0
\$\begingroup\$

Based on this question I created the following circuit:

tp4056_mt3608_ups

Since I made some (component) changes, first I would like to have a validation of the circuit for long-term usage (and some improvements).

Used components (different from linked circuit):

  • 1N5822 diode
  • AO3401 P-channel mosfet
  • 3.7 V 2600 mAh 9.62 Wh (protected) Li-Po battery (from another project)
  • Added a ceramic capacitor and a toggle switch (I hope to the right position)

I made this circuit as a battery backup solution, so 99.9% of the time it will be powered externally, but I don't want the Pi to shut down when there is a power outage.

Here comes my question #1: do Li-Po or Li-Ion batteries last a long time if I connect them to a TP4056 charger constantly?

Currently I'm testing with a Pi 3b+, but I want to upgrade to a Pi 4 later. The circuit works fine when there's external power, but I have issues when I switch to battery powered mode; during boot, the Pi shuts down. If I'm right, the battery protection in the charger module kicks in and saves the battery from over-discharge. When this happens, I get no voltage from the out pins, so I made this modification:

enter image description here

So I connected the battery directly to the MOSFET, because the battery itself has a protection circuit. With this modification I almost get to the desktop, but just before that I get a reboot. If I turn off the external power when the Pi is in idle state, the circuit works until I open an app and then it reboots.

I don't know the C rating of my battery unfortunately; I got it from China. But I have a linear bench power supply.

Question #2: Is it safe to connect it to the TP4056 as a battery replacement to test how much current the boost converter needs in battery mode?

Question #3: I'm thinking about buying a reliable 18650 Li-Ion battery with a high current rate. Do you think it will solve my problem?


As an alternative solution I just connected another charger module in parallel like this (with a second battery), but I saw weird things happen, like they randomly showed the 'charging-done' state when they weren't full. After reconnecting the external power they started charging again. So I would like to stick to one charger/battery if I can.


My question #4: is there a DC/DC boost converter that has a higher efficiency than the MT3608 when using batteries? Also, I don't really need the adjustable output, and actually, that potentiometer looks like a cheap component which is not reliable at all. I just need a stable 5 V output.

Thank you for your help and answers!

\$\endgroup\$
2
  • \$\begingroup\$ Question 4 is off-topic because it asks about component recommendations. Consider asking separate, narrowly focused questions to get better results. \$\endgroup\$
    – JYelton
    Feb 13, 2020 at 23:17
  • 1
    \$\begingroup\$ MT3608 seems a little underpowered for that application electronics.stackexchange.com/questions/445948/… \$\endgroup\$
    – Jasen
    Feb 14, 2020 at 0:28

1 Answer 1

0
\$\begingroup\$

This situation depends on the functionality of the "OUT" pin of the charging module. The following is referencing your original schematic. Your second schematic is invalid, as the charging circuit is expecting the storage cell to be isolated:

If the "OUT" is always active, then you don't need the MOSFET or the diode. If it only switches on whenever the input power is no longer present, then you just eliminate the MOSFET(unless the "OUT" pin doesn't have some sort of PN junction on it's output already, which it likely does) or replace it with another diode.

In either case, the charging circuit you are currently using is only specified to output 1 Amp of current, which is WAY too little to power a RaspberryPi.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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