CC1352R: VC-TCXO problem

Part Number: CC1352R
Other Parts Discussed in Thread: SYSCONFIG

Hi,

We have a design with a cc1352R where we are using this VC-TCXO: NT1612AA. The schematic we follow is the next one:

The Q4 is not mounted, so we use only the Q2.

The GPIO used to power it is the 36.

When the board was received we made a fast test, just powering the TCXO without configuring it as a TCXO, so just before using the RF, power the port and once the RF is closed, unpower it. We saw everything was working fine, so we did not take more care of it.

Now that we are fully developing it, we have entered to configure it in the right way, so we have configure in the sysconfig the next parameters:

  • GPIO36: Output low
  • Enable TCXO Function: Power_enableTCXO

            void Power_enableTCXO(bool turnOn)

            {

                   if ( turnOn ) {

                  // Set corresponding DIO high to enable the TCXO

                   GPIO_write(PWR_CLOCK, 1);

                   }

                  else {

                  // Set the corresponding DIO low to disable the TCXO

                 GPIO_write(PWR_CLOCK, 0);

                }

       }

  • HF Clock Source: External 48Mhz TCXO
  • TCXO Type: Clipped Sine
  • TCXO Max Startup time: 0xFF

With this configuration, the radio is working fine, and the board is not consuming. But, we have an SPI RFID device which stops working after starting the RF. We are not able to receive the interrupt of this chip, but we are able to receive interrupts from other ports. If we go back to the way were we are powering manually without setting in the sysconfig that we are using a TCXO, the SPI RFID device works properly.

Someone can give us a clue of what can be wrong? And which is the relation between this 48 MHz and the SPI channels or the 32kHz xtal?

Thanks in advance.

  • Hi,

    Is it when the RF (static Tx or Rx) is enabled or just when the TCXO is enabled that disturbs the SPI RFID device ?

    Regards,

       Richard

  • Hi,

    At the reboot the RFID works, I make a RF communication (switch on TCXO, send data and switch off the TCXO) and then the RFID is not working anymore, but if I make the same procedure without configuring it as TCXO everything RF and RFID works.

    Thanks!

    Sandra

  • Hi,

    Please double check the GPIO.
    On the schematic it looks like the physical pin 36 is used, and not GPIO36.

  • Hi,

    You are right i wrote it wrong in the question, it is the pin36 the GPIO23.

    Thanks,

  • What is the part number for the SPI RFID device ?

    Is the chip also powered with 3V3 ?

    Have you tried to add a resistance at R8 so the TCXO signal will be attenuated ?

  • Hi,

    We are using ST253911B chip and it is also powered with 3,3V.

    We have not tried to put a resistance, we will try it, which value would you recomend? Thanks!

  • HI,

    The configuration of the CC1352R TCXO has a strong effect on the ST253911B chip. Since changing the settings of the clocking of the CC1352R TCXO affects the ST chip then this could be related to the signal strength level of the TCXO or cross-talk through the layout.

    Test the following to find the root cause:

    1. Use a 10 k resistor at R8 and R9; the TCXO clipped sine output should not exceed 1.7 V pk-pk

    2. Test with two boards, One board with the CC1352R and ST253911B and the other board with the TCXO. Clock the CC1352R from the separate board to determine if there is a RF layout issue.

  • Hi,

    On Monday I will be in the laboratory and we will test this, thanks!

  • Hi,

    Today I was able to make some measurements. Using the default hardware configurations, I have measure the peak to peak voltage in the output of the TCXO:

    - Configuring in the sysconfig, as external TCXO clipped sine type, the peak to peak voltage  was 550mV, and I was able to see the sine signal output

    - Configuring in the sysconfig, as 48 MHz XOSC_HF, the  peak to peak voltage was 530mV, and I was able to see the sine signal output

    Then we took the board to put the resistors of 10k, and we saw that the peak to peak voltage  was 1,3 Volts in the output of the TCXO, but we destroy the pad that goes to the uC and it was not working anymore.

    For the moment we can not do more tests until Thursday that we will receive new boards, but one thing that I do not understand is, how the cc1352R can work with the same TCXO but with different configuration? What is changing in the micro when I configure it with the TCXO or with the XOSC?

    Many thanks,

  • Out of curiosity: What happens if you use Q4 instead of Q2?

    The TCXO syscfg setting is mainly to setup when to switch from the internal RCOSC to the TCXO (to ensure a stable clock signal)  and the required callback. I don't have the full overview but most of the xtal module is unchanged regardless of setting. I think an internal buffer amplifier is turned off when using TCXO but the rest is the same so I'm not surprised it works. Note that even it works not setting the correct settings in syscfg to use TCXO  you have a risk of damage/ bricking the device if you use a TCXO but have not set it in syscfg. 

    To see if this is a power pulling issue you can disconnect the 3V3 going to the TCXO and connect this to a power supply. 

  • Hi,

    Yesterday I was able to make the test of using 2 different boards, one only for the TCXO and the other one for the CC1352R and ST253911B, but I have the same problem.

    Thanks!

  • Hi,

    We do not have the q4 mounted, so we can not test what happens,

    Yesterday we test to power the TCXO with a power supply, but still the same behavior.

    Thanks!!

  • Thanks RGW and TER for your ideas, yesterday we were making all the tests with the new boards without success. But this at least was helping us to put a side the TCXO, and think a bit more what is happening in the firmware.

    We have put some wires in the SPI of the ST253911B, and we have seen that we are able to make the initialization of the chip, later we send some messages by RF, and later on, when we try to access to the ST253911B, in the SPI line we can see all 0x00 are sent.

    In order to recover the chip, we are closing the SPI handle and opening it again after any RF message is being sent. Any clue of why can this happen?? Which can be the internal realtion between the TCXO and the SPI channel??

    Thanks!

  • Questions:

    -

    Yesterday we test to power the TCXO with a power supply, but still the same behavior.
    When the board was received we made a fast test, just powering the TCXO without configuring it as a TCXO, so just before using the RF, power the port and once the RF is closed, unpower it. We saw everything was working fine, so we did not take more care of it.

    It looks like you have tested the board in a very similar way but have got different result?

    - I know that Q4 is not mounted but are you able to do so?

    - Do you have a second noard or a different sensor that you can communicate with over SPI? The reason I ask is if the RFID makes an issue or if this is a general SPI issue.