Sudden battery issues

i2c_mux

Hello there,

This is the first time I've seen this issue happening, and have not seen anything related to it on the forum.

My 1ish year old Pixel 8 Pro (running GOS obviously) had very low battery percentage earlier today (2%), and when I plugged it to my usual charger, I noticed that the battery indicator had turned into a question mark instead of the usual percentage with lightning bolt symbol. The issue persisted with a different charger and cable combo despite the indicator returning to normal with nothing unplugged. Disabling USBC restrictions didn't do anything.

With the charger still plugged in, I rebooted the phone and the question mark was even on the BFU lockscreen. Surprisingly, when I unplugged the phone again (still BFU), the battery percentage was now 50% ! Upon unlocking I got a few crash reports for android.hardware.thermal-service.pixel ("ThermalHAL could not be initialized properly"). I looked on the internet, the few forum posts mentioning the battery-indicator-question-mark issue on Pixel phones were sparse and covered various models (Pixel 4–10). Most people said that they either got the battery replaced or that the issue fixed itself after a reboot or safe-mode reboot.

Next I rebooted the phone in safe mode, the issue was still happening. I installed AccuBattery to check some readings. First weird thing was that on first launch the app failed to accurately calibrate the total battery capacity (it guessed 3000mAh). And in the app itself, the readings made no sense (negative current, power consumption and temperature, and constant voltage).

While still in safe mode, I checked the battery diagnosis in system settings, and it had 100% life expectancy, 0 cycles, unknown manufacturing date and first used date (I checked those a few months ago and they were all defined and nonzero). I enabled developer mode and ADB USB debugging to run adb shell dumpsys battery, and the several values clearly made no sense (see attachment at the end).

When the phone is powered down and I plug it to a charger, a battery symbol with a question mark appears as well. When I unplug it, I once got a 4% read and a charging indicator (couldn't reproduce), otherwise it was the question mark inside battery symbol. Same thing with wireless charging.

I never fully charge my phone (80% charging limit enabled) and the battery percentage rarely gets below 20. My phone is not exposed to unusual meteorological conditions and it never had physical damage (few drops and always with a sturdy case on).

I'm keeping my phone off but charging right now because the battery was nearing 0%. Please me if it's a good idea given the messed up stats. I noticed that the clock got set in 12hr mode in-between reboots (don't know if safe mode causes that), see attachments for proof.

So, could this issue have links with GOS ? How to pinpoint whether it's a software or hardware failure ? Has anyone experienced something similar ? How to fix it ?

Screenshots of phone, adb command output and probably useless crash dump (no user data inside) are available here (password is my username).

TLDR my phone suddenly shows a question mark inside battery indicator when charging and other battery-related problems started happening.

de0u

i2c_mux I installed AccuBattery to check some readings. First weird thing was that on first launch the app failed to accurately calibrate the total battery capacity (it guessed 3000mAh). And in the app itself, the readings made no sense (negative current, power consumption and temperature, and constant voltage).

While still in safe mode, I checked the battery diagnosis in system settings, and it had 100% life expectancy, 0 cycles, unknown manufacturing date and first used date (I checked those a few months ago and they were all defined and nonzero).

The combination of the question mark, and the implausible numbers seen by AccuBattery, and the reading of zero charge cycles, strongly suggests some hardware problem. If you're lucky it'll turn out to be just the battery, but maybe not. The device is fairly likely to suddenly get worse in some way.

As MineralWater indicated, charging the battery may not be safe, so copying all data off and turning the device off is probably best.

MineralWater

I think you can rule out that it is firmware related. If it was, many (or even all) phones of the same build model like yours would be affected.

As Graphene currently runs only on Pixels and there was a bug affecting maybe model 9, many other owners of a Pixel 9 would report the problem.

Or when it is something bigger, all model numbers would report battery issues.

What you experience seems to be a hardware defect-either the battery alone or something on the phones "motherboard" has trouble to charge and read the battery.

If you still have warranty with Google, you could flash the stock os again and send it in.

If you have no warranty, you should not use the phone anymore and hope it's a general hardware defect and Google will announce a repair or refund programm.

An other option would be to replace the battery. But as you don't know if it is the battery alone (and that you probably won't get a genuine oem battery) or the hardware around it, it is a lottery.

So I would suggest taking warranty and if there is none, get a new phone.

Defective batteries are a serious danger-they can start to suddenly burn (though in most cases they swell first). Look if your screen starts to pop out on the sides and/or corners (it should be evenly aligned on all sides).

If that happens, don't charge it. Turn it off (erase your data if needed) and throw it away at a recycling point that accepts batteries and electronics.

i2c_mux

Thank you both for your answers. The phone eventually discharged completely and I still can't charge it (a battery symbol with a question mark shows up and blinks forever). There's no visible swelling, but I still leave the phone unplugged now, just in case.

