.

Please have a look into DS13541 Rev 7 (or later) chapter 4.2.1 section "Passive load modulation".

AN5768 is only valid for reader operation. The AWS is not active in ce mode.

You should also have a look at AN5896. If the tuning in ce mode is too close to the resonance frequency of the initiator, it can have bad influence on the wave shapes.

Also the loading effect or DVR tests may fail.

You should also check if the ST25R3916(B) is able to detect the NFC field of the initiator. If not, adjusting the field detector thresholds might be required.

br Travis.

[JWong.10]

Hi Travis,

Thanks for pointing out AN5896. We used the ST25 antenna matching tool for RF tuning including the VCCs, and matched our resonance to 13.56 MHz in reader mode with RFO set to high-Z and AM turned off. Is this the right procedure for card mode, as we were under the impression that the same matching circuit for reader mode could be used for card mode? Observing the match on the VNA, in card mode our impedance locus still has the resonance at 13.56 MHz in the high-Z RFO state. I'm wondering because AN5896 says "For instance, the ST25R3916/16B uses voltage-controlled capacitors to switch the tuning between card and reader mode", and my resonance for both reader and card mode is at 13.56 MHz.

Or should we manually set the resonance frequency to the center frequency of the load modulation (e.g. for NFC-F at 424 kbps we would tune to (13.56 + 0.424) MHz)?

Another question: can AAT_A and AAT_B be controlled in card emulation mode? We wanted to run the auto-tuning algorithm to eliminate production variation in the tuning. However, it seems the DACs are always set to the midpoints in card mode, and we can only control them in reader mode.

[Travis Palmer · Accepted solution]

Hello JWong,

Since you have varicaps populated we simply need to change the setting between reader and CE mode operation.

Therefore I propose to do the measurement as described in chapter 7.3 of AN5276 (Rev 5) while the device is in card mode.

Item 7 - Place a 2 Ω resistor between the RFO pin to simulate the chip resistance during operation should be skipped! no resistance should be placed between RFO1 and 2.

Then you should change the AAT settings to all 4 extreme:

26h = 27h= 0x00

26h = 27h= 0xFF

26h = 0xFF, 27h= 0x00

26h = 0x00, 27h= 0xFF

You should choose the setting where you have the highest difference to 13.56MHz.

On our discovery board we have ~1.5MHz delta to 13.56MHz.

Being different from 13.56MHz is beneficial for the wave shapes. Having distorted wave shapes can either make decoding errors on the initiator side or on the CE device side.

Since the ST25R3916(B) does not require to be powered from the field, there is no need to go closer to 13.56MHz. You should also make sure, that pt_res and ptm_res have a proper difference.

e.g. pt_res set to 0xFF, ptm_res set to 0x00. It should give the highest LMA.

Please also let me know, if you are targeting any compliance to a standard (e.g. NFC-Forum or ISO).

From your previous post i assume you are having some read range trouble. Can you maybe share some more details about the problems you are facing and what you want to achieve?

BR Travis

[JWong.10]

Hi Travis,

We don't have the setup for ISO10373-6 and we're not targeting standards compliance. However, I'm able to adjust the AAT registers and observe the matching and resonance change on the VNA.

I was able to get the AAT working in CE mode (mode_entry_3_reg in analogConfigTable), and I do see a difference in read range at one end of the AAT_A,B extremes and setting register 0x0029=0xf0 on internal load modulation.

Yes, we are having read range trouble. Small coil, and it's not positioned optimally (not concentric with the reader coil where the read range is much better), and we'd like to avoid changing the mechanics.

We're only using CE mode, not reader mode, so I also changed the RFI divider caps from their original values of 10 pF top and 220 pF bottom to have 100 pF bottom. I realize that in reader mode this is dangerous as the RFI pk-pk exceeds 3V, but this bumped the read range up substantially. Still, we prefer not to do this as it eliminates reader mode from future use.

Do you have any other suggestions given that we aren't targeting compliance?

[Travis Palmer]

Hello JWong,

In principle there are the following possibilities why communication does not get established:

• 3916B does not see the phones field/modulation
• 3916B does not generate sufficient load modulation to trigger the phone’s receiver.

From software perspective you can try to play with:

1. Try to increase passive target modulation strength – existing entries in AC: ST25R3916_REG_PT_MOD: ptm_res and pt_res
2. Try to lower Field detector threshold – existing entries in AC: ST25R3916_REG_FIELD_THRESHOLD.

you would need to check which one is the limiting factor and try to improve it by register settings.

br Travis

[JWong.10]

Hi Travis,

For us it seems that not seeing the reader field/modulation is the case. Our antenna is pretty small, and if I adjust the RFI AC voltage divider as above the read range boosts noticeably. So, the passive target modulation is not the limitation for us.

In the end, we are moving the NFC coil of the card emulator to align with the reader to address the issue.

Thanks for your support!