"Chargie", a battery life saver for non-rooted devices?

[Rob. S. 1,664]

I haven't found much with regard to independent reviews on the net and the company seems to be quite new, but from what I see there (including the feedback on Play Store for the accompanying app), it looks very, very promising. It's basically kind of a USB stick that sits between the charger and the phone, talking to a dedicated app on the phone, stopping the charge at a user-defined limit like 80%.

I've just ordered three, and while I can't test it with a Pro¹ because I haven't got mine yet, I'll be giving it a thorough try. Wish I would have had something like that for the last couple of phones I had, all of which had battery problems much too soon! 

https://chargie.org

[AnnieC 115]

It's interesting, but it seems a little pricey for what it is. I'm eagerly waiting for your review.

[jlavi 112]

According to the Battery University, keeping Lithium battery fully charged at 25C one year degrades it 20%. Limiting temperature and max charge slows the decay an gives more charge cycles.

I think it would be best to keep the battery at 40% charge and keep the phone in the fridge in an airtight bag with activated silica gel packs and only keep and use the phone in fridge like temperatures. Touching the phone with fingers or cheek, no-no-no.

 

[Rob. S. 1,664]

😁

From their measurements, it seems that simply charging up to 80% instead of 100% would prolong the usable battery life by more than 50%. Like 3 years instead of 2. That's already quite nice...

[Jordi 72]

Seems to be a good solution for non rooted devices.

For rooted devices, I just tested "Battery charge Limit" (open source), which seems to work pretty well: I set a limit at 90 % charge and the charge slow down before this limit and does not exceed this value.

[Hook 3,043]

58 minutes ago, Jordi said:

Seems to be a good solution for non rooted devices.

For rooted devices, I just tested "Battery charge Limit" (open source), which seems to work pretty well: I set a limit at 90 % charge and the charge slow down before this limit and does not exceed this value.

If you are not rooted, Accu Battery is a useful app.  It is not as good as Battery Charge Limit as it doesn't do the slowdown and stop, but you can set an alarm for a certain charge limit.  Unfortunately, not great for charge while you sleep, but cheaper and less easy to lose than the Chargie.  😉

[claude0001 1,352]

While I am no expert in Li-ion batteries, restricting the discussion to the question what is the maximum "healthy" charge of the battery seems like over-simplifying the problem to me.

Indeed, it seems evident that consistently keeping the charge of a battery close to some intermediate fraction (around 50%) of its nominal capacity does slow down degradation of that nominal capacity. See e.g. Grolleau et al. (2013) [doi: 10.3390/wevj6030549]. However, the downside of that strategy is just as obvious: this way, you will never take advantage of the full capacity in the first place! Essentially, it's like never wearing your precious jewellery, out of fear of loosing it ...

For me at least, this ansatz is thus not viable: when I get up in the morning I cannot predict how intensely I will use my phone (or laptop) during the day. So I need them fully charged so as to be prepared for the worst. 

Now, if we agree that we do want to use the battery to its full potential, the study by Grolleau et al. shows something more: The time spent at full (100%) charge is not the most important factor when it comes to cell aging. In fact, a battery kept indefinitely at 100% charge looses its capacity more slowly compared to one that is cycled between 100% and some intermediate level, thus only spending a fraction of its life at full charge (see Fig. 3 in [Grolleau et al., 2013]). This challenges a basic concept behind "Chargie" (and probably similar solutions).

A lot of evidence seems to suggest that the most important degradation factor is not the time spent at maximum charge level, but the charging process itself. Especially the heat generated during charging is known to be detrimental for various reasons. In that context, "fast charging" is discussed critically (as it naturally generates more heat). See e.g. Tomaszewska et al. (2019) [doi: 10.1016/j.etran.2019.100011]. While some "fast charging" strategies seem to be more battery-friendly than others (depending also on environmental conditions!) there seems to be consensus that slow charging is generally superior to all when it comes to battery life.

I find that the "fast charging" wall plug that shipped with the Pro1 heats the device to an insanely high temperature -- so that touching the metal case becomes uncomfortable. As I use a rather thick flip case, the problem is likely even amplified by thermal insulation. To me, the consequence is clear: I never use "fast charging". Instead, I charge the phone slowly using my laptop's USB port. The latter can be configured to be "always-on", so the computer does not need to be running for that. This way the phone stays cool, and I need to disconnect it before going to bed (as I use it as alarm clock 🙂).

