This post is older than 2 years and might not be relevant anymore
More Info: Consider searching for newer posts

nRF51 BLE performance issues with Raspberry Pi B+

I am trying to use a Raspberry Pi B+ as Bluetooth to Wifi bridge. But I find that the bluetooth range is very diminished when talking to my Nordic nRF51 and nRF52 devices. But I have some other chipsets that seem to have better blluetooth range with the same Raspberry Pi device. My BLE sniffer shows the channel map shrinking as well, so I suspect interference is involved, but it seems isolated to the nRF devices.

Is anyone else seeing this type of behavior? I see similar discussions on Raspberry Pi blogs/support sites, so it might not only be a nRF issue - it may be both! I'm just curious if anyone has seen this behavior specifically with nRF devices.

Thanks!

  • HI Paul,

    But I find that the bluetooth range is very diminished when talking to my Nordic nRF51 and nRF52 devices.

     Are these custom nRF5x devices or are you using nRF5x DKs? If its the first, have you submitted schematics and layout for review of the RF section?

    My BLE sniffer shows the channel map shrinking as well, so I suspect interference is involved, but it seems isolated to the nRF devices.

     Please attach the sniffer trace, then we can look at the RSSI values and number of re-transmits etc. 

    Best regards

    Bjørn

  • Hi Bjorn,

     Thanks for your reply.  We figured out how to make the connection more reliable.  The issue was that the raspberry pi was setting the connection rate a little longer (~48ms) and the supervisor timeout a little shorter (~450ms) than other central devices we had been using (iOS devices) and so we had very few (~9) connection interval in order to keep the connection alive at longer ranges.  

     I'm still not 100% sure why our "custom" Nordic devices had shorter rang than other devices.  An nrf51 DK we have also had shorter range, but a Nordic "Thingy" we tested had longer range (nRF52 I believe).  

     But for now the crisis is averted.  We have a satisfactory solution by increasing the supervisor timeout and making the connection interval a little faster.

    Cheers!
    PG

  • Hi PG, 

    if the matching network for the antenna of the custom boards has not been tuned then that could be the reason for the reduced range, i.e. that you have losses in the RF path that reduce the radiated output power. 

    However, I am glad that you now have a satisfactory solution.

    Best regards 

    Bjørn

Related