NRF52 RX radio channel scanning receive probability

[astomi]

Hi,

SDK17, SES. NRF52840 host receiver. NRF52810 custom board asynch transmitters, Proprietary radio. 

Assuming there is a handful of NRF52810 asynchrounous "slave" transmitters sending 6 bytes paylod each on a different(pre-assigned) frequency channel (no hopping for now, belongs to a group of ,say, 12 channels out of 100 available) but one "master" receiver.

If there is one master receiver that scans all 12 channels all the time (at what rate?) - is there a reliable scheme to estimate a chance of packets lost? Specificaly:

1. When master's radio RX and HFCLK is always on, just channels frequency is switched in a round robin fashion, is my understaning correct that the reciever just needs to have an overlapping RX window during slave transmitter preamble send on air in order to detect an asynch sender? 
2. Is it possible for a master receiver radio to generate an event upon the successful preamble detection?
3. Would it be correct to assume that if the preamble is successfully detected, the master would need about 2.5ms for the receive (6 bytes payload) and simple ACK back?
4. I presume that 100 frequencies are available under 1 Mbit mode only?

ks guys for your help in advance.

[Håkon Alseth]

Hi,

 

When master's radio RX and HFCLK is always on, just channels frequency is switched in a round robin fashion, is my understaning correct that the reciever just needs to have an overlapping RX window during slave transmitter preamble send on air in order to detect an asynch sender? 

This highly depends on how you have setup your protocol. If it switches RF channel mid-received-payload, then you can potentially get parts of the payload, but not the remainder of the payload.

Is it possible for a master receiver radio to generate an event upon the successful preamble detection?

Not for preamble, but there is an event for address match. See here:

https://docs.nordicsemi.com/bundle/ps_nrf52805/page/radio.html#ariaid-title8

Would it be correct to assume that if the preamble is successfully detected, the master would need about 2.5ms for the receive (6 bytes payload) and simple ACK back?

depends on your overall requirements for re-transmission etc, but for one TX->RX / RX->TX transaction, 2500 us sounds a bit high. Again, this depends on the on-air data rate etc. as well.

I presume that 100 frequencies are available under 1 Mbit mode only?

All channels are available for all supported data rates, but please be aware that your local tele-regulatory requirements will limit this range, typically from 2402 MHz to 2480 MHz.

If you are using BLE data rate, the channel is wider, and you need 2 MHz spacing - effectively cutting the amount of channels in half.

 

Kind regards,

Håkon

[astomi]

Hi,

Thanks for coming back to me on this.

It is my understanding that radio frequency can't be set on the fly, so looking at the specs with fast ramp up, shorts enabled, one byte preamble and minimum address config of 2 bytes + prefix I am getting approx 72 us window = 40 us enable + 32 bit x 1 Mbit rate + 0 us disable. Plus I guess around 5 us to set new frequency. 80 us -ish in total.

Does it sound right?

[Håkon Alseth]

Hi,

 

astomi said:

It is my understanding that radio frequency can't be set on the fly

Yes, that is correct. there is a 40 us penalty when changing channel for ramp-up.

 

astomi said:

It is my understanding that radio frequency can't be set on the fly, so looking at the specs with fast ramp up, shorts enabled, one byte preamble and minimum address config of 2 bytes + prefix I am getting approx 72 us window = 40 us enable + 32 bit x 1 Mbit rate + 0 us disable. Plus I guess around 5 us to set new frequency. 80 us -ish in total.

That is correct. The minimum will be preamble + 3 byte addr, ie. 40+32=72 us.

 

Kind regards,

Håkon

[astomi]

thanks