• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 52 subscribers
  • Views 905 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

CC2640 Reception Issues

Matt Sutton
Prodigy 70 points

We have a 2640 populated on our own board for which we are having what appears to be reception issues.   The device is programmed to operate as a peripheral and we can see it advertising from a large distance away.  However, we cannot reliably connect, nor can we reliably stay connected even at short distances.  Our antenna design is a simple pigtail antenna.  We are using internal biasing and we have configured the biasing as such in the firmware.  We've also configured the antenna as a single ended antenna.  We copied the transmit power table from the single ended, externally biased EM board specification in the firmware code.  Furthermore, we replicated on our board most of the filter network from the antenna to the micro pin as is in the EM board design.  Any ideas as to why we seem to be able to reliably broadcast at large distances but cannot seem to receive?  Any registers that we should be aware of on the device?  

  • 0
    TI__Mastermind 34660 points
    We may need to review your schematic to provide more information. Can you try without POWER_SAVINGS enabled to see if the error goes away? This would provide a good data point.
  • 0 in reply to Tim C
    Prodigy 70 points
    Tim. Thanks for your response. I think we have narrowed our problem down to using the internal low frequency RC oscillator clock. Unfortunately our board design is to small and compact to realistically include an external 32k RC oscillator. Is there any way yo use the 24 MHz oscillator in divide down mode? The docs seem to indicate so, but I'm concerned the RTOS will power down this oscillator. How do we force this oscillator to always run in C code? Thanks
  • 0 in reply to Matt Sutton
    Prodigy 70 points

    As a follow up again, we seem to have been able to circumvent our problem with the internal LF RC oscillator by setting the option on the Power XDC module named calibrateRCOSC equal to true.  However, we don't understand what this option is doing, other than it seems to solve our BLE connection reliability options.  Can we get any insight into what this option does?  We are aware of the errata on the package.  This causes us some obvious concern.  Is there any way we can get more elaboration on the errata?   So, our questions now are:

    1. What does the calibrateRCOSC = true on the Power module really do?
    2. Can we get any additional details regarding the errata concerning the LF RC oscillator given that transitioning now to an external oscillator for our board design is simply not a viable option.

    Thanks.

  • 0 in reply to Matt Sutton
    TI__Guru** 105180 points
    Hello Matt,

    Currently the mode you describe is not supported / endorsed by TI at this time with PG2.2 and BLE Stack 2.0, all BLE designs currently require an external 32kHz XTAL. You may find that it does work, but on some devices you will encounter the published erratum for PG2.2.

    You may want to use a Sleep Clock Accuracy of 500, this will be needed when/if an 32k RCOSC option is supported on future HW/SW.

    Best wishes
  • 0 in reply to JXS
    Prodigy 70 points
    JXS,
    We do happen to use a SCA of 500 now.
    Regards,
    Matt