Advertisement with LE Coded PHY

[Ben]

Hi,

I have recently been evaluating the nRF52840 for the research purpose. I came across a question related to this product that I could not figure it out myself. 

I noticed that when I use Coded PHY, the advertisement take places in uncoded 1M PHY. But this affects the range, as the advertisement is not in Coded PHY. 

I was reading the Bluetooth standards (version 5.1) and I found out that in section 4.4.2.4.4 of the standard, the advertisement is also possible using ADV_EXT_IND on the LE Coded PHY.

Now the question is, does nordic uses this feather? Can we have advertisements on LE Coded PHY in nRF52840?

Best regards,

Ben

[Vidar Berg]

Hi Ben,

Yes, our S140 Softdevice supports CODED PHY. You can easily test this if you have access to two 52840 DK. The ATT_MTU Throughput Example includes this mode. I have also posted a modified version of the ble_app_uart/ble_app_uart_c here: https://devzone.nordicsemi.com/f/nordic-q-a/40476/unable-to-convert-to-long-range-after-looking-at-umpteen-examples--/157300#157300 that sets the max output power as well as using code phy. 

Best regards,

Vidar

[Ben]

Hi Vidar,

Thank you very much for the prompt reply. 

However, I tried to use the CODED PHY for advertisement in ATT_MTU_Throuput example but I am not successful. 

I changed the following line of codes:

adv_params.primary_phy = BLE_GAP_PHY_1MBPS;
adv_params.secondary_phy = BLE_GAP_PHY_1MBPS;

adv_params.properties.type = BLE_GAP_ADV_TYPE_CONNECTABLE_SCANNABLE_UNDIRECTED;

When I run it on the board, it does not throw an error but just simply stops!

Do you know the reason?

Br,

Ben

[Vidar Berg]

Hi,

What did you change the type to? Should be BLE_GAP_ADV_TYPE_EXTENDED_CONNECTABLE_NONSCANNABLE_UNDIRECTED. 

[Ben]

Hi Vidar,

Yes I have changed the type to BLE_GAP_ADV_TYPE_EXTENDED_CONNECTABLE_NONSCANNABLE_UNDIRECTED. 

"

adv_params.primary_phy = BLE_GAP_PHY_CODED;
adv_params.secondary_phy = BLE_GAP_PHY_CODED;

adv_params.properties.type = BLE_GAP_ADV_TYPE_EXTENDED_CONNECTABLE_NONSCANNABLE_UNDIRECTED;

"

But I get this :

"

[00:00:00.000,091] <info> app: Preparing the test.
[00:00:00.000,122] <info> app: Starting advertising.
[00:00:00.000,244] <info> app: Starting scanning.
[00:00:00.050,292] <info> app: Device "Nordic_ATT_MTU" found, sending a connection request.

"

It found the device but then it stops!!

Br,

Ben

[Ben]

Hi,

I also add this to the scan parameter:

"

static ble_gap_scan_params_t m_scan_param =
{
.active = 0x00,
.extended = 0x01,
.interval = NRF_BLE_SCAN_SCAN_INTERVAL,
.window = NRF_BLE_SCAN_SCAN_WINDOW,
.timeout = 0x0000, // No timeout.
.scan_phys = BLE_GAP_PHY_CODED, // BLE_GAP_PHY_CODED BLE_GAP_PHY_1MBPS
.filter_policy = BLE_GAP_SCAN_FP_ACCEPT_ALL,
};

"

And then I added this line of code:

"  err_code = sd_ble_gap_scan_start(&m_scan_param, &m_scan_buffer);" 

to scan_start() function.

But still, the code doesn't work and it stops after finding the Nordic device!!!

Br,

Ben

[Vidar Berg]

Hi Ben,

It looks like the example will require a few modifications to advertise and scan with CODED PHY. Did you try the other example I linked to?

[Ben]

Hi Vidar,

I have not done the other. Since I am interested in The Throughput example. We are doing a heatmap of RSSI in the indoor environment for studying BLE 5 in the university. 

I would appreciate if you help me with this example! If you could share the code that works with COED PHY, really would save my time for now. 

Br,

Ben

[Vidar Berg]

Here you go: ble_app_att_mtu_throughput.zip. I made CODED PHY the default. 

Edit: It's based on SDK v16.0.0

[Ben]

Hi Vidar,

Thanks a million. This is brilliant. 

I have a question though!

1) Why changed the att_mtu from  NRF_SDH_BLE_GATT_MAX_MTU_SIZE to BLE_GATT_ATT_MTU_DEFAULT ?

2) I also noticed that in test_begin function "m_test_params.phys.tx_phys" is 2M PHY!! This means that the test starts with 2M PHY!!? I changed it to coded PHY. 

Best regards,

Ben

[Vidar Berg]

Hi Ben,

1. It wasn't a necessary change. It is possible to use longer MTUs with this PHY.

2. Good that you noticed it. I didn't. The test will still be conducted with CODED. At least on the first run because the test starts in default mode.