Hacker News new | past | comments | ask | show | jobs | submit login
Battery power management for power users (2023) (akc3n.page)
32 points by laserstrahl 7 months ago | hide | past | favorite | 15 comments



No, this is bad advice. Yes, battery management software has gotten better, but you still can't get around physics. Keeping lithium polymer cells at close to 4.2V per cell for prolonged periods of time will degrade them quickly, regardless of any clever "maintenance mode" hacks that make the cell dance around the top 5% of it's capacity. But also, never charging past 80% will confuse the charge controller and start giving you incorrect percentages if it's not expecting you to do that.

Charge to 80% most days, charge to 100% occasionally, avoid discharging beyond 10 or 15%.


Maybe I'm just on a nostalgia kick today, but it's funny to think of when I was first introduced to that caution in the form of cyborg video- game dialogue.

> JAIME REYES: What kind of pain?

> WALTON SIMONS: Behind the eyes. A sharp burning. Almost electrical.

> JAIME REYES: How's your bioelectric level?

> WALTON SIMONS: It's always at a hundred percent. I like to stay prepared.

> JAIME REYES: That's probably it right there. Free radicals. You should charge your systems only when they've been significantly drained.

> WALTON SIMONS: I wasn't informed of that.

> JAIME REYES: It's a lot like an electric razor. If you leave it plugged in all the time, the battery loses its zero point. Just watch your levels.

> WALTON SIMONS: Interesting. Thanks, Doctor.

-- Deus Ex (2000)


> the GrapheneOS golden rule applies: don’t try to be a ‘power user.’ The devices GrapheneOS supports are thoroughly modern and will manage their own battery health automatically, far better than you ever could manually

So this advise is distorted by an apparent hate for "power users".

It's strange to suggest that modern embedded hardware is perfect. But more importantly, there's a conflict between the longest battery life your product can offer in initial reviews, and longevity years later. And while it's true that "fast" charging isn't necessarily bad at all, there is similarly a conflict between offering the fastest charging for reviews/comparisons, and keeping the temperature optimal for longevity.

So for their "power users bad" philosophy they should have just said "plug it in whenever you want, it'll be perfectly fine". To say you could get the best possible battery longevity by keeping the phone plugged in at all times (with the original charger?!) is absurd, especially combined with the fact that apparently grapheneOS doesn't have the usage-pattern-learning feature to keep charge at 80% until just before you unplug each day ...


I do all the things Peter Easton, the original author, says not to - running my phone between 20 and 80%, with an occasional full charge, and it's usually charged off a 1A charger, despite being capable of charging at 3x that.

It's a six year old iPhone 8, with the original factory battery, an estimated 76% of its original capacity, and its internal resistance is still low enough that it's running in 'peak performance' mode.

Easton knows nothing about electronics or batteries and has a terminal case of Dunning-Kruger.

"It's not 2005 anymore" really? Then why does the iPhone 15 offer the option to cap the max battery charge at 80%? https://support.apple.com/en-us/108055

Your phone is not going to explode in flames because of a malfunctioning power supply and it's beyond absurd to suggest it might.

USB is a POWER SUPPLY to the phone's mainboard - it powers the charger, it isn't the battery charger itself and the two are not directly connected. If the USB plug supplies the wrong voltage or negotiation doesn't happen, the phone simply won't charge. Furthermore, many cell phone batteries have an internal BMS that will disconnect the battery if it goes overtemperature, overvoltage, or overcurrent - charge or discharge. I've changed the battery in several android phones and each had a small battery protection board on it.

Charge rate absolutely impacts battery lifetime. If Easton opened a single datasheet from any lithium-ion battery manufacturer he could verify that himself. Manufacturers prioritize charging time, not battery lifetime. It's well known among iPhone users that if you constantly use fast-charge power adapters you'll wear down the battery much more quickly than if you use a low-power 5W adapter, say every evening when you go to bed and the phone has 6-10 hours to charge.

As long as the cable meets USB specs, it doesn't fucking matter who makes it.

"Back in 2005, gadgets with built-in lithium batteries, did not have their own onboard battery management which necessitated manual battery management."

He's talking about coulomb counters, aka 'battery gas gauges' and they've been in iPhones and Android phones for nearly a decade and a half.


