Throughput sample project slow speeds

[jspiller]

I am running the throughput sample project on sdk 1.9.0 using two nRF5340 dev kits and am getting a throughput of 197 kbps using 2M PHY, Data Length 251 and Interval of 7.5ms. According to the specifications from https://www.novelbits.io/bluetooth-5-speed-maximum-throughput/ and this thread https://devzone.nordicsemi.com/nordic/nordic-blog/b/blog/posts/throughput-and-long-range-demo it should be much higher.

[haakonsh]

Yeah, that is way to low of a throughput. Do you have a log or a sniffer trace?

[jspiller]

Yes, here is one I just ran that gave me 294kbps. I am getting quite inconsistent results now anywhere from 200kbps to 350 kbps, but still much slower than the expected values.

throughput 294kbps.pcapng

[jspiller]

I was able to reach 1292 kbps but the boards require no distance between them nor any obstructions. Does this seem correct? Even having just a dev board between them for a sniffer trace brought it down to 300kbps. It was my impression that it should have better range than this for these speeds.

[haakonsh]

A PCBA between the dk's will definitley reduce the throughput considerably, as the signal will not have to rely on reflections from nearby walls in order to reach the peer's antenna. 

At 1292kbps you are very close to the theoretical maximum (at zero range), but the throughput will fall drastically as you increase range. 300kbps is what I'd expect across a large office space filled with other 2.4GHz radios like wireless headphones, wifi devices, etc.