By rjacobs


2016-01-11 23:44:14 8 Comments

There have been a lot of discussions about lithium-ion battery life across the internet. However, within these discussions I have found contradictory details about the way power is regulated between battery needs and device needs when a mobile/Android device is fully charged. It's clear that the charging process itself is a "smart" process (multiple stages with different current, overcharge protection, etc.), but what happens when the charging stops and the device is still in use? My understanding is that in these situations most laptops have regulators that divvy AC power between directly powering the device and "topping-off" the battery as it looses charge to ambient factors. Do modern Android devices (and iPhones for that matter) do the same thing?

Consider the following hypothetical situation:

  • Device is plugged-in to AC mains with 1 Amp supply
  • Battery is charged 100%
  • Current ongoing device use draws < 1 Amp
  • Battery naturally looses 5% charge in 3 hours when not in use (I have not idea what an actual real word value would be here, but the point is that there is a slow loss even when no power is drawn from the battery)

In this case is the battery bypassed completely as long as the active needs of the device do not exceed 1 Amp, or does the battery remain an active power source regardless of charge state and cable connection status? If the battery is in-fact bypassed, does the device wait for some charge-level threshold (say 95% charge) before touching the battery to minimize accumulating (micro)charge cycles?

It's possible that this is more of an electrical engineering question, and though it has certainly come up in that context (such as here), the related discussion seems either too broad or too specific to laptops.

Update

I've been learning a bit more on the correct terminology (thanks to @beeshyams). I see that the concept of "parasitic load" plays a very important, and detrimental, role in charging efficiency. However, I am led to believe that it is not necessarily a key variable in the fully-charged state described above. Anyway, without getting too deep into semantics, it seems like the key variable would instead be "self-discharge" (the natural 5% loss in my example).

My original thought was that this self-discharge effect would be so small, and lead to extremely infrequent topping-off (micro)charge cycles, that it would have a negligible effect on the cumulative charge cycles of the battery. Therefore, by isolating the active load of the device to AC mains (keeping it plugged in most of the time), instead of drawing from the battery, one could potentially prolong battery life. Even though everyone says "don't do that" they never really say why, and that's really what led to this question.

What I'm now starting to see, and which someone will hopefully confirm, is that any charge cycles (even small and infrequent) are really really bad news when they happen near full charge:

  • They happen much faster than one would expect (the effects of self-discharge are increased near full charge).
  • They are really detrimental (it looks like a 5% cycle between 95%-100% can lead to several magnitudes more "wear" on the battery than a 5% cycle between, say, 45%-50%.
  • They exacerbate the negative effects of heat (heat at high change seems to be much worse for battery wear than the same heat at low charge).

So I suppose it's safe to say that these negative factors, even if only considering self-discharge as a catalyst for cycling, nearly always outweigh the benefit of diverting normal operational power away from the battery?

1 comments

@beeshyams 2016-01-12 04:03:02

Good question and to give a definitive reply is not possible since OEMs use different battery charging algorithms both at hardware level and OS level. Details of these are deemed "proprietary" and not available on public domain.

By way of providing the background to understand:

  1. Battery Charging circuits have a dedicated "Power Management IC (PMIC) to take care of battery charging operations. See PMIC from Wikipedia. The complexity of these chips has increased with multi core processors. These ensure that Power required for charging is pulled rather than pushed- meaning to say that chargers draw what is required rather than what they can possibly draw from the outlet
  2. Battery Charging circuits "isolate" power required for charging the Battery and providing power required to use an application. For a detailed understanding of this,see my answer https://android.stackexchange.com/a/131169/131553. For purposes of this question, keeping the device on while charging constitutes a load, however small it may be
  3. Refer Battery University post on charging Li-Ion batteries, which clearly depicts at Fig 1, various stages of battery charging-Constant Current stage, Saturation stage, Ready no current stage and Standby mode

Coming to your question,

*In this case is the battery bypassed completely as long as the active needs of the device do not exceed 1 Amp, or does the battery remain an active power source regardless of charge state and cable connection status?

  • Battery remains as an active power source. That is what is depicted in standby mode of referred battery charging article which shows current flow to " top up" depletion of charge. As to what level this topping starts send to be dependent on OEM implementation. For my Huawei Honor 6 device, it starts at around 97% and for Samsung Note 2, it starts way below at around 91%. In both cases, cable is plugged in and charger is switched on
  • Cable connection status as asked in your question is answered partly above, but if the cable is disconnected, there is no point of discussing charging!!

    If the battery is in-fact bypassed, does the device wait for some charge-level threshold (say 95% charge) before touching the battery to minimize accumulating (micro)charge cycles?

