How to deal with the power accuracy issue with nRF7002 and nRF7000?

[JoeFu]

Hi Nordic,

I test with nRF7002-DK and IC mark on nRF7002 is below:

N7002
QFAAB0
2252AA

 

Here is my test data, showing about a 3dB discrepancy between power setting and measurement power (left sheet). When I set a higher power, I also get a higher measurement power (right sheet). Therefore, I would like to ask if there is a calibration mechanism to ensure that when I set 16dBm, the measurement power is also 16dBm, or how to deal with the power accuracy issue? I believe that a difference of less than 0.5dB between the power setting and measurement power is reasonable. Otherwise, there will be issues with Tx certification due to mismatching between power setting and measurement power.

 

[JoeFu]

Hi Ketilijo,

Thanks for your feedback. I want to know AntGain is peak gain or average gain.

I can find how to set antenna gain. However, there is no way to set Kconfig value. Could you explain that?

[ketiljo]

Peak gain, as this will affect EIRP.

JoeFu said:

However, there is no way to set Kconfig value. Could you explain that?

Have you looked here? https://docs.nordicsemi.com/bundle/ncs-latest/page/nrf/protocols/wifi/regulatory_certification/antenna_gain_compensation.html#setting_antenna_gain_parameters

[ketiljo]

Hod do you measure the power? It should be peak power. You also need a SA with sufficient bandwidth, >20 MHz, or integrate the power over the channel.

[JoeFu]

I use Litepoint IQ-MW to test average and peak power. I can share my test result: www.dropbox.com/.../Wi-Fi-test.xlsx

We have always used average power for testing, and with other solutions, there is no issue of power not matching the settings. If there is a mismatch of around +-1dB, you can adjust the power by setting Tx_gain through commands.

[JoeFu]

You provide command is about "compensate" not "calibrate". Is there a command about "calibrate"? 

[JoeFu]

any update for this case?

[JoeFu]

I use different fw setting to test Tx power. One is CONFIG_NRF70_ANT_GAIN_2G=CONFIG_NRF70_ANT_GAIN_5G=0, and another is CONFIG_NRF70_ANT_GAIN_2G=CONFIG_NRF70_ANT_GAIN_5G=6. I find there is still a gap form setting and measure bt 3.5~1.5dB. In 3590RER004_DRAFT - EN 300 328 V2.2.2.pdf p9, the setting is wifi_radio_test_tx_power 20 and we get 15~19 dBm Max Burst EIRP. So, is Nordic willing to admit that a 3dB difference between set power and measured power is normal and does not require adjustment?

Thanks.

 3590RER004_DRAFT - EN 300 328 V2.2.2.pdf

		Kconfig=0	Kconfig=6		Kconfig=0	Kconfig=6
nRF7002-DK	2.4GHz	Avg. power (dBm)	Avg. power (dBm)	5GHz	Avg. power (dBm)	Avg. power (dBm)
Power setting 16dBm	1M	13.2	13.6	VHT20_M0	12.4	13.5
	11M	13.4	13.7	VHT20_M7	12.8	14.0
	6M	13	13.6	HE20_M0	13.1	14.3
	54M	13	13.6	HE20_M7	13.2	14.5
	HT20_M0	13.1	13.5
	HT20_M7	13	13.5

[Amanda Hsieh]

Hi, 

Ketil is out of the office, so I take this case.

It is not clear if you have compensated for the PCB loss in these measurements. Can you clarify?

1. For the nRF7002-DK, we have typically measured ~1.5dB PCB loss in the 2.4GHz band and ~3dB PCB loss in the 5G band.

2. Note that all the power numbers in the Product Specifications are at the nRF70 chip RF pin and not at the PCB antenna port - Just FYI (and hence need proper compensation for PCB loss and any other connector, cables losses, etc) to get accurate results.

3. Assuming the TX power numbers they are quoting at the antenna connector and not compensated for PCB loss, adding the nRF7002DK PCB loss, brings them close to the PS numbers.

-Amanda H.

[JoeFu]

Hi Amanda,

Thanks for your feedback. In the above test, I have already calibrated the RF cable loss and did not add any extra PCB loss. The PCB loss you mentioned is unbelievably large. I don't think any product would use an IC chip pin directly without an antenna trace. Is there an RF test report for the nRF 7002-DK measured at the PCB antenna port or in-line switch? I need you share your measurement environment and testing methods then I can compare it with my data to see if they are consistent.

Thanks.

[Amanda Hsieh]

See nRF7002 Errata [8] RADIO: TX output power deviates from specified value. Try to get the IC sample(after trace code 2333) for testing.

[JoeFu]

Hi Amanda,

In this test, I follow the power table from datasheet below. I use one EVB with different trace code nRF7002 chip for testing. The test results indicate the need for a calibration mechanism to adjust power levels. If the specification is ±2dB, there seems to be a significant margin at 2.4GHz. Without calibration, how can we resolve the issue of unbalanced settings and measurements?

Thanks.

nRF7002-DK	2.4GHz	Setting	EVB N7002 2252AA Avg. power (dBm)	EVB N7002 2518AA Avg. power (dBm)	5GHz	Setting	EVB N7002 2252AA Avg. power (dBm)	EVB N7002 2518AA Avg. power (dBm)
Transmitter characteristics	1M	20	18.50	18.32	VHT20_M0	13	12.74	13.32
	11M	20	18.58	18.41	VHT20_M7	13	12.83	13.89
	6M	15	13.67	13.49	HE20_M0	13	13.14	14.12
	54M	15	13.37	13.43	HE20_M7	13	13.26	14.23
	HT20_M0	15	13.53	13.34
	HT20_M7	15	13.22	13.3

[Amanda Hsieh]

JoeFu said:

I use one EVB with different trace code nRF7002 chip for testing.

May I know the trace code?