0
\$\begingroup\$

I'm building a battery-based wireless weather monitor for hobby project.

for longer battery life, I want to control the power of each subsystem separately.

Let's take a NRF24 modem for example: Previously, I have configured a separate Power regulator for modem. regulator had an EN pin, which could be switched on or off by changing pin level. (High-side control) this method was easy and worked well.

now, I wanted to control power the module using MOSFET instead of EN pin of regulator. I thought of using N-channel MOSFET to control the power. like this:

(Note: DMG2302UK was N-channel MOSFET)

Scheme

so, My question is:

  1. Is this a good method? I've tried running fans and heaters in this way, but I don't know what's going to happen to modules with ICs.
  2. it seems that there is a pull-up resistor inside the module, but can this cause harmful operation?
  3. I don't have much experience in electronics. is there any other problem I didn't expect in this regard?
\$\endgroup\$

3 Answers 3

3
\$\begingroup\$

I'm not sure that it's a good idea to switch the ground side of the power supply. Doing so just leaves the module floating. Far better to devise a way to switch the 3.3V supply.
In this article it states

"And the best part is that it consumes only 26 µA in standby mode and 900 nA in power down mode. That’s why it’s the go-to wireless device for low-power applications."

Is this a low enough? Can you find the commands in the API to do this? The advantage would be that you could then put the controlling microprocessor into sleep mode as well as it wouldn't have to maintain the drive signal to the MOSFET. In fact you wouldn't need the MOSFET and its associated components at all.

\$\endgroup\$
5
  • \$\begingroup\$ Thanks for the detailed answer! My module is equipped with LNA, it seems to require little more current even if the NRF24L01 chip is off... one additional question, what can happen if module floating with GND blocking? \$\endgroup\$
    – Yukihyo
    May 21, 2023 at 23:20
  • \$\begingroup\$ @Yukihyo Unpredictableness since your circuit which is supposed to be OFF, now doesn't "know" what OFF potential is. It still has all the connections to other powered equipment and those introduce voltages to your device which now have no reference state. So inside your device you might get weird half powered portions while others might be powered but floating, etc. There are sneak paths all over the place such as to the power rails from ESD diodes. \$\endgroup\$
    – DKNguyen
    May 21, 2023 at 23:37
  • \$\begingroup\$ @Yukihyo You also get similar issues even if you disconnect the highside but in that case it's a bit more predictable. Current from signals gets injected into the power rails through the ESD diodes. \$\endgroup\$
    – DKNguyen
    May 21, 2023 at 23:44
  • \$\begingroup\$ @DKNguyen Thank you for your detailed reply. yes, my board is equipped with a ESD protecton diodes. sometimes I have experienced measure data becoming abnormal when control with GND pins as a power source. now I know the cause of this why. \$\endgroup\$
    – Yukihyo
    May 22, 2023 at 0:11
  • \$\begingroup\$ @Yukihyo ICs have these diodes internally. \$\endgroup\$
    – DKNguyen
    May 22, 2023 at 1:01
0
\$\begingroup\$

One needs to be careful when adding a switch in the ground path. In the functional mode when the switch is ON, there will be an IR drop across the switch. If the ground current is noisy, it will manifest as ground noise due to the switch. You need to analyze if this can degrade performance of your IC.

Regarding the case where the switch is OFF, the state of the input and output pins of the IC is indeterminate. So, you'll need to ensure that this does not cause additional leakage and/or functionality failure of your system.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you for the detailed explanation! You gave me a clear answer to what I was most curious about. \$\endgroup\$
    – Yukihyo
    May 22, 2023 at 12:06
0
\$\begingroup\$

Likely your module will turn off this way, but if its outputs are connected back to the MCU, they will tend to be seen by the MCU as a 'high'; this might confuse the MCU. If the MCU tries to drive any of the pins to (its) low (== GND), this will provide a leakage path and leave the module on in a strange configuration.

The system will be clearer if you use a high-side PMOS switch. You also don't need R106 in the above circuit.

\$\endgroup\$
2
  • \$\begingroup\$ Thank you for your detailed answer! I think I forgot the importance of the GND... and, I added a R106 for protect the GPIO pin. Is it okay to omit this? I'm not using a FET driver IC. \$\endgroup\$
    – Yukihyo
    May 22, 2023 at 12:04
  • \$\begingroup\$ As-drawn, with a common GND there is no need for protection - there is no path to damage anything \$\endgroup\$
    – jp314
    May 23, 2023 at 12:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.