2024-07-18 11:11 PM
We developed custom PCB with CR95HF on it.
PCB on itself was tested and RF circuit tuned(PCB and antenna design is based on development kit), antenna frequency is at 13.56MHz(+/- 0.5KHz), at around -60dB, uart communication working, can put CR95HF to sleep, wakeup, set register and protocols, but the problem is that we can not detect NFC tag (we use 14443A).
Also we can't successfully finish calibrate process, described in datasheet, since the tag is never detected. We don't have a lot of experience with RF system, so looking for any kind of help.
Commands we use:
-protocol setup 0x0202020000
-tag respond request 0x04022607
Kind regards,
Klemen
Solved! Go to Solution.
2024-07-29 02:05 AM
Hi,
0x8700 means no response from the tag. This means the tag is likely not able to understand the command sent by reader.
If the same application works with RFID Click with the same tag, this means the suspect is the matching circuit.
Rgds
BT
2024-07-19 12:59 AM
Hi Klemen,
what are the return codes of the protocolSelect command and of the sendRecv command?
Can you connect a logic analyzer on SSI_0, SSI_1, UART_RX / IRQ_IN, UART_TX / IRQ_OUT and SPI_SS and provide a trace from the power-up up to the protocolSelect and sendRecv commands?
With an oscilloscope, can you probe the RF field after the protocolSelect command (just connect the gnd of the probe to the tip of the probe and put the probe in the field and you should be able to capture the field)?
"PCB and antenna design is based on development kit": can you share details: which development kit: X-NUCLEO-NFC03A1 or other kit?
Which tag do you use (manufacturer/model)? Can you check the tag is alive (e.g. use the NFC reader from a smartphone with STNFCTap app)?
Make sure to follow AN5248 ST25R95 transceiver antenna tuning circuit with EMI filter
Also, I would recommend to base your FW development on top of ST RF abstraction layer (RFAL). See STSW-ST25R017 ST25 embedded NFC library.
Rgds
BT
2024-07-28 11:50 PM - edited 2024-07-28 11:51 PM
Hi,
Thank you for information and sorry for late response, due to vacation.
- I dont have option to use logic analyzer, only logic probe for my oscilloscope (SDS-8034X).
- In attachment are screenshots from spectrum analyzer(SSA-3021X), of field before protocol select, only used echo command (0x55) to check if its up and after selecting protocol (0x0202020000).
- It is based on evaluation kit RFID Click from MikroElektronika, since there are no official STM boards available (all are marked as obsolete or out of production and there is no old stock). Also the serial control software was developed and tested with that board and on there the tag detection works. This is what pointed us toward the "detection calibration" issue.
- Just fyi, on our custom board, using sleep and protocol select commands, we can turn field on or off, so that part of firmware is working.
- Will check this abstraction layer to review our firmware.
@Brian TIDAL wrote:Also, I would recommend to base your FW development on top of ST RF abstraction layer (RFAL). See STSW-ST25R017 ST25 embedded NFC library.
Kind regards, Klemen
2024-07-29 12:15 AM
Hi,
"since there are no official STM boards available", X-NUCLEO-NFC03A1 (expansion board based on CR95HF/ST25R95) is not obsolete and is still available from distributors.
What is the return code of the sendRcvd command when you try to poll the tag on your custom board?
I believe the issue is due to your matching network. Make sure to follow AN5248 ST25R95 transceiver antenna tuning circuit with EMI filter.
Rgds
BT
2024-07-29 01:15 AM
Hi,
- Oh, guess we missed this kit in the research.
- Return code is 0x8700.
- Will go through tuning circuit and calculation one more time to check it.
2024-07-29 02:05 AM
Hi,
0x8700 means no response from the tag. This means the tag is likely not able to understand the command sent by reader.
If the same application works with RFID Click with the same tag, this means the suspect is the matching circuit.
Rgds
BT
2024-07-31 06:04 AM
Thanks for your help, went through the AN5248 again and found error in EMI filter calculations which moved the cutoff frequency to around 13.6MHz, which was to close to 13.56MHz. Removed the filter capacitors and inductors and now works. Will recalculate the correct values and apply them.