2
\$\begingroup\$

I have an old phone (nexus 5) that I want to use for a project. I need to replace the battery with a permanent power supply, I've found the 3 pins of the battery and made a breakout connection with 1/10" headers instead of the stock battery connector. The phone will boot fine with a battery connected this way. The phone runs at 4.2v, ~500±100mA.

The issue is the phone won't boot on my bench PSU. Recovery mode works but Android powers off midway through boot. I suspected it was the lack of a thermistor in the third battery pin that made it shut down. I measured the resistance and added a fitting resistor between the battery monitor pin and ground (56k). However, the phone wont boot either.

I also suspected a power spike that my power supply could not handle could be the culprit. But after adding a 4700μF cap the issue persists.

I have anoter nexus 5 motherboard that behaves just the same.

What else can be causing the phone to act differently with the power supply and the battery?

Ps: I have access to te UART port of the phone, the kernel does not print anything relevant and the phone shuts down at random boot events, sometimes glitching the line, so I think this is hardware or at least isolated from the main processor.

\$\endgroup\$
7
  • \$\begingroup\$ leave the battery in and measure V on each pin near 3.7V \$\endgroup\$ May 22, 2021 at 13:31
  • \$\begingroup\$ Why not just leave the battery alone and power over USB? I used a Nexus 5 as a time lapse camera before and it'll run forever through USB power. \$\endgroup\$ May 22, 2021 at 16:04
  • \$\begingroup\$ @Tony does this mean i need a series resistor with my battery? \$\endgroup\$
    – edmeme
    May 22, 2021 at 16:36
  • \$\begingroup\$ @user1850479 None of the boards will boot from usb if the battery isnt on. Does yours work without battery? \$\endgroup\$
    – edmeme
    May 22, 2021 at 16:37
  • \$\begingroup\$ See android.stackexchange.com/questions/213722/… and android.stackexchange.com/questions/26123/… for debugging boot loop issues. You can use adb or fastboot or some recovery versions to see the logs and that may point you at possible battery error messages. \$\endgroup\$
    – Passerby
    May 22, 2021 at 17:42

2 Answers 2

3
\$\begingroup\$

Not on nexus 5, but I've done this on several phones. I highly doubt the 500mA+100mA figure for peak usage is close to real power draw. From what I've googled, it is probably a tad north of 2A depending on what phone you have. A bare minimum then would be more like 3A to give some headroom. The strange part is, an ordinary bench PSU should provide more than plenty current to cover that.

The BMS circuit on the battery does a few things like overvoltage/temperature (high and low) protection and temperatature probing. With my phones, I have done with and without the BMS, and it powers up and runs all the same regardless. Just be in the know that without it, any overvoltage could kill the phone in matter of seconds if not less.

When extracting the BMS, I like to use a "donor" battery from the same phone, to make it easier to fit in the battery compartment. Locate B+ and gnd pins on the BMS, solder a pair of <18awg wires, while making sure the orientation of the wires doesn't impede closing the battery cover and/or causes short circuit with other terminals on the BMS.

I was told android phones have no way of directly measuring the reserve capacity on the battery; the guesswork is done through software algorithm. With a non-factory battery which radically different voltage-current pattern, the battery management component of the OS could get confused and turn the phone off falsely believing that you are running out of charge. In my experience, keeping voltage in 3.9v - 4.2v range helps minimize chances of this happening. I was never able to boot with anything lower than 3.8V, so there could be your problem. Go for 3.9V and see how it goes.

\$\endgroup\$
1
\$\begingroup\$

The battery might have a management IC inside it. This is how some phones can tell if you are using a factory-original or aftermarket battery. The chip might be just a gas gauge, or it could be a complete charge controller.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ There is definitely some circuitry on the battery, however the third pin really looks like a thermistor. Its resistance is strongly correlated with temperature and I can make it swing by 10K just by holding it in my hands. Ill try to snoop the power lines while on battery and see if there's any data going through \$\endgroup\$
    – edmeme
    May 22, 2021 at 16:35

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.