My question is pertaining to running of sniffer to show my device name

[khosla]

50 MHz, 500 M Samples [16].logicdatamainc.txt(/attachment/98d67704ad9c5affad8d5bce2ef93315)(/attachment/f20245cb8f1c01edb4b17875b6eedc4b)01200555.pdfI am using the radio with ppi to transmit as well as receive packets through the same code. I need to record the time events.I have enabled a special button functionality to distinguish between the transmit and receive events.

I am attaching the mainc file under. I have thoroughly checked the address, and other values that i put in the packet. If you could let me know what could be the problem behind my device named "Raghav" not being showed on sniffer, even though the code is built and run properly. I even try to set break point at timer init and commented out 'app button is pushed' function but no change occurs.

I am attaching my mainc file .

[endnode]

Hi, your source code seems to be missing here. However from your description more items are unclear: what is your code based on? Was it ever working (meaning was it ever advertising)? The best approach is to start from some proven SDK example before writing bunch of code from scratch...

[khosla]

I have attached two files in my initial comment, one mainc and the other pdf of the research paper i am using. seeing fig 1 in that paper, u will know what i am doing.

to summarize, i am basically working on synchronizing time between two neighboring nodes as well as between distant nodes. so for that i am timestamping each of the send and receive events at the nodes. Considering two nodes, 1 and 2, the timestamp first being generated when the packet is sent from 1 to 2. then, when 2 receives it, another timestamp is generated. There is some delay, when 2 sends an acknowledgement back to1. so again now just when the ack is sent form 2, now 2 being the transmitter, another timestamp at node 2 and similarly when the ack is received at 1, then another timestamp collected. so total 4 timestamps being collected in each cycle. The packets being transmitted on radio carry the timestamp info generated by the hardware timers used at each node.

Fo that i need to see the time samples in wireshark and analyser. i am using sdk 12.2 and nrf51 board in segger studio. I need to basically capture the transmit and receive events.

[endnode]

OK, so you hope to do some kind of "time of flight" demo on nRF5x and for that you are (obviously) not using Soft Device but you need to command radio front-end directly. That looks correct. Have you used any SDK example as source? Why do you assume that you will see your packets in BLE sniffer? Are they even valid BLE packets? What sniffer are you using?

[khosla]

i am using radio with ppi functionality, since i am using harware timers to capture events and do need CPU intervention. i would be seeing the packet receive times in wireshark corresponding to my device. That would be advertising ind packets. Then i ought to see the toggle event times on the analyser. I am using the button functionality as u see in my code to distinguish between transmit and receive events.

I am just using wireshark and analyser for verification. Ofcourse i can't rely on their time samples, as i don't know which clocks they are using. But i would be using Segger RTT viewer to communicate with my board and then Pynrfjprg interface to transfer the data from the board to the PC. i AM USING THE STANDARD NRF Bluetooth smart sniffer from nrf51 downloads

Earlier the sniffer was running okay at the initial stages and my device showed up, but lately i had to modify the functionality to load the timestamp in the packet prior to it being sent as well as calling the function app button is pushed in the main function which represents a transmit event. So i think the address and the payload values i loaded in the mesh demo packet are same as earlier, but now after everything seems okay in the code, i can't see my device. If u could analyse the problem and get back to me.

[khosla]

Hi, I even tried to resolve this by setting a break point on "timer init". But when debugging while running in SES, and pausing, there is no call stack. This seems to be quite strange.