Hopefully, replacement batteries will be available some day, so we can relax about this topic ... 😉

Edited February 7, 2021 by claude0001

[VaZso 1,998]

34 minutes ago, claude0001 said:

I find that the "fast charging" wall plug that shipped with the Pro1 heats the device to an insanely high temperature -- so that touching the metal case becomes unconfortable. As I use a rather thick flip case, the problem is likely even amplified by thermal insulation. For me, the consequence is clear: I never use "fast charging". Instead, I charge the phone slowly using my laptop's USB port. The latter can be configured to be "always-on", so the computer does not need to be running for that. This way the phone stays cool, and I need to disconnect it before going to bed (as I use it as alarm clock 🙂).

That is why I have bought a charger which has two USB connectors:
- One for Quick Charge compatibility
- Another one which only supports 5V 1A

I use the latter to charge my phone and it can charge it in a reasonably fast rate which allows me to plug it to charger for evening and have it charged in the morning (after a few hours it is done).
The other port still makes possible to use fast charging when I may need it (if the phone is depleted in the morning as I may have not connected it in time).

Also, I have a QC compatible charger in my car in case of "emergency", so I may charge it up quickly when necessary.

...but basically I only use my 5V 1A charging port as it does not heat up the battery and still fast enough for daily use.

[Rob. S. 1,664]

18 hours ago, Hook said:

If you are not rooted, Accu Battery is a useful app.  It is not as good as Battery Charge Limit as it doesn't do the slowdown and stop, but you can set an alarm for a certain charge limit.  Unfortunately, not great for charge while you sleep, but cheaper and less easy to lose than the Chargie.  😉

Indeed I'm right now testing AccuBattery on my not-rooted Motorola phone and indeed it looks good, for all a not-rooted device allows it to do. That said, I already have too many things I need to watch or somehow keep in mind, so I really don't need more of it and that's why I'll gladly pay a bit for anything that prevents the addition of another one to all those things 😉 

[Rob. S. 1,664]

8 hours ago, claude0001 said:

While I am no expert in Li-ion batteries, restricting the discussion to the question what is the maximum "healthy" charge of the battery seems like over-simplifying the problem to me.

Indeed, it seems evident that consistently keeping the charge of a battery close to some intermediate level (around 50%) of its nominal capacity does slow down degradation of that nominal capacity. See e.g. Grolleau et al. (2013) [doi: 10.3390/wevj6030549]. However, the downside of that strategy is just as obvious: this way, you will never take advantage of the full capacity in the first place! Essentially, it's like never wearing your precious jewellery, out of fear of loosing it ...

For me at least, this ansatz is thus not viable: when I get up in the morning I cannot predict how intensely I will use my phone (or laptop) during the day. So I need them fully charged so as to be prepared for the worst. 

 

Obviously a charge limit option then is not for you...

My usage is completely different; my phone is within reach of some charging device most of the time every day, and usually stays connected to one as long as I'm in the office (whether it's the real office or, like most of these days, home office...).

When I need or want to go out for an unclear amount of time, I can make an exception and up the charge from 75% or 80% (and I wouldn't set the charge limit below that) to 100% within no more than five minutes or so when quick charging (which I normally never do), potentially even in the car on my way to wherever I expect to leave the vicinity of a charger, without really hurting the overall effect of the charge limit regime over the battery's lifetime .

I've looked into your source where you say "a battery kept indefinitely at 100% charge looses its capacity more slowly compared to one that is cycled between 100% and some intermediate level". I'm not sure this is fully applicable to our case here. For one, when limiting the charge in a use case in which keeping a battery "indefinitely at 100%" is an option (like mine, but seemingly not yours),  we're not "cycling between 100% and some intermediate level" – we're keeping the charge indefinitely at 75% or 80%. And even if we were "cycling", we would be cycling between 75/80% and some even lower level, which again is something completely different regarding battery health. It's exactly what is supposed to make the difference.  

By the way, subjectively I've found most of my past phones' batteries deteriorate unnaturally quickly despite my keeping them at 100% much of the time...

