Hello, I am wondering how I can power my ESP32, which has a sleep current of about 10μA with a battery. I am asking because I can't seem to find a high efficiency buck converter for extremely low currents. I need to have a stable 3.3v for the ESP32. Should I use 2 separate buck converters, one for normal operation and one for the sleep mode, or a single IC (which I can't find)? I am trying to achieve the highest efficiency possible. 
Thanks!
How about an LDO linear regulator for your sleep mode? Still takes 10 µA out of your battery but will work fine.
The big problem is going to be to switch between the two power supplies - how do you plan to achieve that?
What battery are you using?
If I am using 2 systems, I can use a GPIO which will be pulled high during operation to shut down the LDO and enable the buck converter. When the ESP32 goes to sleep, I can pull the GPIO low to shut down the buck converter and enable the LDO.
edit: I am using a 200mAh LiPo battery
Depending on your on time, and how much power the ESP draws then, it may not be worth it using a buck converter bringing 3.3-4.2V down to 3.3V. There are LDO regulators that have a minimum drop just of 100-200 mV and very low quiescent current. Probably a better bet.
epicface2304:
edit: I am using a 200mAh LiPo battery
Even if the LiPo is constantly as high as 4.2V the linear regulator has an 'effciency' of around 80%.
With such a low differance between input and output the switcher could, overall, be less efficient.
I would stick with a decent LDO.
does the SC662K-3.3V look like a good option? I'm wondering if it's possible to somehow use the the 0.5 volts left.
Here is the discharge curve:
No idea, you would have to test it.
The ESP32 is very fussy about the supply and regulator when the WiFi is in use. An inadequate supply will cause it to brown out when the WiFi turns on.
The worst supply I have found for this effect is low capacity lithium batteries at less than full charge.
Thankfully, I don't plan on using WiFi, just Bluetooth.
srnet:
The ESP32 is very fussy about the supply and regulator when the WiFi is in use. An inadequate supply will cause it to brown out when the WiFi turns on.The worst supply I have found for this effect is low capacity lithium batteries at less than full charge.
I use at least 470uF worth of tantalum capacitance on the VCC pin of the ESP to fix that problem.
Leo..
Wawa:
I use at least 470uF worth of tantalum capacitance on the VCC pin of the ESP to fix that problem.
Leo..
Its not just a capacitance problem, although adding them does help.
With no additional capacitance apart from the ceramics on the input and output of the regulator I get these results with different supplies;
3 x AA Alkaline - OK.
4 x AA Alkaline - OK (the intended supply for the board)
5V USB power bank - OK.
4 x AA NiZn rechargeable - OK.
Lithium Ion 2000mAhr Freshly charged - OK, but just.
Lithium Ion 2000mAhr 1/2 discharged - Resets
And in the case of the situation where there were resets with the Lithium Ion, adding low ESR electrolytics did not cure the problem.
Regulator was a MCP1700.
Did you connect a scope to the regulator input and ESP VCC, and see how much each/both dip.
Maybe you should try a different regulator, if only the output dips.
Leo..
Wawa:
Did you connect a scope to the regulator input and ESP VCC, and see how much each/both dip.
Maybe you should try a different regulator, if only the output dips.
Leo..
Yes I did check with a scope, and several regulators.
One day I will write it up.
You can see the progress of the project here;