I contacted Google support. They seemed helpful at first but ended up saying they couldn't help me. They stated it was because my IMEI was "inaccessible". I don't know what that means, but it's probably because I never signed up with a Google account on the phone. After this, I added the phone IMEI to a dummy Google account to check the warranty status, it had expired in December. Anyway, I'm now the happy owner of a $500 rare earth and glass fancy brick... Lucky me I had backed everything up. @de0u

@MineralWater On the possibility of replacing the battery, iFixit sells OEM batteries. But as you said, it might be a faulty motherboard or another hardware defect (my battery was fine before the incident, with 100% estimated capacity).

Heck I hoped it was a loose battery connector when I stumbled upon this YouTube video. The guy said that by pressing hard on both sides of the phone, a loose battery connector could be reconnected. Almost all the comments mention that his method worked. It didn't for me unfortunately.

Apparently there was an acknowledged issue with Pixel 4XL devices that forced Google to add a 1 year warranty to these models because of the question mark battery indicator. "Google probably fixed it by now on the next generations", I thought. Well maybe not...

I found several Reddit posts mentioning this issue on recent models (plus the YouTube video above). Here's a non exhaustive list :

There's also an AndroidAuthority article for the Pixel 10 Pro Fold on the topic : https://www.androidauthority.com/google-pixel-10-pro-fold-battery-question-mark-3609650/

As these posts and comments show, this issue does not seem exceptionally rare. Moreover, lots of these Reddit posts are less than 1 year old. Although a majority of the problematic devices seem to be the Pixel 9 Fold and Pixel 10 Fold, there are mentions of non-Fold Pixels, including Pixel 7 and 8 models. Some users reported that it happened right after the warranty expired.

To wrap this up, perhaps the problem I encountered is planned obsolescence, who knows ? I'm disappointed that one of the few phones with 7 years of updates and GrapheneOS compatibility died in less than a year, just a few months after the warranty expired. I hope the future Motorola device will not be subject to such defects.

One last comment for the GrapheneOS development team : in the Battery tab inside the Settings app, there was a message saying "Problem reading the battery meter." in a red rectangle, with a red "Learn more" text. The "Learn more" text/button did not work, nothing happened when I clicked it. I don't know if the scope of this bug is the AOSP or GrapheneOS. A screenshot of the corresponding page in the settings app is available as battery_settings.png here (my username is the password).

de0u

i2c_mux It's no fun to have a phone to fail, especially suddenly, especially if the cost hasn't been amortized yet.

If, given your situation, avoiding e-waste is a higher priority than saving money, it's possible that a repair shop could restore the device for the cost of a battery replacement. Or perhaps a refurbisher might buy it for parts.

i2c_mux I hope the future Motorola device will not be subject to such defects.

They will. All hardware components eventually fail, and for every component some of the individual failures will happen much earlier than others. Some will fail inside the warranty, some just after the warranty expires... some components don't get the opportunity to fail because some other component in a device fails first.

Manufacturers work from a statistical/population perspective. If there are a lot of early failures for some part, they may extend a warranty, offer a discount, etc. But there will always be some component failures right after warranty expiration.

It's even plausible that Motorola phones might contain batteries made in exactly the same factory as Google Pixel batteries.

de0u

i2c_mux In the Battery tab inside the Settings app, there was a message saying "Problem reading the battery meter." in a red rectangle, with a red "Learn more" text. The "Learn more" text/button did not work, nothing happened when I clicked it. I don't know if the scope of this bug is the AOSP or GrapheneOS.

Yes.

I suspect the bug itself is in AOSP, because Google may not have wanted people running an AOSP-derived OS on a Samsung device to click and be taken to a Google web site. I suspect fixing it in GrapheneOS would need to be the GrapheneOS project, but I also suspect the developers are more focused on things like VPN DNS leaks, Bluetooth regressions, etc., and that situation may continue for a good while. If you want to file an issue on the GrapheneOS issue tracker I think that would be fine (include the screenshot). If you don't have and don't want to have a GitHub account, let me know and I can.

i2c_mux

de0u Thank you for suggesting this. I do not own a GitHub account, so if you feel like this issue is relevant, you can go ahead and post it on the GrapheneOS issue tracker.

About the hardware failures in general, I think you are right. What I meant was more toward the repairability aspect of modern phones. For instance, changing the battery on my Pixel 8 Pro requires very careful operation and specialized tools, not to mention that opening the phone requires changing an inner graphite sheet plus replacing adhesives. I had in mind the 2027 enforcement of removable batteries in consumer devices by the EU. My point about the upcoming GOS-ready Motorola phone was that I hoped it would enforce this measure. And ship replacement parts.