 

[Jacob_S 128]

8 hours ago, claude0001 said:

Now, if we agree that we do want to use the battery to its full potential, the study by Grolleau et al. shows something more: The time spent at full (100%) charge is not the most important factor when it comes to cell aging. In fact, a battery kept indefinitely at 100% charge looses its capacity more slowly compared to one that is cycled between 100% and some intermediate level, thus only spending a fraction of its life at full charge (see Fig. 3 in [Grolleau et al., 2013]). This challenges a basic concept behind "Chargie" (and probably similar solutions).

😉 (Emoji doesn't want to be removed?)

Figure three of the paper from Grolleau et al. you are citing actually strongly supports the idea of not charging the battery fully.

Scenario 1: Lot of charging cycles to 100% lead to fast degradation

Scenario 2: Reducing the number of cycles by using the battery until lower percentage and only then recharging still lets the battery age slightly faster than at static storage conditions (at 100%)

Scenario 3: Same number of charging cycles as in Scenario 1, but keeping the battery at 20-60%: Slightly improves battery life compared to the static storage.

So you were kinda right, when you were saying that keeping a battery at 100% for longer isn't that bad. What is really bad is GETTING to 100% in the first place. And that's why limiting the maximum charge to something like 90% (compromise between not charging fully and still allowing as few charging cycles as possible) is a very reasonable thing to do.

[claude0001 1,352]

On 2/6/2021 at 2:44 PM, Jordi said:

[...] I set a limit at 90 % charge and the charge slow down before this limit and does not exceed this value. [...]

 

On 2/6/2021 at 3:50 PM, Hook said:

[...] you can set an alarm for a certain charge limit.  [...]

 

10 hours ago, Rob. S. said:

[...] we're keeping the charge indefinitely at 75% or 80%. And even if we were "cycling", [...]

 

 

10 hours ago, Jacob_S said:

Figure three of the paper from Grolleau et al. you are citing actually strongly supports the idea of not charging the battery fully. [...]

 

Yes, as I wrote (and never disputed), it is well-known that not using a battery to its full cell capacity saves lifetime. Also, please do not get me wrong: feel free to experiment with your devices as much as you want. Limiting the maximum charge will probably never hurt anything.

That said, let me add that I believe the Pro1's battery management system already limits the maximum charge to 90% nominal cell capacity anyway.

The Pro1 spec sheet advertises a battery capacity of "3200 mAh", which likely corresponds to the maximum cell capacity that could theoretically be used. However, the Linux kernel reports a "design" capacity of only 2900 mAh:

puma@rostkatze:~> upower -i /org/freedesktop/UPower/devices/battery_bms            
  native-path:          bms                                                        
  power supply:         yes                                                         
  updated:              Sun 07 Feb 2021 11:03:51 CET (112 seconds ago)                
  has history:          yes                                                              
  has statistics:       yes                                                              
  battery                                                                                   
    present:             yes                                                                  
    rechargeable:        yes                                                                    
    state:               discharging                                                               
    warning-level:       none
    energy:              12.4099 Wh
    energy-empty:        0 Wh
    energy-full:         12.6632 Wh
    energy-full-design:  12.6632 Wh
    energy-rate:         1.47598 W
    voltage:             4.39208 V
    time to empty:       8.4 hours
    percentage:          98%
    temperature:         25 degrees C
    capacity:            100%
    icon-name:          'battery-full-symbolic'
  History (rate):
    1612692231  1.476   discharging

Note that 12.6632 Wh / 4.39208 V = 2.88319 Ah. This is quite precisely equal to 90% of the advertised cell capacity (3200 mAh x 0.9 = 2880 mAh).

I do not think this is surprising -- many devices are known to limit cell charge in order to save battery lifetime as discussed above. However, it could mean that further manual intervention is not all that necessary ...

Edit: Turns out this conclusion is wrong (as pointed out by @Jacob_S below): the voltage reported by upower reflects the actual (instantaneous) voltage of the battery which changes as a function of charge level. At full charge (as in the figures above) the voltage is highest, which falsifies the calculated charge (in Ah) towards smaller numbers. 

P.S.: How to you create a code box for listing terminal output in a nice way? Ok, thanks @Rob. S., that was too obvious ...

