Hi, I'm i'm sniffing the NRF51822 s110, using the Nordic's BLE sniffer (on s120).
The BLE device is in a connection, and do a connection parameter update to a new interval of 108.75ms (indicated as 87 in the LL_CONNECTION_UPDATE_REQ).
I'm trying to look in the sniffer and to match this value to the measure delta time, and there is a constant difference between the theoretic and practical value.
In the attached screenshot, please look at packet #142 and #141. by looking at the delta time (start to start) we see that the total delta time from packet #140 and packet #142 is measured as 109,054us (=108,672 + 382).
This value should represent the connection interval, and i would expect it to be much closer to the nominal 108,750us.
So the difference between the actual value and the nominal value is 304us.
Why there is such a big difference?
This difference appears to be very common (although its value may change a bit). I currently assume that the timing is measured by the s120 MAC timestamp.
Here is the attached screenshot of the wireshark:
thanks, Hagay.
Update:
If i compute the difference using end to start delta i get much more logical information.
end of #141 to start of #142 = 108,440us
end of #140 to start of #141 = 150us
#140 and #141 packet length (empty PDU) = 9 Bytes + preamble = 10 Bytes
#140 and #141 total TX time = 108+108 = 160us
total time = 160+150+108440 = 108,750.
I think you have an issue with the timing, and probably it relate to the added data by nordic (nordic sniffer BLE meta) causing the packet to be with wrong length?
