Questions about ESB RSSI value

[Lurn_Z]

Hi all,

In this link, it said the rssi value should be a negative value, but the register stored a positive value.

So when I read a value of 60 for example the RSSI is -60dBm.

But in my test, I use this code to choose a  relatively stable channel.

static int esb_get_rf_channel_rssi(uint8_t ch)
{
    int rssi_reading = 0;
    nrf_esb_set_rf_channel(ch);
    nrf_esb_start_rx();
    nrf_delay_us(300);
    for(int i = 0; i < 10; i ++)
    {
        NRF_RADIO->TASKS_RSSISTART = 1;
        while(NRF_RADIO->EVENTS_RSSIEND == 0);
        rssi_reading += NRF_RADIO->RSSISAMPLE;
        nrf_delay_us(100);
    }
    nrf_esb_stop_rx();
    nrf_delay_us(300);
    return rssi_reading/10;
}

the result is here

I scanned it multiple times, and choose a channel with a smaller value change.

At first, I choose a minimum value(e.g. 44) to set the channel and the packet lost rate almost 50%, at 1.5M without retransmit

Then I choose the max value( e.g. 100 or something), in this channel I found the packet lost rate about 10%, 1.5M without retransmit.

I think the - 60dBm(for example) should better than -100dBm.

Note: the channel with minimum value has a wider range of fluctuations than the channel with max value.

So, Is there something wrong with my test? Or my My test environment is bad?

The SDK is 17.1.0, nRF52832.

Best regards,

Lurn

[ovrebekk]

Hi Lurn

If you take a look at the register documentation here you will see that it reports positive values in the range 0-127, and in order to get the actual RSSI value you need to multiply by -1. 

In other words when you read a smaller value such as 44 this actually means that the interference is stronger, and that this is a poor channel compared to the others. 

In your particular case it seems that channels 9-12 are hit by pretty heavy interference, and if you can avoid using these channels it should greatly improve the robustness of your communication. 

Best regards
Torbjørn

[Lurn_Z]

Hi Torbjørn,

Maybe I misunderstood this value, what I think is that this value is a negative dBm value for the channel. So the smaller value such as 44 actually is -44dBm. And for the dBm value -44 should better than -100.

In your explanation, I think the value should be a strength of the interfering signal about this channel, in this way my test results can be explained.

Thanks a lot.

Best regards,

Lurn

[ovrebekk]

Hi Lurn

Exactly, most likely you are not receiving a valid packet at this point so the RSSI you see is the signal strength of the interference. In other words a stronger signal means more interference, which will lead to worse performance. 

Best regards
Torbjørn

[Lurn_Z]

Hi Torbjørn

Yes, at that time I was not receiving packet just read the rssi value, so you are right, the value  I saw is the signal strength of the interference. 

Now I understand it, thanks for your help.

Best regards

Lurn

[ovrebekk]

I am happy to help Lurn. The best of luck with your project