Edited February 7, 2021 by claude0001

[Jacob_S 128]

19 minutes ago, claude0001 said:

 

 

 

Yes, as I wrote (and never disputed), it is well-known that not using a battery to its full cell capacity saves lifetime. Also, please do not get me wrong: feel free to experiment with your devices as much as you want. Limiting the maximum charge will probably never hurt anything.

That said, let me add that I believe the Pro1's battery management system limits the maximum charge to 90% cell capacity already anyway.

The Pro1 spec sheet advertises a battery capacity of "3200 mAh", which likely corresponds to the maximum cell capacity that could theoretically be used. However, the Linux kernel reports a "design" capacity of only 2900 mAh:

puma@rostkatze:~> upower -i /org/freedesktop/UPower/devices/battery_bms            
  native-path:          bms                                                        
  power supply:         yes                                                         
  updated:              Sun 07 Feb 2021 11:03:51 CET (112 seconds ago)                
  has history:          yes                                                              
  has statistics:       yes                                                              
  battery                                                                                   
    present:             yes                                                                  
    rechargeable:        yes                                                                    
    state:               discharging                                                               
    warning-level:       none
    energy:              12.4099 Wh
    energy-empty:        0 Wh
    energy-full:         12.6632 Wh
    energy-full-design:  12.6632 Wh
    energy-rate:         1.47598 W
    voltage:             4.39208 V
    time to empty:       8.4 hours
    percentage:          98%
    temperature:         25 degrees C
    capacity:            100%
    icon-name:          'battery-full-symbolic'
  History (rate):
    1612692231  1.476   discharging

Note that 12.6632 Wh / 4.39208 V = 2.88319 Ah. This is quite precisely equal to 90% of the advertised cell capacity (3200 mAh x 0.9 = 2880 mAh).

I do not think this is surprising -- many devices are known to limit cell charge in order to save battery lifetime as discussed above. However, it could mean that further manual intervention is not all that necessary ...

P.S.: How to you create a code box for listing terminal output in a nice way?

Ah, yeah, sorry. Slightly misunderstood your post. Now that I read it again I understand what you meant. 

On your recent post:

I'm not entirely sure if you can calculate like this, as the voltage provided by the cell significantly lowers on low charge levels. Wouldn't you have to insert some mean value for the voltage? Something around 3.9V would lead to the advertised 3200mAh.

I think that's another example why using Wh would just be superior to using Ah (also for comparison of the batteries of devices running on different voltage). 

[claude0001 1,352]

24 minutes ago, Jacob_S said:

I'm not entirely sure if you can calculate like this, as the voltage provided by the cell significantly lowers on low charge levels. Wouldn't you have to insert some mean value for the voltage? Something around 3.9V would lead to the advertised 3200mAh

I see your point, but I cannot remember seeing a much different value reported by upower. Maybe it is even some average already? However, I also do not recall ever looking  the numbers at deep charge level. Will report.

[jlavi 112]

I've ordered the spare battery for Pro 1-X and need to store it for couple of years and minimize its aging.  Keeping it in 40% charge at fridge and avoiding moisture sounds the best bet.  In those condition the battery should age in one year the same amount fully charged battery ages in room temperature.

[Rob. S. 1,664]

3 hours ago, claude0001 said:

P.S.: How to you create a code box for listing terminal output in a nice way?

I'd just use the Code button (between the Quote button and the Smiley button)  with the "no syntax highlighting" option 😉 

[claude0001 1,352]

9 hours ago, Jacob_S said:

I'm not entirely sure if you can calculate like this, as the voltage provided by the cell significantly lowers on low charge levels. Wouldn't you have to insert some mean value for the voltage? Something around 3.9V would lead to the advertised 3200mAh.

Turns out you are right: now, at 85% charge level, upower already reports 4.13 V battery voltage. With the above 12.6632 Wh this would already correspond to 3066 mAh design charge. I'll keep monitoring the reported voltage as it discharges, and will calculate some approximate average. I can now easily imagine that we end up very close to the expected 3200 mAh.

[EskeRahn 5,471]

Did not read all the latest activity here, but generally keeping a battery in good condition is a bit of a science.
If you turn off a phone, so only the circuit waiting for a power-button press is active, a phone can last for ages. It is not unusual that it is still well above 90% after half a year.

