NCS - Very short range even with Coded PHY

[GiulianoFranchetto]

Hi,

We are currently trying the Long Range (Coded PHY) on our advertised message, but we can't see any improvement (range of about 50meters). The range can even be smaller while using Coded PHY, which seems odd. 

We are using NCS and Zephyr and not the SoftDevice.

The configuration of the advertiser is

CONFIG_BT=y
CONFIG_BT_SMP=y
CONFIG_BT_SIGNING=y
CONFIG_BT_PERIPHERAL=y
CONFIG_BT_ATT_PREPARE_COUNT=2
CONFIG_BT_PRIVACY=y
CONFIG_BT_BROADCASTER=y
CONFIG_BT_CTLR_TX_PWR_PLUS_8=y
CONFIG_BT_EXT_ADV=y
CONFIG_BT_EXT_ADV_LEGACY_SUPPORT=y
CONFIG_BT_KEYS_OVERWRITE_OLDEST=y
CONFIG_BT_EXT_ADV_MAX_ADV_SET=3
CONFIG_BT_LL_NRFXLIB=y

The advertiser parameters are:

 

BT_LE_ADV_PARAM(
		BT_LE_ADV_OPT_EXT_ADV | BT_LE_ADV_OPT_CODED,
		BT_GAP_ADV_FAST_INT_MIN_2 * 2,
		BT_GAP_ADV_FAST_INT_MAX_2 * 2, NULL)

The scanner scans the packet with the correct PHY:

[DEVICE]: 54:7c:02:d1:fc:a8 (random),
RSSI -47
Prim: LE Coded,
Secn: LE Coded

What could explain having such small range? What configuration can I set to increase the range?

Regards

Giuliano

[run_ar]

Hi,

What is the length of your advertisement package? and what kind of environment are you testing in?

[GiulianoFranchetto]

Hi.

Does anyone have an idea? Can we get a nordic guy here?

Regards

Giuliano

[run_ar]

Hi,

I am sorry for the late reply, I've been trying to find out if we have evaluated coded phy with the zephyr controller.

Regardless coded phy gives a ~6 dB improvement in link budget if I remember correctly. This helps a lot if you have line of sight, without too much reflections (i.e https://blog.nordicsemi.com/getconnected/tested-by-nordic-bluetooth-long-range) However it might not help at all inside where you could have hard barriers such as concrete walls or outside in areas where there walls/cars and similar that causes reflections. One draw back with coded phy that is often neglected, is that it has much longer on air time (x8) so if there are Wifi networks in the area, coded phy is more likely to have errors as the S8 coding won't help against certain strong blockers (wifi or lte) as they will disturb the signal for a long time. In addition the antenna direction can affect the received signal strength with only minor changes in direction, and multi path reflections is another issue. I guess what I am trying to say is that it is hard to say what could go wrong with your test, but in certain environments you will not get an improvement with coded phy. What HW are you using? is this nordic development kits?  are you only looking at the range? or do you also monitor the RSSI? How much do the RSSI change when you move 0.5 meters, 1 meter, 2 meter etc at the range limit in your test?