Low LQI also in neigbour table (TZ-1100)

[xight155]

Answers checklist.

• I have read the documentation ESP Zigbee SDK Programming Guide and tried the debugging tips, the issue is not addressed there.
• I have updated ESP Zigbee libs (esp-zboss-lib and esp-zigbee-lib) to the latest version, with corresponding IDF version, and checked that the issue is present there.
• I have searched the issue tracker for a similar issue and not found a similar issue.

IDF version.

IDF 5.1.3

esp-zigbee-lib version.

1.4.0

esp-zboss-lib version.

1.4.0

Espressif SoC revision.

ESP32-C6

What is the expected behavior?

In older zigbee lib versions the LQI was must higher, also in the neighbour tables.

Also the distance >20meter can’t reach any more. Now +/- 15 meter with LQI 0 is max.

What is the actual behavior?

low LQI, distance 5meter with LQI +/- 20

Steps to reproduce.

test

More Information.

No response

[xieqinan]

@xight155

Regarding this issue, I need to confirm the following points:

• Could you please provide more details about your test environment?
• Which older version are you using for comparison with v1.4.0?
• Have you printed the RSSI in the neighbor table, which is more accurate than the LQI?
• Could you please modify the ieee802154_receive_done() function in ~/esp-idf/components/ieee802154/driver/esp_ieee802154_dev.c to check if the original RSSI value meets the requirement?

static void ieee802154_receive_done(uint8_t *data, esp_ieee802154_frame_info_t *frame_info)
{
    // If the RX done packet is written in the stub buffer, drop it silently.
    if (s_rx_index == CONFIG_IEEE802154_RX_BUFFER_SIZE) {
        esp_rom_printf("receive buffer full, drop the current frame.\n");
    } else {
        // Otherwise, post it to the upper layer.
        frame_info->process = true;
        esp_rom_printf("rssi: %d\n", frame_info->rssi);
        esp_ieee802154_receive_done(data, frame_info);
    }
}

[BaWe155]

@xieqinan
Thank you for the quick reply, we are testing the network (with 2/3 hops) now in a warehouse with a free line of sight.

1.1.2 works well, with a join lqi of 5. I will test the code below, normally I get the Lqi and neighbour table from the control bridge/coordinator

[BaWe155]

I tested the function with a CB and three nodes (as router), two routers with a distance of 1 meter of another and a router with a distance of 15/16 meters

Router A with mac:404C ----- 1028, RSSI -48 and -87

Distance A and B is 1meter

Router B with mac:5432 -----38f4, RSSI -62 and -101

I think the module batch start with MAC:5432 have a poor/weak RX (I tested several modules starts with 5432). How can I check if the batch with 5432 is a 100% original Espressif C6 module?

[BaWe155]

Please check the refreshing of the neighbour table (of a router with V1.4.0) and lqi, Lqi is very low compare with a router running with 1.1.2 Zigbee version.

[xieqinan]

@BaWe155 ，

I think the module batch start with MAC:5432 have a poor/weak RX (I tested several modules starts with 5432). How can I check if the batch with 5432 is a 100% original Espressif C6 module?

I think you can run esptool.py -p /dev/ttyUSB1 flash_id to get the BASE MAC, and then verify the MAC address on the OUI page.

Please check the refreshing of the neighbour table (of a router with V1.4.0) and lqi, Lqi is very low compare with a router running with 1.1.2 Zigbee version.

Were the test results for esp-zigbee-sdk v1.4.0 and v1.1.2 generated in the same environment? If so, please modify the ieee802154_receive_done() function according to the instructions in #418 (comment) and then provide us with the test results.

[BaWe155]

Hi @xieqinan,

I will test it, can you check/compare it if there is anything changed with the neighbour table and refreshing of it?
I see nodes in the table who are longtime offline and there are better neighbours available.

[BaWe155]

Hi @xieqinan
I checked the Mac bove are original ESP32-C6 modules but a difference of +/- -20dBm, how can I get in the future the modules with a good RX level? maybe ask a colleague about the hardware?