This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

CC2640R2F: No output from DCDC_SW

Part Number: CC2640R2F
Other Parts Discussed in Thread: TLV840

Hi Team,

We have a sensor built with CC2640R2F which is already in mass production for some periods.

Recently, we got three failure units from field. The failure phenomenon was kept so we can do some basic measurement and we found there is no output on the DCDC_SW.(and no VDDR)

The phenomenon is gone when we power off it and power on it again.

From the record, these three units are working good originally and suddenly stop working(they were kept in this failure state and sent to us).

We know the customer will put a magnet on the back of the sensor to fix it to the metal wall.

So we're wondering will the magnet cause this issue?

We have tried to put a very strong magnet close to the chip and the inductor, but we didn't see the same failure.

Any suggestion for debugging this issue?

  • Hi,

    You might be observing the same issue as advisory 09 of errata www.ti.com/.../CC2640R2F

    regards

  • Hi,

    no it's different from advisory 09.

    The advisory 09 draws around 2.25ma however this issue draws almost nothing.

    Would you please share other possible causes of DCDC_SW no output suddenly in running ?

  • Hi,

    Is this a battery based module? I wonder if this issue happens in some power mode? e.g standby.

    Can you verify that?

    regards,

  • Hi,

    1. It's a battery based device. (battery is full, no BOD concern)

    2. It's not in standby mode since all software components fail.

    (butter press triggered interrupt no response, timer based behavior no action .. complete stop)

    3. Would you please explain in more detail about your theory ?

    We want to know any possible cause to try to reproduce the issue fast.

    (it runs for weeks perfectly than suddenly no DCDC_SW output)

  • As Sam's reply, our sensor is a battery powered device. Do you have any feedback for Sam's reply? 

    What's the possible reason that the DCDC_SW has no output suddenly?

    (the SW setting is following the default value of SDK,

    #define SET_CCFG_MODE_CONF_DCDC_RECHARGE                0x0        // Use the DC/DC during recharge in powerdown

    #define SET_CCFG_MODE_CONF_DCDC_ACTIVE                  0x0        // Use the DC/DC during active mode

    )

  • HI,

    I am discussing this issue internally with colleagues. Will get back to you as soon as possible.

    Thanks,

  • Hi,

    Can you share the schematic of the design?

    Also, VDDR = 0 when device is in shutdown. Is your code putting the device to shutdown?

    Are any of the IOs exposed to negative voltages, perhaps due to switching operations?

    Thanks

  • Hi ,

    There is no shutdown mode in our design.

    And our device is a compact temperature sensor so no IO is exposed.

    This sensor was used for monitoring the temperature of refrigerator(normally, 0~20 degC) or freezer( normally -20~5 degC).

    Our customer will put a magnet on the back of the sensor and then apply it to the refrigerator or freezer.

    The schematic is attached. We're using a customized CC2640R2F module. 

    6708.schematic.pdf

  • Hi,

    I am reviewing the schematic.

    Also, wondering if the magnetic field might be having an effect on the 10uH inductor on VDDR line.

    Have you seen failures on boards without magnets?

    Regards,

  • Hi, 

    This failure is only observed on this customer's units which have magnets on the back.

    But we have tried some extreme test on the inductor. It seems the requirement of this inductor is quite loose.

    1. We have tried to use a very small inductor(nH) and it works.

    2. We have put several units with strong magnets just on top of the inductor. But they're still running without problem so far(over 10 days)

    3. As running, we've tried to short the inductor with tweezers. But it ran without problem.

    4. Onle when we short the inductor with 0 ohm before power on, it will be fail to run.

  • Hi,

    Regarding the point 4 above, the inductor is required to run in DC/DC converter mode and that value is based on extensive characterization.

    See section 11 of this app note https://www.ti.com/lit/swra640

    Also, if you do not want to use that inductor, consider using global LDO mode (section 11.3 from same app note) for this application. It seems like the magnetic filed might be causing the inductor to miss-behave on some boards.

    regards

  • Hi,

    There are more units failed and returned from the field.

    So far, 6 units have the same problem, there is no power on VDDR( and DCDC_SW) when it fails.

    And power cycling them will make them work again.

    There is one thing in common that they are all failed when the environment temperature rises.

    As mentioned earlier, the units are put in the refrigerator or freezer to monitoring the temperature changes.

    The failure happens when the temperature rises from -20 degC to 0 degC( freezer) or from 0 degC to 20 degC( refrigerator).

    The rising of temperature takes around half an hour.

    Any suggestion for this clue?

    Is it possible that the recharge is not enough due to the temperature rising, something similar to Errata 04?

    Although we're not using the sensor controller.

  • Any update?

  • Hi,

    How fast are you running the sensor controller?

    Depending on the frequency of wakeups, there may not be significant temperature change during the interval to cause this errata issue.

    Thanks

  • Hi,

    no we don't use sensor controller at all.

  • Hi,

    You can try varying the recharge settings from this app note (see section 4.2) :www.ti.com/.../swra486.pdf

  • Hi,

    1. Do you think recharge is possible cause of DCDC_SW no output ?

    2. I didn't see clear explaination on recharge setting, would you please share more detail parameters ?

    3. How to tune recharge parameter to reproduce the issue for aging verification ?

    4. Can global LDO mode avoid recharge issue ?

    We need to know possible cause, the situation is killing our project.

