Use NPN transistor to switch NRF24L01. Is this OK or will it fry?

Experimenting with a low-power temp/humidity wireless sensor.

Brief H/W description: ATMEGA328P chip (bare, int.8MHz clock) powered by LiPo cell. DHT22 with Vcc connected to Arduino Pin. NRF24L01+ powered by a 3.3V LDO regulator (BA33BC0).

Brief code description: DHT on, sleep 1s, read temp/humidity from DHT, DHT off, radio on, transmit data, radio off, sleep 7s.

I want to switch the Vregulator and NRF24 off while sleeping to save some power.

All I have in my parts pile are some NPN transistors. I can make a low-side switch, with the Vreg and NRF24 grounded to the NPN's collector. When the transistor is on that should bring the collector to ground and allow current to flow.

When the transistor is off, though, I think the NRF24 ground will be at Vcc and I worry that current can flow through the NRF24 via the signal pins and zorch something.

Can anyone offer advice on using NPN transistors for switching in this situation?

When the transistor is off, though, I think the NRF24 ground will be at Vcc and I worry that current can flow through the NRF24 via the signal pins

Yes it will.

The way round this is to make all outputs to the NRF24 into inputs before you turn it off with the FET.

Most people find it convenient to use a PNP transistor to control power to small modules, because it has several advantages over your approach. Google "high side switch".

But there is no need to do that at all, because you can put the radio module to sleep, in which case the current draw is negligible (~900 nA).

Thanks for your responses! I agree the NPN is not ideal, I chose it because it's what I had in my parts pile.

But there is no need to do that at all, because you can put the radio module to sleep, in which case the current draw is negligible (~900 nA).

I measured 400uA @ 5V, after putting the radio to sleep, DHT off and the MCU in powerdown. Perhaps the library I use (RadioHead) radio.sleep() puts it in standby (26uA), not full power-down (0.9uA). I think the Vreg is wasting the majority of that, which is why I want to power-off the Vreg AND the radio.

I suppose the LDO Vreg is not the ideal way to power the radio but I had them handy and my electronics engineering fu is too weak to come up with something better for a LiPo powered project.

I ended up getting some P-MOSFETs instead, thinking they would also be useful to have around as reverse-polarity protectors for other projects (with projects using NRF24 radios and LiPo cells, the forward-voltage drop of diodes for RPP is a significant disadvantage). They did the job very well. I can no longer measure the sleeping current with my existing equipment.

(I should get a uCurrent from EEVBlog if I keep doing these kinds of projects)

You can also get low-dropout regulators that have an enable or shutdown pin, and do away with the extra switching transistor all together.

But you still need to shut down any outputs to the chip as I said in reply #1