sd_ble_gap_tx_power_set not working

[eduardoescarti]

Hello! I'm using nRF5_SDK_15.0.0_a53641a to programm my Nordic NRF52832 with Segger Embedded Studio.

I'm trying to increase the BLE Tx power to 4dBm. At the moment I've tried the following:

1. CHANGING THE TX POWER AFTER INIT SEQUENCE RIGHT BEFORE THE MAIN LOOP

[... init stuff ...]

err_code = sd_ble_gap_tx_power_set(BLE_GAP_TX_POWER_ROLE_SCAN_INIT, BLE_CONN_HANDLE_INVALID, -20); APP_ERROR_CHECK(err_code); 
for (;;) {

}

RESULT: Absolute no change in advertising RSSI signal value

2. CHANGING THE TX POWER WHEN INITIALIZING ADVDATA

static void advertising_init(void) {

[...]
uint8_t power_level = 4; 
init.advdata.p_tx_power_level = &power_level;
[...]
}

RESULT: Absolute no change in advertising RSSI signal value

3. CHANGING THE TX POWER AFTER CONNECTING TO BLE DEVICE

case BLE_GAP_EVT_CONNECTED:

 

[...]
err_code = sd_ble_gap_tx_power_set(BLE_GAP_TX_POWER_ROLE_CONN, m_conn_handle, 4);

APP_ERROR_CHECK(err_code);

RESULT: Absolute no change in RSSI Signal value after connection

Am I missing something here???

[MartinBL]

Hi,

How are you doing your tests?

Case 1 & 3 should have an effect, but unless you use proper test equipment and an RF chamber then small changes in TX power (from 0 to 4 dBm e.g.) might not be practically measurable. However, if you switch between e.g. -30 and +4 dBm the effect should most definitely be noticeable. 

Case 2 does not actually change the TX output power of the radio at all. It just updates a value in the advertising packet. 

If this doesn't help, do you mind uploading some code I can use to test exactly what you are doing? 

[eduardoescarti]

Hi MartinBL,

Im currently testing on the PCA10040 Development board.

For the RSSI Signal measurement I'm using the iOS nRF Connect App.

The modules I'm initializing are:

int main(void)
{
    bool erase_bonds;
    
    NRF_POWER->DCDCEN = 1; 
    
    timers_init();
    buttons_leds_init(&erase_bonds);
    power_management_init();
    saadc_init();
    ble_stack_init();
    gap_params_init();
    gatt_init();
    services_init();
    advertising_init();
    conn_params_init();
    peer_manager_init();
    spi_init();
   
    advertising_start(erase_bonds);

    err_code = sd_ble_gap_tx_power_set(BLE_GAP_TX_POWER_ROLE_SCAN_INIT, BLE_CONN_HANDLE_INVALID, -40); 
    APP_ERROR_CHECK(err_code);
    
    for (;;)
    { 
        /* left empty for testing purposes */
    }
}

 

changing the parameter of sd_ble_gap_tx_power_set form -40 to 4 does absolutely nothing to the RSSI signal value.

This being around -42dBm when I place the phone about 10cm away from the devel board and around -60dBm when the board is about 1m apart.

I suspect that one particular module or my particular BLE configuration are getting in the way.

I'll test with the examples provided to see if I can reproduce the behaviour.

Best,

Eduardo.