I am designing a circuit for low power Iot device and trying to get as much as possible from the battery (Lipo).

I placed bq25170 as battery charger and tps7a20(3.3v) as LDO. I would need to power a nrf52840 module.

Now about switching between USB (when connected) and Bat before the LDO, most solutions use a p-mosfet. I think I could use LM66200 or not?

Are there better solutions to do that? My first concern here is saving power and pull as low as possible from the battery when no usb is connected. I could change the LDO if you have better alternative which leads to save power.

EDIT: seems that I was not able to describe the question correctly most of the solutions are using something like this: enter image description here My question is there a better way to achieve same functionality but more energy efficient than using a transistor with a diode? I thought of putting LM66200 and connect the battery and the VBUS as its two inputs.

  • \$\begingroup\$ "pull as low as possible from the battery when no usb is connected" - when USB is not connected, isn't that when power has to come from the battery? Please draw a circuit diagram (there's a built-in editor). \$\endgroup\$
    – Sim Son
    Commented Jan 18 at 20:07
  • \$\begingroup\$ Sorry for the bad describtion language. I modified it now \$\endgroup\$
    – user130227
    Commented Jan 18 at 22:05

1 Answer 1


When you look at the data sheet of your NRF, you'll see that universally, the power consumed is lower when the supply voltage is 1.8 V than when it's 3.3 V.

So, you'll want a high-efficiency (especially in low-load) switch-mode power supply that brings down your battery voltage to 1.8 V.

Your supply switching question makes little sense to me, sorry – it seems like you want to control something with a circuit that at that point has no power, and that can't work.

  • \$\begingroup\$ Thanks for this point about the nrf chip. seems to be really important. I will check it carefully. So you think switch-mode would not affect the stability of the nrf module. Most of the ble modules use LDO. Is this a point that they consider or it is something else ? about the supply switching question I think I was not able to describe it correctly I modified the question now. Hope this is better now. Sorry \$\endgroup\$
    – user130227
    Commented Jan 18 at 22:02
  • \$\begingroup\$ I don't see any problems with a modern switch-mode supply. It's what basically everyone uses to drive their devices off a lithium battery, and for the actual internal voltages, the NRF chip contains on-chip LDOs, so having an external LDO really has no advantage. \$\endgroup\$ Commented Jan 18 at 22:06
  • \$\begingroup\$ Great info. I looked now to the Power Profiler of the nrf and calculated the wattage in both cases. you are totally right. Do you have a suggestion for a switching mode supply unit? \$\endgroup\$
    – user130227
    Commented Jan 18 at 23:50
  • \$\begingroup\$ @user130227 I'd probably just note down the maximum current you'll draw, multiply it with a safety factor of 1.5, go to ti.com, select products->power management->DC/DC switching regulators->Step down (buck)->"View all products" and filter by what I need (so, input between 3.7 V and the maximum charge voltage, output 1.8 V), sort by quiescent current, pick one that is cheap and scroll down to see whether I can actually buy it. \$\endgroup\$ Commented Jan 19 at 8:43
  • \$\begingroup\$ so, something like TPS62A01DRLR might possibly work. Of course, TLV627432 is even lower in quiescent current, but increases BOM cost and might be harder to integrate into your assembly chain. \$\endgroup\$ Commented Jan 19 at 8:51

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.