So if that is what is meant by "kept at 100%", it is clearly harmless. But if it is plugged in and applied a high voltage constant trying to pull it up, it is not a good idea.
So if kept plugged in it is clearly a good idea to stop at roughly 80-90%, as it is the last part that is hardest, and that is also why most chargers slow down with lower current here, to reduce the strain / but the strain is still there.

Generally one of the worst combinations are heat during charging, That alone does fast charging and wireless charging very problematic, unless it is actively cooled meanwhile. LiIon batteries are a bit like humans in that they prefer 'room temperature'. Already as low as 30°C is considered warm for a LiIon battery.

One above had a not wearing your jewellery analogy.  I rather see an analogy in our own feeding habits. It is not really a substantial health problem if we occasionally eat fast food and/or eat way too much. But it is not good for us if we do it on a daily basis. So use the extremes like 0%-100%-0% cycling when you really need it, and the same for fast charging. But don't do it daily. I personally do like @VaZso above a charger with BOTH a slow 5V 1A and a QC3.0 socket. And normally use the 1A socket.

A LOT can be found on batteryuniversity.com. I made several posts in my own blog derived from that and personal experienced. Start e.g. here

[jlavi 112]

> Generally one of the worst combinations are heat during charging,

Would fast charging be ok if the battery were actively cooled and kept below room temperature? If one would charge the phone in fridge?

[claude0001 1,352]

2 hours ago, jlavi said:

Would fast charging be ok if the battery were actively cooled and kept below room temperature? If one would charge the phone in fridge?

Not according to the review paper I have linked above. Charging of cold batteries seems to be especially problematic. E.g., as charging heats the battery from within, aggressive cooling from outside leads to thermal gradients that seem to cause their own problems.

For these reasons, batteries of electric cars are actually warmed up using dedicated heating systems during (and prior to) charging in winter ...  

Edited February 8, 2021 by claude0001

[Rob. S. 1,664]

Here's a small update. Got the "Chargie" today and the first tests are encouraging. The device seems very much thought through, so far everything works as expected, without much setup at all. The app is well designed and user-friendly. For us geeks, it keeps three hours or so of statistics about what actual charging power is being used to up the charge to the configured limit. If someone insists on quick-charging their phone even though it has a tendency to be detrimental to the battery health the "Chargie" is supposed to preserve, it does that, too, it is even compatible with the non-standard "Turbo Power" quick charge mode of my Motorola phone, which is a rare thing even for third-party chargers and cables. Max specs are 16V/6A.

 

 

Edited February 11, 2021 by Rob. S.

[Rob. S. 1,664]

For what it's worth, there's a new version of the hardware device out (or, rather, available for preorder), the Chargie C, a Power Delivery (PD) compatible, USB C charge limiter for phones which also limits the PD charging power to 8W. As our device doesn't use PD but rather Qualcomm's Quick Charge, I don't suppose the charging power limit will work for us; otherwise the device might have made sense even if we're rooted (as I have been for a while with my Pro1, too, so no need to charge my Pro1 through the Chargie anymore).

The description on the website is far too long and there's no concise feature list or technical data, either, but something like that has been promised to be added over the next few days.

[VaZso 1,998]

23 minutes ago, Rob. S. said:

As our device doesn't use PD but rather Qualcomm's Quick Charge,

...as far as I know, theoretically Pro1 supports PD but I don't know if it really supports it.

Anyway, SD835 supports QC4 which has PD support and also F(x)tec shows PD under technical details of Pro1.

So it may happen also power limit works if it does not  have a fallback option for QC which allows higher power.

[EskeRahn 5,471]

I'm pretty sure the Pro1 does NOT support PD. You can charge the device though a PD port though.

[dreamflasher 120]

I've ordered a Chargie a couple of months ago, because I had problems with AccA (if you're root that's the best battery control app). At first it worked great, but after a week I had regularly all kinds of problems: The app not connecting with the chargie, chargie not stopping to charge (so battery was always at 100), or chargie not charging (so the battery drained). Overall I am fairly disappointed with chargie, and I am glad that with an update of AccA/acc it works again. Also chargie claims they could work without the app, but I never managed to get this to run.