Here's a quick and easy fix for the turning off at 40% battery issue that people are seeing after running Linux. This is effectively the same fix as unplugging the battery, but does so without having to take apart the Switch.

Step 1: Download attached fusee-linux-battery-fix.rar
Step 2: Extract and load the binary with your favorite RCM Smash tool (I'm using TegraRcmSmash).
Step 3: You should see the typical fusee screen, at this point unplug the USB cable. The screen should go off.
Step 4 (optional?): Hold the power button for 10 seconds. Your switch won't turn on at this point, that's desired.
Step 5: Plug in a USB cable. Switch will turn on, and you're done.

This worked to fix my Switch, let me know if it works for you.

Edit:

Source is here:

github/crystalseedgba/BatteryFix/blob/master/fusee/src/main.c (Sorry it's not letting me post hyperlinks)

This doesn't fix the Linux bug, this is really just a crude workaround. At first I suspected that Linux is somehow mangling the battery gauge but this doesn't seem to be the case. The battery gauge reports around 40% and decent battery voltage when my Switch would turn off. I think Linux is misconfiguring some undervoltage lockout somewhere, but I don't know where.

Thanks to jjbredesen for his "How to make a custom Fusee payloads" tutorial.

 

[Notex]

What exactly does this do to fix the battery? Did you code this or did you find it somewhere? I'm kinda anxious about just running a random binary on my switch.

 

[x0x0]

Lol 1 post guy, dont use it...
EDIT: THIS IS WORKING THANK YOU OP

 

[Soluble]

Yeah I don't have the issue, but this is a good time to remind everyone NOT to use random payloads we know nothing about.

 

[Benoit934]

Source code ?
I mean can you please give us the source code ?

 

[LeRepex]

just tried it out for me it seems to work. having my switch playing some games now to waste some battery and see if it turns off

 

[Notex]

just tried it out for me it seems to work. having my switch playing some games now to waste some battery and see if it turns off

You are brave. I guess you are our guinea pig

 

[Clydefrosch]

or he's another sock puppet trying to sucker you into bricking your switches

 

[LeRepex]

You are brave. I guess you are our guinea pig

it was more just "yes finaly i can fix my switch than thinking about the risks "

also i can say my switch is running super mario odysee right now with 35% without turning off so it seems to have fixed it before the limit was about 40%

--------------------- MERGED ---------------------------

or he's another sock puppet trying to sucker you into bricking your switches

If you want i can post pics

 

[Notex]

also i can say my switch is running super mario odysee right now with 35% without turning off so it seems to have fixed it before the limit was about 40%

That's great news. I am going to wait until a few more people try it first though before I try. It just seems very sketchy that someone new to the forum has posted this fix.

 

[LeRepex]

That's great news. I am going to wait until a few more people try it first though before I try. It just seems very sketchy that someone new to the forum has posted this fix.

Thought the same at first but only the chance to fix my switch without taking it apart was enough for me to try it

 

[Benoit934]

If you want i can post pics

Even a video capture wouldn't be enough because you can run a dummy payload letting the think it's working but actually sending a bad payload.
I don't know why someone might do this but for avoiding the paranoia around here the best way is to have the source code that we compile ourselves after checking after what it's doing.
They fear TX too so ...

 

[LeRepex]

Even a video capture wouldn't be enough because you can run a dummy payload letting the think it's working but actually sending a bad payload.
I don't know why someone might do this but for avoiding the paranoia around here the best way is to have the source code that we compile ourselves after checking after what it's doing.
They fear TX too so ...

Yes thats actually a good point but still if you want i can post pics maybe also a video with download process

EDIT: 30% and still running fine

 

[Benoit934]

@LeRepex I beleive you and I'll not blame the newcomer for being a newcomer but it would be great if he share the source code, it's a fix there is no reasons to not share it

 

[rajkosto]

Source code posting along with the binaries is actually REQUIRED since this is based on the fusee payload code base (GPLv2 licence)... don't run this until someone else has compiled it and verified the binaries are similar.

 

[LeRepex]

@LeRepex I beleive you and I'll not blame the newcomer for being a newcomer but it would be great if he share the source code, it's a fix there is no reasons to not share it

Yes thats actually true

EDIT: 20% and still running Mario Odyssey just fine
EDIT 2: 10% Still running i will now charge my switch up again. who ever made this fix i love you

 

[xXxSwagnemitexXx]

doubtful, ill probably try it though

 

[rajkosto]

  init_i2c_controller(0);
  i2c_read_register((int)v2, 1u, 0, 0x6B, 0xA);
  if ( v2[0] != 0x2F )
  {
    printk("ERROR: charger part number mismatch got: 0x%02x want: 0x%02x\n\n   ", v2[0], 0x2F);
    while ( 1 )
      ;
  }
  i2c_read_register((int)&v2[1], 1u, 0, 0x6B, 5);
  v2[1] &= 0xCFu;
  i2c_write_register(0, 0x6B, 5, &v2[1], 1);
  i2c_read_register((int)&v2[2], 1u, 0, 0x6B, 7);
  v2[2] |= 0x20u;
  i2c_write_register(0, 107, 7, &v2[2], 1);
  printk("\n\nBATFET is now disabled, disconnect the USB cable.\n\n");

reverse engineered code, so doesnt seem too malicious.

 

[x0x0]

doubtful, ill probably try it though

GL post results.

  init_i2c_controller(0);
  i2c_read_register((int)v2, 1u, 0, 0x6B, 0xA);
  if ( v2[0] != 0x2F )
  {
    printk("ERROR: charger part number mismatch got: 0x%02x want: 0x%02x\n\n   ", v2[0], 0x2F);
    while ( 1 )
      ;
  }
  i2c_read_register((int)&v2[1], 1u, 0, 0x6B, 5);
  v2[1] &= 0xCFu;
  i2c_write_register(0, 0x6B, 5, &v2[1], 1);
  i2c_read_register((int)&v2[2], 1u, 0, 0x6B, 7);
  v2[2] |= 0x20u;
  i2c_write_register(0, 107, 7, &v2[2], 1);
  printk("\n\nBATFET is now disabled, disconnect the USB cable.\n\n");

reverse engineered code, so doesnt seem too malicious.

I will be really surprised if this is a real fix

 

[dark_samus3]

EDIT: looks like it was already posted above :|