That 80% option on iPhone 15, which Apple describes thuly:

> With iPhone 15 models, you can choose between Optimized Battery Charging, 80% Limit, and None.

> When you choose 80% Limit, your iPhone will charge up to about 80 percent and then stop charging. If the battery charge level gets down to 75 percent, charging will resume until your battery charge level reaches about 80 percent again.

> With 80% Limit enabled, your iPhone will occasionally charge to 100 percent to maintain accurate battery state-of-charge estimates.

is fairly easy to achieve on other iPhones and on iPads. Plug your charger into a smart plug that can be controlled from Apple Shortcuts. Shortcuts lets you make automations that trigger on battery level.

Make an automation that triggers on battery level goes about 80% that turns off the smart plug. Make another that triggers on battery level goes below 75% that turns on the smart plug.

I've been doing this for a while (just the 80% part) and it works great. It wasn't my idea. I saw someone here a few weeks ago mention it, and they got it from some blog.

I'm using TP-Link Tapo Mini Smart Wi-Fi Plugs [1]. I don't have a Matter hub so had to install the Tapo app to use them, but the Tapo app is controllable from Shortcuts.

I even get the "occasionally charge to 100% behavior", because sometimes Shortcuts seems to lose the ability to talk to the Tapo app. :-) Opening Tapo and turning something on or off seems to fix it.

If you don't want the "charge to 100%" behavior make a Shortcut to turn on the plug. When you are about to put the device on the charger use that shortcut to turn on the plug. It that works you know Shortcuts and Tapo are still working together.

If you buy smart plugs, I recommend getting ones with the same current and power rating as your outlets, which would be 15 A and 1875 W for a typical US home. There are smart plugs with much lower power limits meant just for lights. Those would be fine for a phone or tablet charger but then you've got remember that you can't use them for higher loads.

[1] https://www.amazon.com/TP-Link-Tapo-Compatible-Certified-P12...


> If you want to keep your phone’s battery health in the best possible condition it possibly could be in, you must use the Google charger and cable and keep them all plugged in around the clock.

Correct me if I'm wrong but the best state of charge for long term storage (not use) is around 60%.

While modern Apple devices (and probably Pixels too?) cap their charge at 80% if permanently plugged in, a. this indicates that 80% is better than 100%; and b. this partially matches with what I've heard of users with permanently plugged in devices/iPads in meeting rooms ending up with swollen batteries. Heck, my Framework has an option in the bios to limit max charge, and recommends that to be used if kept plugged in very often.

There is the counter-fact that if your system doesn't support a lower state of max charge, you may need to unplug the device and use it, ideally draining it from 80%-40% and back (as per Apple).

However, I rarely find swollen batteries in phones used in everyday life but have heard about it in the cases of permanently plugged devices. I think I'd rather just use the charge cycles of the device.

> Popular advice going around has been to slow charge with an older and slower charger rated for less current, to reduce the heat on the battery.

> If the battery’s getting too hot or the battery’s own conditions aren’t ideal for fast charging, the phone will slow down to preserve the health of the battery.

I'm sorry but this doesn't make too much sense. Heat is generally acknowledged as bad for batteries. Saying "the system will throttle charging if it gets too warm" doesn't mean it's necessarily good to hit that throttle temp in the first place. If 25 degrees is good, 30 degrees is not bad and 40 degrees is bad, it's great that it's not going to hit 40, but if I can restrict it to 25 that's even better.

Additionally, anecdotally the possible cause of "wireless chargers degrade batteries" is due to thermal degradation (again, what I saw probably on reddit). I have a 5w ikea wireless charger. It takes time to charge my Pixel 5, but my phone doesn't get too hot.

I'll be very happy to hear from the more experienced members of their views.


>> If you want to keep your phone’s battery health in the best possible condition it possibly could be in, you must use the Google charger and cable and keep them all plugged in around the clock.

> Correct me if I'm wrong but the best state of charge for long term storage (not use) is around 60%.

Your reaction is a non-sequitur to the author's point; they were being sassy by saying if you want optimal battery life, plug in and let the device manage voltage/battery health forever.

