The nRF7002 datasheet lists that the current draw for maximum power transmits is ~170 - 260mA depending on frequency and if you are using the CSP or QFN package. Please confirm that this current is drawn from the VBAT net and that it includes the complete power draw of the unit, not just the transmit sub-section.
Please also confirm the maximum current draw of IOVDD.
Similarly, this link says that the wifi transmission power has NO control?
Is it true that we will have no control over the transmit power of the device? Especially given that this website shows that in 'DTIM' mode the maximum power draw is much lower than in 'TWT' mode.
https://devzone.nordicsemi.com/power/w/opp/14/online-power-profiler-for-wi-fi
The current draw for tx is only the current for the amplifier to transmit and not the rest of the chip. It is the power drawn on VBAT for Tx only and doesn't include the rest of the circuitry.
IOVDD current is not specified because it is very low, in the noise relative to the Wi-Fi work. IOVDD only supplies the SPI port, the interrupt line and coexistence pins all of which should be generally high impedance so I would estimate current on average to be very low, below 1mA by a bunch.
The page you reference is about Matter over Wi-Fi and appears to pertain to that context. You can specify maximum Tx power in the devicetree. You can use the nRF7002-EBII as a reference.
&nordic_expansion_spi {
status = "okay";
nrf70: nrf7002-spi@0 {
compatible = "nordic,nrf7002-spi";
status = "okay";
reg = <0>;
spi-max-frequency = <DT_FREQ_M(8)>; /* 8 MHz */
/* Maximum TX power limits for 2.4 GHz (dBm) */
wifi-max-tx-pwr-2g-dsss = <21>;
wifi-max-tx-pwr-2g-mcs0 = <16>;
wifi-max-tx-pwr-2g-mcs7 = <16>;
/* Maximum TX power limits for 5 GHz (dBm) */
wifi-max-tx-pwr-5g-low-mcs0 = <13>;
wifi-max-tx-pwr-5g-low-mcs7 = <13>;
wifi-max-tx-pwr-5g-mid-mcs0 = <13>;
wifi-max-tx-pwr-5g-mid-mcs7 = <13>;
wifi-max-tx-pwr-5g-high-mcs0 = <12>;
wifi-max-tx-pwr-5g-high-mcs7 = <12>;
iovdd-ctrl-gpios = <&nordic_expansion_header 5 GPIO_ACTIVE_HIGH>;
bucken-gpios = <&nordic_expansion_header 4 GPIO_ACTIVE_HIGH>;
host-irq-gpios = <&nordic_expansion_header 14 GPIO_ACTIVE_HIGH>;
wlan0: wlan0 {
compatible = "nordic,wlan";
};
};
};
TWT is substantially lower power than DTIM. You can sleep for days at low uA. With DTIM, you need to ping the router constantly, at a minimum once per second. The difference can be many orders of magnitude lower with TWT.
Hello,
Lowering the output power will reduce the current consumption, you can see this for instance by adjusting the TX power steps in the online power profiler for a TX beacon (though the online power profiler only have 3 steps):
https://devzone.nordicsemi.com/power/w/opp/14/online-power-profiler-for-wi-fi
The nominal voltage is 3.6V, using 4.2 is fine, since it's lower than the maximum 4.5:
https://docs.nordicsemi.com/r/bundle/ps_nrf7002/page/chapters/recommended_op_conditions/doc/recommended_op_conditions.html
You can find expected variation in output power (in dB) as function of supply voltage and temperature here:
https://docs.nordicsemi.com/r/bundle/ps_nrf7002/page/chapters/elspec/doc/electrical_specification.html
Hope that helps,
Kenneth
Thank you. these are helpful.
One more follow-up question: The power limits in that code snippet you sent (and in our code that we used) are stated to be in dBm in the note. However we find that setting these numbers even as low as 0 still has the unit drawing ~140mA on the EVK. Is that variable actually in dBm, or does it have an offset from actually outputting 0 power?
Yes, it's dBm with no offset, so 0 means a 0 dBm ceiling. Current doesn't scale down toward zero with output power, because much of the TX current is fixed.