  • As clarified above, it seems to be device and OEM dependent

  • Coming to micro cycles accumulating and resultant parasitic load (not asked specifically by you but micro cycles of charging and discharging cause parasitic load)- this aspect had been investigated by me and vetted by Battery University, in my answer is-power-consumed-from-battery-to-run-the-phone-when-Charging/. In this experiment, a load equivalent to mobile data download (by keeping screen on at 50% brightness" was loaded when charging was ongoing to study micro cycles and parasitic load and results submitted to Battery University for their views. Quoting Bruce Huang verbatim (emphasis mine)

I don't think the parasitic load is going to prevent the current drop forever. I guess the parasitic load only delay the current drop for a period of time. In this test, the parasitic load may be too small (turning on screen with 50% brightness is not huge load) to create a significant delay to the current drop.

tl:dr;

  • "Smart" management of battery charging is done by Power Management IC (PMIC) in Android phones. Key aspects relevant to question are:

    • Battery charger pulls power it needs

    • Battery charger isolates charging requirements from usage requirements and unless the latter is extremely high, will not affect adversely

  • At what level of battery will charging re-commence to top the battery seems to be OEM dependent

  • Micro cycles accumulation (and consequent parasitic load) are negligible in the situation described by you

  • Is it safe to keep the Battery plugged in after full charging? Short answer NO , due to heating effects which causes internal stress. Detailed explanation on this available at Battery University article linked above, which clearly shows the deleterious effects of heating on battery health (Related: See this is your battery gauge lying to you (and it's not such a bad thing)

Edit: In response to edit by OP

As you have rightly concluded but packaged it differently, I do agree with your line of thinking but haven't come across any study which confirms it. You could consider this portion as a separate question in SE Electrical, which may elicit a suitable reply. My approach (simplistic?) would be to factor the potential damage at high voltage to the battery cycle life and that should be of more importance from end user perspective

@rjacobs 2016-01-12 17:44:48

This is a very thoughtful answer with good added references. One thing I'm not clear on however is the statement that "Battery remains as an active power source [at 100% charge and still plugged in]." Is this active as in "ready to provide current if device needs exceed the 1 Amp the charger can deliver", or active as in "always providing some of the power demanded by the device"? If the former it would seem that leaving the device plugged in would always reduce battery cycling... with the only (though notable) detrimental factor being the increased heat during any usage.

@beeshyams 2016-01-12 18:23:25

Thank you. I meant readiness with reference to the battery primarily to connect with the last stage of graph in battery university link which shows current again being pulled and dropping to zero. Having said that, the distinction of readiness between the amount of some/more than 1A is immaterial since the Battery and usage by Battery are isolated and powered separately (as those detailed tests in my linked answer show ). Of course, battery being on itself while charging is a small load. To sum it up it doesn't matter as 1A load > some small load

@beeshyams 2016-01-12 18:34:12

About battery cycling reduction- it is the opposite. Your battery is worse-off if it is charged to full voltage capacity. The battery life is reduced if you do this compared to charging at less than full. Personally, I stop charging at 85%. Table 4 of this article makes it very clear. This aspect is not commonly known/implemented. I trust this answers and yes, do consider upvote and "accepting" the answer to reflect it has answered you and has merit for others who refer later (your choice of course)

@rjacobs 2016-01-14 06:59:15

I did upvote as you've definitely exposed the key variables that I was missing. I was getting too distracted by the concept of parasitic load, which (as noted) is probably irrelevant here. What has taken time for me to understand however is the concept of self-discharge (when all active device power is over cable), and the non-intuitive notion that it could somehow actually be worse than drawing the same amount of device power via battery. Anyway, I updated my question to capture this. That said, my edit may simply be re-summarizing some of your points from a slightly different perspective.

@beeshyams 2016-01-14 08:15:10

It is rare and good to find someone who is getting into these kind of details but as noted in the edit to my answer, it is unlikely that those points would be sufficiently answered here and SE Electrical is a better bet. I would also suggest that you mail battery university (not post on their articles, that is rarely replied). You may also get lucky in getting a reply as I did and that could be the second and probably faster way

@rjacobs 2016-01-14 16:53:58

Thanks for the follow-up. I suppose some of this is starting to border on the realm of theory, as building a really precise understanding of all these variables sorta has diminishing returns when you consider all the variables at play in real world situations. Either way, I enjoy building an understanding of these things, especially given the huge variety of (sometimes conflicting) info available out there to the casual observer.

Related Questions

Sponsored Content

1 Answered Questions

[SOLVED] Change battery meter behavior

1 Answered Questions

1 Answered Questions

[SOLVED] Is the fast charge safe for long-term usage with Li-Poly battery?

1 Answered Questions

[SOLVED] Android definition of charge cycle?

2 Answered Questions

[SOLVED] Is battery calibration important for Android phones?

1 Answered Questions

2 Answered Questions

[SOLVED] What is the difference between USB charging and AC charging?

Sponsored Content