tag reading with the NFC X-NUCLEO-NFC08A1 extension connected to an STM32L073RZ.

[med_benhaj1]

 

Hello,

Tag reading was working perfectly on my previous setup.

I now want to migrate my project from an STM32G0B1RE to an STM32L073RZ. Since the Arduino UNO R3 physical pinout standard is respected, the X-NUCLEO-NFC08A1 shield (which uses SPI, IRQ, LEDs, etc.) remains pin-compatible across different STM32 Nucleo boards (G0, L0, L4, etc.), as long as they follow the UNO R3 layout.

So, I created a new project in STM32CubeMX targeting the STM32L073RZ, and reused the configuration from my working G0-based project.

Current status:

Initially, execution would stop inside `demoCycle()`, initialization was successful (I could see `Initialization succeeded` in the UART), but no tags could be read.

After fixing the IRQ setup, execution no longer reaches `demoCycle()`, it now stops at the call to `MX_X_CUBE_NFC6_Init()`.

Any idea what could be preventing `rfalNfcInitialize()` from completing correctly on the STM32L073RZ?

Thanks in advance!

[Brian TIDAL]

Hi,

please check the pin assignment. See UM1724

SPI CS pin is likely not properly assigned.

Rgds

BT

P.-S. your trace are not a logic analyzer trace but a scope traces. Make sure to use a logic analyzer.

 

[med_benhaj1]

Hi,

Yes, you're right — the CS pin is on PB6, and that was indeed my original configuration before I changed it.

I've now restored PB6 as the CS pin, and it's still not working.

What I observe is that the blue TX LED on the X-NUCLEO-NFC08A1 is blinking, but I'm still not getting any tag detection.

Rgds

[Ulysses HERNIOSUS]

Hi, 

please try to be systematic:

1. Verify basic communication. Can you read register?
2. Can you write registers?
3. With SELFTEST enabled: Does rfalInitialize() succeed (not return an error)
4. Only now you can consider looking into tag reading

If you have issues in 1-4, then you need to look at the signals. What comes out as expected and what is missing?

I fear a two channel scope is not a good debugging tool for this. 4 channels are minimum (MISO,MOSI,SCLK,CS) but for further debugging also INT pin will help a lot. That's why we recommend a scope only for signal integrity. By the way your scope shots seem to only show a 50Hz from your main supplies at completely insufficient levels.

So our recommendation is a logic analyzer with 8 channels. There are cheap ones around. Please get one.

Regards, Ulysses 

[Brian TIDAL]

Hi,

this is almost the same configuration except for the drive strength of the GPIO (the "speed" parameter): GPIO_SPEED_FREQ_LOW in your project and GPIO_SPEED_FREQ_VERY_HIGH in my test. With GPIO_SPEED_FREQ_LOW, when reading the IC identity register 3Fh, the last received bit is 0 instead of 1. Thus the IC Identity check fails.

This SPI behavior is documented in the section 2.15.3 Corrupted last bit of the  ES0292 STM32L07xxx/L08xxx device errata. See the workaround section (e.g. use GPIO_SPEED_FREQ_VERY_HIGH).

This should help to fix your issue: this was not an easy one.

Rgds

BT