If you are interested in peak current drawn by the nRF7002, the best is to look at the online power profiler I linked. For instance configuring TX beacon in the online power profiler the peak current drawn by nRF7002 @2.4GHz is: 0dBm ~120mA, 6dBm ~124mA and 16dBm ~162mA. Same numbers @5GHz seems to be about ~70mA higher peak.
Kenneth
Hello Kenneth,
I had looked at the profiler and it was some helpful. My concern was that it may have some sort of nuance I wasn't understanding since 16 dBm is over the listed maximum power for MCS0 and MCS7 mode but below max power for the DSSS mode. So it was unclear why 16dBm was being used as a characterization point and default in the code for the 'MCSx' modes.
So to make sure I set up our EVK with a current monitor on the 7002 (only the 7002, not the whole EVK) and set it to transmit. I got 182mA peaks at 16, which is very close to the datasheet prescribed typical 187 for the QFN in MCS0 mode, even though the code said 16dBm instead of the datasheet's maximum of 15dBm. It is also 12% higher than the profiler's output at that power level. So I tested at '0dBm' and I got 140mA which is 13% higher than the profiler's output of 124mA of the profiler.
So I just wanted to clarify to make sure we are on the same page:
The datasheet does not appear to have any data about the unit to unit variance in power draw for a given power setting. The profiler is giving readings 12-13% below my experimental readings. Should I expect a +/- 10% or greater variance in current draw from unit to unit for a given power setting?
Hi Samuel,
A few thoughts.
A 10-15% spread across units under same conditions is typical for process variations and RF power amplifiers. If you factor in temperature range and supply voltage variations, I expect the spread can even be higher.
Other reasons why your measurements might be 12-13% higher than the Online Power Profiler:
"Peak" vs. average TX burst: The datasheet and profiler numbers represent the average current across the active TX sequence, not the instantaneous peak. If your measurement tool is logging the absolute "max" spike, it's likely catching DC-DC switching ripple or brief ramp-up transients rather than the actual TX level.
Voltage drop (IR drop) at the chip: At 100+ mA, test wires and contacts can cause voltage drop. Because the nRF7002 uses internal DC-DC converters, if the voltage sags at the chip input, it has to draw more current. It's worth scoping the supply rail directly at the nRF7002 VDD/VBAT pin during transmission to check if the voltage is dipping.
Test lead inductance: Fast current steps through longer jumper leads introduce inductive overshoot, which measurement tools can misinterpret as higher peak current.
The devicetree property wifi-max-tx-pwr-2g-mcs0 = <16>; only sets the software ceiling limit. The Wi-Fi firmware and internal factory calibration will clamp the actual output power to the hardware limit (15 dBm typical for QFN MCS0).
Best regards,
Kenneth
Hi Samuel,
A few thoughts.
A 10-15% spread across units under same conditions is typical for process variations and RF power amplifiers. If you factor in temperature range and supply voltage variations, I expect the spread can even be higher.
Other reasons why your measurements might be 12-13% higher than the Online Power Profiler:
"Peak" vs. average TX burst: The datasheet and profiler numbers represent the average current across the active TX sequence, not the instantaneous peak. If your measurement tool is logging the absolute "max" spike, it's likely catching DC-DC switching ripple or brief ramp-up transients rather than the actual TX level.
Voltage drop (IR drop) at the chip: At 100+ mA, test wires and contacts can cause voltage drop. Because the nRF7002 uses internal DC-DC converters, if the voltage sags at the chip input, it has to draw more current. It's worth scoping the supply rail directly at the nRF7002 VDD/VBAT pin during transmission to check if the voltage is dipping.
Test lead inductance: Fast current steps through longer jumper leads introduce inductive overshoot, which measurement tools can misinterpret as higher peak current.
The devicetree property wifi-max-tx-pwr-2g-mcs0 = <16>; only sets the software ceiling limit. The Wi-Fi firmware and internal factory calibration will clamp the actual output power to the hardware limit (15 dBm typical for QFN MCS0).
Best regards,
Kenneth