Skip to main content
Visitor II
December 9, 2018
Solved

Error 87 issue while reading tag using CR95HF-VMD5T NFC transceiver chip.

  • December 9, 2018
  • 10 replies
  • 3026 views

Hi,

We have developed a NFC reader design using the CR95HF-VMD5T chip. Our design is similar to the X-NUCLEO-NFC03A1 development board, its just we have used a flex NFC antenna by PulseLarsen with part number W7001.

The issue we are facing is that the board is unable to read a tag and returns error 87, with the firmware we have written.

We have tested our firmware on the X-NUCLEO-NFC03A1 development board and we are able to read a tag without any problem.

We also probed the antenna pads of both our board and X-NUCLEO-NFC03A1 using an oscilloscope just to see if we see any difference in the signal. What we noticed was that the X-NUCLEO-NFC03A1 board was showing a 10V peak-to-peak sine wave at 13.5Mhz whereas our board was showing around 3.4V peak-to-peak around the same frequency.

We wanted to know if this is the reason for our problem and a solution for that or if something else is going on that we're missing.

I have attached the antenna's datasheet for reference.

Thanks

Brosnan Gomes

Hardware Developer Engineer

+917972046516

Mrinq Technologies

    This topic has been closed for replies.
    Best answer by Brian TIDAL

    Hi,

    here is the answer of our expert:

    Antenna tuning circuit can be verified “in system�? using a network analyzer connected between TX1 and TX2. To do this, reader board has to be electrically floating (all signals lines and power lines shall be disconnected, board power switched off is not sufficient). In such conditions, CR95HF TX1 and TX2 are in open state and input impedance of antenna circuit can be measured without cutting connection between antenna circuit ant CR95HF. By living board floating, network analyzer GND/signal port will measure the differential input impedance of antenna circuit between TX1 and TX2 with no risk of electrical short cut between network analyzer GND and power line or signal reference lines.

    A power frequency sweep impedance measurement done with network analyzer allows to compare real life with impedance simulated with our tool. A good impedance value is between 27 and 50 ohms

    Let me know if you have further questions on this subject.

    Rgds

    BT

    10 replies

    Technical Moderator
    December 10, 2018

    Hi,

    I guess that the ErrorCode 0x87 is received when you sent SendRecv command (0x04) to the CR95HF.

    ErrorCode 0x87 means Frame Waiting Timeout i.e. a frame has been sent to the tag but no response is seen by the reader (typically no tag in the field).

    If the same software is working on the X-NUCLEO-NFC03, this looks like an antena/sensitivy issue.

    In order to support you in the best way, can you give some additional informations regarding your application:

    • is your software based on ST CR95HF driver or do you have your own driver directly using the communication commands?
    • which tag technology do you use: ISO15693 and/or ISO14443-A and or ISO14443-B and/or FeliCa?

    Meanwhile, I'll dig regarding the antenna.

    Regards

    BT

    BGomeAuthor
    Visitor II
    December 10, 2018

    Hi Brian,

    Thanks for your reply.

    Actually we got it to work, the matching circuit was the issue apparently. We had kept our matching network values as recommended in the W7001 flex antenna datasheet as we were using that part. But as I had mentioned earlier the peak-to-peak voltage was very low.

    We changed the values to those given in the X-NUCLEO-NFC03 reference board and saw the sine wave increase to 12V peak-to-peak and then it work and recognized the tag.

    Can help us with an explanation why it possibly had happened?

    Thanks

    Technical Moderator
    December 10, 2018

    Hi,

    the default values provided in the W7001 Datasheet are for a certain output impedance which does not correspond to the CR95HF output impedance.

    As the W7001 antenna impedance is very close to the X-NUCLEO-NFC03A1 one, using values from X-NUCLEO-NFC03A1 will give acceptable results.

    AN3394 (Antenna design and impedance matching guidelines for CR95HF multiprotocol contactless transceiver IC) provides helpful information and calculation tool regarding matching network. See https://www.st.com/content/st_com/en/products/nfc/st25-nfc-rfid-tags-readers/st25-nfc-rfid-readers/cr95hf.html to retrieve the various application notes related to the CR95HF

    Our expert gave the following feedback  regarding your schematics in addition to the matching network: C2 value to be increased a little bit (value to be adjusted by measuring the input impedance).

    Regards

    BT

    BGomeAuthor
    Visitor II
    December 12, 2018

    C2 is the one in series or parallel? Which schematic should I be referring?

    Technical Moderator
    December 18, 2018

    ​Hi,

    Nucleo antenna inductance is 1.1µH while W7001 antenna is 0.9µH. Hence, he would make sense to increase a bit C32+C33+C34 (called C2 by our expert) to take this antenna impedance difference into account if more fine tuning id needed.

    Regards

    BT

    BGomeAuthor
    Visitor II
    January 7, 2019

    Hi Brian,

    Thanks for your advice.

    I wanted to know what parameter should I be looking at to know that the match is fine, should I look at the peak voltage across the antenna or the return loss on a network analyser?

    I'm new to the 13.56Mhz matching circuit so wanted to know what will help me decide the if the match is good enough.

    Thanks

    Brosnan

    Technical Moderator
    January 7, 2019

    Hi,

    here is the answer of our expert:

    Antenna tuning circuit can be verified “in system�? using a network analyzer connected between TX1 and TX2. To do this, reader board has to be electrically floating (all signals lines and power lines shall be disconnected, board power switched off is not sufficient). In such conditions, CR95HF TX1 and TX2 are in open state and input impedance of antenna circuit can be measured without cutting connection between antenna circuit ant CR95HF. By living board floating, network analyzer GND/signal port will measure the differential input impedance of antenna circuit between TX1 and TX2 with no risk of electrical short cut between network analyzer GND and power line or signal reference lines.

    A power frequency sweep impedance measurement done with network analyzer allows to compare real life with impedance simulated with our tool. A good impedance value is between 27 and 50 ohms

    Let me know if you have further questions on this subject.

    Rgds

    BT

    Technical Moderator
    January 7, 2019

    Hi,

    in addition to my previous answer, I would recommend the reading of AN4327 (https://www.st.com/content/ccc/resource/technical/document/application_note/8e/99/77/a7/4f/58/4d/d6/DM00089926.pdf/files/DM00089926.pdf/jcr:content/translations/en.DM00089926.pdf) in particular §4. This gives comprehensive information on practical tuning.

    Rgds

    BT

    BGomeAuthor
    Visitor II
    January 9, 2019

    Thanks Brian,

    Can you also tell me about what kind of cable is recommended to use between the board and the antenna.

    Right now we were using simple two wire connection from the board to the antenna.

    Thanks

    Brosnan

    Technical Moderator
    January 9, 2019

    ​Hi Brosnan,

    if I understand correctly (please confirm), you plan to use a cable between the output of the matching circuit (called ANT1 and ANT2 in your schematics) and the flexible antenna?

    In that case, you will have to take into account the capacitance of the cable in the matching circuit.

    There are different possibilities for the cable depending on price, mechanical and EMC constraints:

    • unshielded or shielded twisted pair
    • coax

    At least an unshielded twisted pair should be preferred to a simple 2 wires.

    Rgds

    BT