Electronics > Projects, Designs, and Technical Stuff
Low power, battery operated design with dual linear regulators
(1/4) > >>
Srijal97:
Hello,

I'm looking to build a project around the ESP32 module. It would be powered from a single LiIon cell and I want the ESP32 to be in a deep sleep state when the project is turned off so that the inbuilt RTC on the ESP32 remains active and maintains time. I plan to use 2 linear voltage regulators: First the TL1963A - 3.3V, 1.5A for powering the ESP32 + peripherals (Display, IMU, camera, etc) under normal operation and a second TPS797 - 3.3V, 50mA with a very low quiescent current (1uA) to power just the ESP when project is turned off.

Here's the power switching flow I have in mind. Please refer schematic attached with the post.
1- Everything is on and the 1.5A regulator (TL1963A) is active.
2- User switches the power switch off, which is detected by an ESP32 GPIO.
3- ESP32 performs the necessary shutdown routine (save some data to memory), shuts down power to all peripherals (using the FET), shuts down wifi and other internal hardware blocks (WiFi, Bluetooth, etc).
4- ESP32 now draws less than 50mA of current so turns off the 1.5A regulator (TL1963A) and the smaller 50mA regulator (TPS797) is used. The Schottky diode will create a 0.3V drop so ESP32 Vdd will drop from 3.3V to 3V (as per my understanding). Will this voltage sag create a problem?
5- The ESP32 now enters a deep sleep state ("Hibernation mode" as per datasheet) and draws about 5-10uA with the RTC and certain GPIOs active for waking the device back up when the power switch is turned ON again.

I need some comments on whether something like this would work reliably? I am expecting a minimum battery life of 6 months to a year with a single 18650 cell.  How can I improve the design to get the same functionality?

Thank you very much for your time!
ledtester:
How about using LiFePO4 batteries and ditch the regulators?

https://youtu.be/heD1zw3bMhw?t=4m35s

(Video was made in 2016 which is why Andreas says they are not readily available.)
ledtester:
Another idea...

How about using the TL1963A only for power the peripherals and the TPS797 only for the ESP32?

The ESP32 enables the TL1963A only when it is fully awake. The two 3.3V rails don't have to be connected.


Srijal97:

--- Quote from: ledtester on September 15, 2019, 02:28:58 pm ---How about using LiFePO4 batteries and ditch the regulators?

https://youtu.be/heD1zw3bMhw?t=4m35s

(Video was made in 2016 which is why Andreas says they are not readily available.)

--- End quote ---

I saw that solution somewhere but the device I am making would have several peripherals in the future, all of which won't be able to run at a voltage lower than 3.3V, as the ESP. I need a stable 3.3V rail. Perhaps I can have a step up circuitry for them? But I don't like the idea of doing that.
Srijal97:

--- Quote from: ledtester on September 15, 2019, 02:44:23 pm ---Another idea...

How about using the TL1963A only for power the peripherals and the TPS797 only for the ESP32?

The ESP32 enables the TL1963A only when it is fully awake. The two 3.3V rails don't have to be connected.

--- End quote ---

The TPS797 won't be able to sustain the ESP32 at all times. The ESP32 can draw up to 500mA with spikes up to 800mA during operation of WiFi and Bluetooth blocks. If I have a separate regulator of that rating for the ESP alone, I lose out on quiescent current.

Also, a friend pointed out that the Schottky diode won't operate at such low uAs of current, so I need a different hardware circuitry.
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod