BLE performance of BlueNRG-2N

[path321]

Hello everyone,

I am designing a custom board (STM32L476 MCU paired with BLE connectivity, among other components) and after search decided to use BlueNRG-2N module along with Antenova A5839 Antenna.Tried to follow AN4819, despite the tight PCB space constraints. However,I get quite poor performance (~2 BLE devices found in 10s scan), especially when compared with X-NUCLEO-BNRG2A1 (which is equipped with BlueNRG-M2SP module, ~10 BLE devices found in 10s scan) , while running the same SW. Respective pics of board and schematic attached below. Any idea to improve the PCB design ?

Thank you in advance

[Laurent LOUAZON]

Hello,

Looking at schematics I see few potential concerns

• Seems High speed xtal selected is having a 10pF load cap.

If so, C16/C17 should be much more around 18pF.

So may be High speed oscillator is a bit shifted (See AN4818)

• L2 value (100nH) seems really high.

Idea is to prevent 2.4Ghz RF pollution in High Speed oscillator

Recommended value is much more 2.2 or 3.3nH.

• Finally, no RF matching is present (between balun and antenna) so maybe antenna is not radiated properly.

From layout side, I see no real GND plane here.

It could mean again that radiated perfs are not really good.

As you mention, seems problem is related to slow discovery of your PCB.

I would suggest to look at

• If you are facing a range problem due to poor radiated perfs (ground plane or antenna matching). Do you have same problem if devices are set close one from each others?
• High seep oscillator centering

Then you may contact your ST sales representative or distributor in order to get access to ST support for deeper analysis eventually.

Regards

Laurent