nrf52833 + Long Range + Scanner

RE: nrf52833 + Long Range + Scanner

embeddedER — Thu, 16 Jun 2022 16:23:04 GMT

Hi Amanda,

Appreciate your detailed response.

Few clarifications from my end:

- I am aware of the factors affecting range (txpower, antenna, environment etc). That's the reason I mentioned I am using same environment to test default vs coded phy.

- I used mobile, just to verify that I am sending data using Coded PHY. In the nRF app, I do see primary and secondary phy as "LE Coded" which means we have setup transmitter correctly with Coded PHY. For our testing we are using 2 nrf52833 dev boards (fanstel). On the scanner as well I am printing primary/secondary phy for advertisements which shows "LE Coded".

- We have already tested txpower=4/8 with nrf52833 and planning to tune that as per our final need. Since txpower comes with a trade-off of high power consumption and our solution is battery operated device.

Ultimately the answer that we are trying to get here is: with the same setup/antenna/env/txpower how much coded phy out perform default phy?

Since our application uses advertisement packets with our data stuffed in it, we are trying to get max range with these adv packets.

Changes you mentioned in bt_conn_le_phy_update(), does it affect the advertisement packets as well? My assumption is, S=2/S=8 will come into affect only after connection, is that right?

Is there any other configuration (apart from txpower/antenna) that can help to improve the range of Coded PHY advertisement packets?

TIA!

RE: nrf52833 + Long Range + Scanner

Amanda Hsieh — Thu, 16 Jun 2022 13:22:55 GMT

Hi,

Bluetooth range is affected by many factors, and an issue when trying to predict what kind of range you get when phones is involved is that the radio performance of the phone itself is an unknown factor.

[quote user="harshal034"]I've confirmed adv packets are coded PHY in the scanner side by printing the data from adv packets on terminal and also using the nRF mobile App.[/quote]

I would suggest using two nRF53833 to test instead of the mobile device. What is the device, custom board or nRF53833DK? Can you test with nRF52833DK?

If you want to test with the phone, Please check the phone support Coded PHY or not as this post indicated. from nRFConnect mobile app manual (left top) -> Device information.

[quote user="harshal034"]My question here is, how do we configure S=2 or S=8 in adv packets?[/quote]

If you use I.E. the HR_Coded examples in NCS, the simplest way to implement S=2 would be to do it as part of a connection PHY parameter by calling bt_conn_le_phy_update(). See https://github.com/nrfconnect/sdk-zephyr/blob/main/include/zephyr/bluetooth/conn.h#L491. The preferred coding scheme is part of the PHY options: https://github.com/nrfconnect/sdk-zephyr/blob/main/include/zephyr/bluetooth/conn.h#L90

[quote user="harshal034"]Is there anything else that needs to be configured for the maximum range?[/quote]

Please make sure in the logs or sniffer that you in fact have a coded phy connection.

what is the TX power for each set to? The TX power is 0 dBm by default in the existing Bluetooth: Peripheral Heart Rate Monitor with Coded PHY and Bluetooth: Central Heart Rate Monitor with Coded PHY sample. To change this you have to send a HCI command to the Bluetooth controller as demonstrated by the Bluetooth: HCI Power Control sample. This post might help.

If you do in fact use Coded PHY but still only see this tiny increase in range, here are a few reasons for this:

Your test environment is noisier at >700m.
The custom board is not able to transmit longer than this for some reason. Not optimal antenna tuning or the application is not transmitting messages on the Coded PHY optimally.
Weather conditions during testing (rain/snow will decrease range considerably)

-Amanda

RE: nrf52833 + Long Range + Scanner

embeddedER — Thu, 16 Jun 2022 01:08:42 GMT

Hi Amanda,

I am currently measuring the range in the lab (not with clear line of sight). And it's around 30 meters where I see packets getting dropped with a default phy. At the same location, I am expecting Coded PHY to perform better. Also, I've confirmed adv packets are coded PHY in the scanner side by printing the data from adv packets on terminal and also using the nRF mobile App.

I've exact same configuration as mentioned in the sample application. My question here is, how do we configure S=2 or S=8 in adv packets? Is there anything else that needs to be configured for the maximum range?

TIA!

RE: nrf52833 + Long Range + Scanner

Amanda Hsieh — Wed, 15 Jun 2022 08:47:26 GMT

Hi,

What is the distance you get? Can you compare the distance while using Central Heart Rate Monitor with Coded PHY and Peripheral Heart Rate Monitor with Coded PHY?

-Amanda

RE: nrf52833 + Long Range + Scanner

embeddedER — Mon, 13 Jun 2022 17:42:33 GMT

Hi Amanda,

Thanks! I got it working after checking the configuration in the sample. With changes, I do see primary and secondary PHY of my adv as "LE Coded" now. However, I don't observe much change in the range compared to default phy.

I've kept the advertiser and scanner at a distance where I see drop in few packets with default phy. Then I replaced the default phy with Coded PHY and was expecting better performance (in terms of packets dropped) at the same distance. But I don't see much difference, even at the same distance, I see packets dropped with both default/coded phy. I am using same txpower for both the case. And experiments are done using nrf52833 on both the sides.

Is there anything else I need to configure to increase the range of advertisement packets with the Coded PHY?

This is what I changed to make my advertisements coded PHY.

	struct bt_le_adv_param param =
		BT_LE_ADV_PARAM_INIT(
				     BT_LE_ADV_OPT_USE_NAME | BT_LE_ADV_OPT_CONNECTABLE | BT_LE_ADV_OPT_ONE_TIME | BT_LE_ADV_OPT_CODED | BT_LE_ADV_OPT_EXT_ADV,
				     BT_GAP_ADV_FAST_INT_MIN_2,
				     BT_GAP_ADV_FAST_INT_MAX_2,
				     NULL);

TIA!

RE: nrf52833 + Long Range + Scanner

Amanda Hsieh — Thu, 09 Jun 2022 07:59:22 GMT

Hi,

Did you enable CONFIG_BT_CTLR_PHY_CODED=y in the scan sample? I would suggest taking a look at the prj.conf and main.c of the Bluetooth: Central Heart Rate Monitor with Coded PHY sample.

Regards,
Amanda

RE: nrf52833 + Long Range + Scanner

embeddedER — Tue, 07 Jun 2022 23:27:46 GMT

Just to add some more information, our adv parameters are:

    struct bt_le_adv_param param =
        BT_LE_ADV_PARAM_INIT(
                     BT_LE_ADV_OPT_USE_NAME | BT_LE_ADV_OPT_CONNECTABLE | BT_LE_ADV_OPT_ONE_TIME | BT_LE_ADV_OPT_CODED | BT_LE_ADV_OPT_EXT_ADV,
                     BT_GAP_ADV_FAST_INT_MIN_2,
                     BT_GAP_ADV_FAST_INT_MAX_2,
                     NULL);

So, it's extended Advertisements. And we can see them on nrfConnect App on S22 with primary and secondary PHY set as "LE Coded".

Are there any changes that we need to sniff/scan extended adv packets with nrf52833?