    Thanks,

    Sam

  • Hi,

    Global LDO mode will not avoid recharge as this process is used in standby modes.

    With GLDO, the DCDC inductor can be removed and VDDR is supplied internally by LDO.

    Also, the temperature gradient, in your case, is not as fast to cause the errata to play here (the RTOS adjusts the recharge duration automatically). 

    Please submit your design for review here www.ti.com/.../SIMPLELINK-2-4GHZ-DESIGN-REVIEWS

    Thanks

  • Hi 

    To get rid of this issue, we have an idea that needs your help to confirm.
    We want to add a reset chip to monitor the VDDR.
    The chip we plan to use is TLV840MADL08DBVR which consumes less than 1uA normally
    and has a 0.8V threshold with open drain RESET(LOW active) output.
    When VDDR falls below the threshold(0.8V), the chip will reset the CC2640R2F.
    Then when it further goes down to below the working voltage of TLV840, the RESET signal will be pulled up again and CC2640R2F will boot again.
    Is this idea workable?

  • Hi,

    This could work but we don't recommend connecting external circuitry to VDDR pin. This may cause error in recharge time calculation.

    If you want to use external regulator mode, you could use this method which basically will monitor the main supply and reset the device if VDDR drops below threshold. (see https://www.ti.com/tool/CC2650EMK for example).

    Also see if this thread helps as it looks like you may be seeing this issue https://e2e.ti.com/support/wireless-connectivity/bluetooth-group/bluetooth/f/bluetooth-forum/569563/cc2640-voltage-on-vddr-dcoupl-down-to-zero

    Regards

  • There is no external components that consume lots of power and it works for months and suddenly DCDC_SW stop output. We understand put a reset IC on VDDR might cause recharge time calculation error however anyway the recharge error occurs that make DCDC_SW stop output and it cause un-recovery hangup. We need a urgent workaround. If putting reset IC can we finetune recharge margin to make more conservative recharge period ? How to do that ?


    I saw there are several posts on the forum about no DCDC_SW and we think there is something weak on recharge algorithm.

  • Hi,

    Its worth trying but we cannot guarantee that it will fix this.

    Also, the recharge calculation is done automatically by RTOS which takes into account the temperature gradient and sleep interval.

    I am not sure if there is a way to tweak that unless you use non-rtos solution.

    Regards

  • Hi,

    Is PowerCC26XX_config.vddrRechargeMargin a parameter to make recharge more conservative ?

    BTW if we don't use reset IC to monitor VDDR, what alternate choice do we have to recover from DCDC_SW stop since TI didn't give us possible cause and any suggestion to avoid this ?

  • Part Number: CC2640R2F

    Hi,

    The customer has a question below since we have an issue in rarely chance to reproduce that CC2640R2F will go into shutdown mode (VDDR is gone) in low temperature(-20 degree) condition with full power battery(CR2032). We think it may due to the battery voltage is too low cause the BOD/POR is not working properly cause this in shutdown mode.

    Below is customer 's assumption and qiz:

    we would like to use TI  TLV840 to monitor VDDR, the reset pin is using signal chain to pull low. The full part number is ""

    Assume the VDDR is lower the threshold (may 0.8v) the reset IC send assert to pull low reset pin.

    Then the voltage is too low so this reset IC is going to inactive by external pull high, the system reboot successfully.

    Q1: Is this idea workable?

    Q2: Is there any side effect once we put this reset ic in VDDR? The power consumption of TLV840耗 is <1uA(typical 0.125uA)

  • Hi,

    Thank you for reaching out. I have a few follow up questions. It seems that the issue is that the battery voltage becomes too low for the device to operate and it shuts down, correct? If this is the case, then why is VDDR being monitored instead of VDDS? I believe the TLV840 will provide the functionality you are looking for, but if the input voltage is too low, then the device will not boot up.

    Best Regards,

    Jan

  • Hi,

    I have notified an expert to comment further.

    Best Regards,

    Jan

  • Hi,

    I was looking at the schematic you attached and it seems like the 10uH inductor is placed off board. We do not recommend this as stated in section 11.2 of this https://www.ti.com/lit/swra640

    This combination has to be placed as close as possible to the chip for short current paths.

    In this case I suggest you use GLDO mode (section 11.3) and see if this fixes the proble.

    Regards

  • Hi,

    When the issue happens (suspected something wrong in recharge), is it possible in shutdown mode and could be waked up by wake-up pin ?

    Regards

  • Hi,

    There is a configuration ccfg.c

    #define SET_CCFG_MODE_CONF_VDDR_CAP 0x3A

    0x3A ==> 5.8uF

    It seems recharge algorithm use this value to calculate recharge interval.

    Is it a valid option to change the value lower (e.g. 1uF) to make recharge period more aggressive ?

    Regards

  • Hi,

    Yes, you can wake up device from shutdown with pin interrupt (see example code in the SDK).

    Changing the value of VDDR_CAP is not recommended. This is used by RTOS to automatically adjust recharge timing. Further more, this value is determined by the capacitor on VDDR.

    regards