Any additional debugging msg available for checking connection issue?

[Anonymous]

Hi,

  I can manage to produce one weird GATT connection symptom by using LightBlue and BCM20732 module.

  Is there any additional debugging message available from LL / PHY for diagnostic?

  BCM20732 act as a peripheral (running application 'hello_sensor') and LightBlue is running on iPhone as a central.

  Steps:

  1. Use LightBlue to create connection to BCM20732

  2. When connection is made, disconnect it from LightBlue. From console output of BCM20732, the connDown callback is not yet called immediately after disconnection made at LightBlue side. (Any one can explain this?)

  3. Wait for while, the connDown callback is truly called (observed from console output of BCM20732).

  4. Trying to make connection again from LightBlue. Then it will failed with 'Disconnection from peripheral' immediately and there is no any connection callback available from BCM20732.

  5. Usually, it will take a while then you can connect it again (connUP callback available from BCM20732). Not sure exactly how long is this period. But it would take roughly 30 sec or more.

[Anonymous]

Hi,

  Another finding is that it seems to be difficult to connect to peripheral when there are a lot of BLE devices around.

  In my environment, there are roughly 10 BLE devices is at advertising state.

  Is there any limitation for BLE in initiating connection from channel competition when several peripherals are advertising at the same time?

[VictorZ_46 · Accepted solution]

Tried to play with Lightblue and hello sensor.  Indeed it looks that every other time Lightblue disconnects after 10 seconds rather than right away.  Really have no idea why.  If you enter the same device right away there is no reconnection and you can read characteristics.  So lightblue did not even try to disconnect.  Meanwhile I cannot reproduce your problem with failed connection.  After disconnection every time I try to connect it works like a charm.  If you include following 2 lines in the beginning of the create function, there should be a bit more traces coming from the stack.  But I am not sure what you will see for the problem like that.  Over the air traces maybe the only choice.
 

extern UINT32 blecm_configFlag ;

blecm_configFlag |= BLECM_DBGUART_LOG | BLECM_DBGUART_LOG_L2CAP | BLECM_DBGUART_LOG_SMP;

[Anonymous]

Thanks, victorz.

I'll try your suggestion for gathering more log message if I can still reproduce it.

I'm also not able to see connection issue with different environment case, I doubt that this might related to the crowded situation of bunch peripherals around my device.

Thanks for your suggestion.