> this indicates that 80% is better than 100%

Better for what? 50% is better for long-term disconnected storage -- actually, voltage is better than % and for storage you're aiming for 3.7v/cell (which is roughly 40-50%) which is where the discharge properties slow almost to a halt. 80% is the kind of standard "I'm not discharging this charge immediately"-recommended state (see: EV charging recommendations). What the author mentioned in passing and didn't really explain is overprovisioning, which is where the device reports e.g. 100% charge, when the cells are really e.g. 92% charged, which some devices do to maintain battery health. So depending on your device and its configuration, trying to keep charge capped at 80% might in reality be keeping it much lower than you intended.

Anyway, the point of the article is for batteries with good battery management systems -- like modern phones -- it is better to not try and micromanage SoC and let the system do its thing. Those of us that have lots of naked cells should be doing that micro for those cells, but when built into a device with proper management there's no need and it could hurt more than it helps.


> Your reaction is a non-sequitur to the author's point; they were being sassy by saying if you want optimal battery life, plug in and let the device manage voltage/battery health forever.

I don't think that's a non sequitur; the author is wrong. Leaving the device plugged in won't give you optimal battery life, it will at best simply avoid it mattering. The firmware can reserve room and say it's at 100% when it's physically 90%, but that's still going to degrade the cells faster than 80% or 60%.


The fact is we (you and I) don't know what the Android phone's behavior is when plugged in 24/7 (note the author is a community mod for an Android-compatible OS). The UI can/does lie, and it's probably different from model to model, so at best we can say we're not sure if he's right or not. Technically using DC power instead of battery is best for battery health (dis/charge cycles are worse for battery health) but like I said before, it's just a snarky suggestion on his part.


>> this indicates that 80% is better than 100%

> Better for what? 50% is better for long-term disconnected storage -- actually, voltage is better than % and for storage you're aiming for 3.7v/cell (which is roughly 40-50%) which is where the discharge properties slow almost to a halt. 80% is the kind of standard "I'm not discharging this charge immediately"-recommended state (see: EV charging recommendations).

I meant better for the battery health in general, if kept permanently plugged in at that level. I agree 50% is likely much better than 80%, which is still better than 100%. It's just that very few OSes/software/BIOSes allow you to set it at 50 or 60%, while a few (like Apple) do fallback to 80% which is at least better than 100%.

Interestingly, I think my Framework BIOS allows me to set it at any value between 1 & 100%. I have set mine to a relatively "lower" 75%.


I charge my iPhone 13 pro (2021) with zero regard for battery health, and mostly via Qi, and iOS reports battery health at 89% which is much better than the expected -10%/year (I say expected because Apple will replace your battery if you lose 10% capacity within the first year).

I charge my Tesla whenever, usually limiting to 90% but doing 100% maybe 10% of the time. 2021 model, Teslafi shows 8% capacity loss.

Anecdotally I'm just not convinced that modern batteries with modern BMS need my micromanagement. Dealing with raw cells (18650s and the like) is a different story, but phones -- no way. But I'm not going to look down on someone if they feel passionate about it.


(I would greatly appreciate it if some clarification can be provided on what I said wrong, as I regularly deal with 21700 li-ion cells and likely will deal with more, and I’d like to do so safely. I’m assuming I was downvoted because I said something wrong, which in the world of lithium batteries, can be quite dangerous.)


> They were designed for certain assumptions taking into account certain limitations, and taking into account what most users tend to do.

...and that right there is why most[0] of this article is bad advice. It's not about being a "power user"[1], it's that I know more than the machine. I know when I'm going to want battery power available and when I won't. I know when I need to charge as fast as possible and when trickle charging is an option. For bonus points, Google's been trying to let the phone figure these things out... so then these folks apparently rip that out, too, lest the poor firmware get any help knowing what to do.

[0] The bit about not discharging to zero is good; I don't miss NiCd.

[1] Side note: "the GrapheneOS golden rule applies: don’t try to be a ‘power user.’" is... interesting advice for an OS that is almost by definition only used by power users; I would love to hear a definition of the term that somehow doesn't include people who replace their phone's OS.


This is in stark contrast with what has been said elsewhere: https://electronics.stackexchange.com/a/623375


The article suggests not charging your phone to 80% and unplugging, but doesn’t cover the recent iOS setting which allows capping your charge to 80%. That’s what I do and it offers best of both worlds.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: