Hello Dialog,
There is a bug in firmware 1.0.8 for 14681.
When requesting the buffer size over HCI the chip responds with 27, which should be 251 as max packet size over air in BT4.2 is 251 bytes using the "Data Packet Length Extension".
This reduces throughput vastly, as it literally makes the extensions useless, as packets are then fragmented in 27 byte pieces.
Attached is a trace of the incorrect response during boot.
Notice the frame
Note: If the response is ignored, the chipset does actually accept up to 251 bytes of ACL payload as it should.
有另一个bug, of completed packets is reported incorrectly, but that is another bug.
Thanks,
/pedro
Device:
Rename the PDF file to CFA to see the trace. Strange filter :-).
Ah it accepts ZIP, so here is a zipped version
Hi Donpedro,
Thank you for posting this question and providing detailed description.
I will look into this and I will get back to you as sson as I have an update for you.
Regards
TN_Dialog