Check if PN5190 is alive? SPI cmd doesn't work

[Haddock]

I am trying to get a life sign from PN5190B1 via SPI interface. MISO gives nothing back. Using GET_VERSION me. I send data and clock signal. I can see that on the oscilloscope. I think I have tried EVERYTHING. PN5190 is like dead. I must have missed something in the manual.
MCU: SAMD20

The steps I do are:
VEN, low and then high
Transmits 0x7F, 0x27 with SPI_Write, 2 bytes
Waiting for SPI ready
Sends 0xFF,0x00...0x00, SPI_Write, 6 bytes (the response)
Waiting for SPI ready
Examining the buffer. Only zeros.

I put SS low before each SPI command. Callback puts it high again.

Do I have to say that I have tried every possible combination? Shouldn't I be able to ask about the SW/HW version before anything else? Or is some initialization of the EEPROM/Register required before anything else?

All tips are gratefully received.

[Nor311]

Hi i would also like to attach the schematic of this project.

What we have found is that the internal voltages VDDC and VDDNV are zero. The voltages are cross measured against the developement kit where the voltages are present.

After a lot of hardware troubleshooting maybe you could spot the problem.. The crystals function is not verified but my assuption is that the internal voltages would be present anyway, or could that be the culprit?

Thanks in advance.

//Måns

 

[Julián_AragónM]

Hi @Nor311,

Based on the schematics and comparing it to the schematic from the evaluation board, no issues arise.

Some guidelines from the datasheet may be useful.

"Some pins are connected to blocking supply capacitors. PCB traces to these capacitors need to be as short as possible, and a low-ohmic grounding of the GND-side of the capacitors is required for optimized RF performance."

Could you please elaborate on "The crystals function is not verified"? And could you clarify the voltages issue... Are they zero on your custom board, but not zero on the evaluation board while testing? 

Best regards,
Julián.

[Julián_AragónM]

Hi @Haddock,

It is unusual to not get any signal from the SPI. Could you share some information about your project?

Does the PN5190 have a custom SW, or are you basing this on an existing example from the Reader Library?

Are you utilizing a custom board? Could this be a hardware design problem?

Could you also attach the oscilloscope or a logic analyzer of how the data is sent?

Best regards,
Julián

[Haddock]

The light blue is SS. Purple is the clock and yellow is the data (from MOSI).
I have never seen a puls in the dark blue, MISO.

[Julián_AragónM]

Hi @Haddock,

Are you using a library on your MCU to send the SPI commands, or are you sending the data raw?

The NFC Reader Library is recommended, as this is the official support for these readers. This community post showcases the porting of the reader library to a FRDM K64F board, which can be helpful in migrating to a non-NXP MCU.

There is also this recorded training.

Unfortunately, as this is an external MCU, it is out of this community support focus. I apologize for any inconveniences this may cause.

Best regards,
Julián

[Haddock]

Thanks for your reply or should I say questions? The 5190B1 circuit is supplied, I hope, with factory software?  I have not loaded it with any software. Is it missing software from the factory? I have seen that it has the option to update the FW, however. Haven't used that feature.
The 5190 is on a custom designed PCB with a MCU communicating with the 5190 via SPI.
As I said, the only thing I do is ask PN5190 for HW, SW version (0x27) as a test. Tried sending the command in different combos in case I misunderstood something. I never see anything happening at MISO. (VEN, MOSI, CLK and SS are checked and present). So I thinking if it is necissery to obtain some kind of parameters to power supply registers? We using power opt 1. Everything that needs power is connected to 3V3.

The way I see it, the fault is in one or more options 1) the voltage, 2) the crystal, 3) design or 4) bad soldering under the IC circuit.

The only thing this PCB should do is check that the RFID tags (testobjects) are activated and alive. 

I will return with osc. pictures.

I am grateful for any input regarding this because I am running out of possible solutions. Colleagues have reviewed the design in several rounds. I have tried different variations on protocol and also the SPI settings on the MCU. I have also searched for suitable PDFs that can give me suggestions on what a short init should look like. 

Thanks