- 12-12-2011, 01:32 AM #6
Looking at the app developer's website, Battery Health is a measure of how much energy the battery stores relative to the manufacturer's claim, see below. So, if RIM has been sticking batteries in some units that don't meet its own specs, you would see figures below 100% even on new Playbooks. Given that most manufacturers exagerate specs somewhat and that RIM of late has a worse than average record when it comes to the truth, figures well below 100% would not surprise me. Another thing to consider is that if the Playbook is fairly new, the battery may still not be taking a full charge and may improve soon--don't know why, but that seems to be the case with some lithium ion batteries.
On my Playbook, Battery Health is reported as 82%
"What is "battery health"?
In short, it's how much charge your battery holds compared to the manufacturer's specifications.
The PlayBook's two batteries, combined, are designed to hold 5400mAh of energy. In actual fact, no lithium battery ever makes it out of the factory with the full capacity as printed on the label (i.e. with 100% health) and, over time, this capacity slowly drops further.
Most batteries even a month old will likely have dropped to around 90%, and over the next couple of years this will drop gradually to around 50%. At that point your battery will provide only half the designed capacity."
- 12-12-2011, 09:30 AM #10
Consider 94% for a fairly new device to be pretty typical. There's an older thread where people were reporting their values...
I currently have one with 28 recharges, at 91% health, and another with 153 recharges and 89%.
Note also that my comments on the web site, about the "manufacturer" and nominal values, refer to the battery manufacturer, not RIM. RIM has no direct control over the battery in that sense, so the stuff about "no lithium battery every gets out of the factory with 100%" health is pretty much not their fault. If you want to beat someone up for not truly having 19.98Wh in your battery, find who makes the battery and send them a nasty-gram. :-)
The drop to 61% in the first few days, then settling back up around 85-95% is probably not unusual either, since the software needs a few cycles to calibrate itself properly, and would also be affected by your recharge patterns. If you never let it drop below a battery level of, say, 75% in the first few days, it might not be showing as accurate a "health" calculation as it might if you drained it right down once or twice.
I'll summarize: none of the readings mentioned before this post look very unusual or "kind of low", so I don't think anyone here has cause for concern.
- 12-12-2011, 09:40 AM #11
If you look the 2 post above yours(f2 and king benny)they have similar charge counts but different health. Mine is 127 and 87%. Why is that? Just a matter of battery construction or does charge habits play a big role?
I would also like to know from F2 and king benny if you have any charging habits or is just whenever.
12-12-2011, 11:00 AM #12
- 19,436 Posts
Just whenever but usually overnight.
I do need to do more followup with Peter. Mine tends to lose about 13% overnight on standby. I'd like to see if I can figure out why. The other night I turned off wifi/bt and still the same loss. I don't know if it is partying while I sleep or visiting strange websites or being a bot for some Gambian banker!
- 12-12-2011, 05:36 PM #14
The devices as shipped may have significant differences. It doesn't appear that battery manufacturing, at least for this sort of product, is much of a science.
I think there's also a large element of "the measurements just aren't that precise", which is why I'm telling people not to be concerned about 5 or 10% difference with other people, at least for now. What matters more is whether your reading is stable.
That said, what we probably could do is develop a bit of a test for battery life, and something that might exercise the thing the right way and trigger a recalibration of the readings (mainly an update to the health estimate).
1. enable airplane mode
2. restart tablet
3. launch only Battery Guru (or another app with a Keep Awake feature)
4. plug into charger
5. charge fully (not just to 100%)
6. disable auto-dim, set "Battery mode" brightness (backlight) on minimum
7. enable Keep Awake on Battery Guru's screen
8. remove from charger
9. let it sit untouched until it turns itself off (approx. 9h)
10. plug into charger and recharge fully again, preferably re-launching BG at the start
For one thing, that should be a good test (if I haven't overlooked anything) of relative battery life for people.
For another, it may trigger the recalibration of the health value, as RIM's knowledge base article suggests. (Though I think that can occur any time within one or two more recharge cycles afterwards, not necessarily right away.)
Planning to try this tonight if no one else has a suggestion for something I overlooked, which could make the results variable. Note that the point isn't to maximize the discharge rate... setting backlight to 100% would get the job done in only 6h but the results would be less accurate. I'm trying for a test that should give very close results on two units with similar actual battery life.
If anyone else tries this, please do an export afterwards and send me the file. I'll do some stats on the results and maybe give us all a benchmark.
(For reference, under these conditions mine uses 1.93W at 0% brightness, compared with 2.93W at 100% brightness.)
(Edited, to mention doing this without Battery Guru. Use some other app, like AIR Browser or Stay ON, for the keep awake function if you don't want to buy Battery Guru.)
Last edited by peter9477; 01-11-2012 at 08:49 PM. Reason: add note about doing this without Battery Guru
- 12-12-2011, 05:50 PM #15
I'll give it a try. So guru will keep the log even after the PB battery drains?
Are you using the time it takes to drain from full as the measurement of battery health?
One more thing. Should the app be full screen? That would keep the light on for the duration correct?
Last edited by Chaddface; 12-12-2011 at 05:57 PM.
- 12-13-2011, 08:41 AM #16
Chaddface, Battery Guru can log only as long as it's running, so as soon as the battery is fully drained and shuts the tablet off, that's the end of that logging session (but also of any drain). You do need to relaunch it manually after restart too, of course, since we have no way to register apps for auto-start or running as background services for now.
The first time you use the Keep Awake feature in the latest version, you'll notice an informational dialog telling you how it must be used. As noted there, some portion of the Battery Guru window must be visible, either fullscreen or minimized, even if it's just one edge of it because you have it shoved off to the side. Even a single pixel showing is adequate, but if it's completely hidden then it doesn't work. That is, as noted there, an OS limitation and there's nothing anyone can do about it.
The "health" measurement is not something Battery Guru calculates. In fact, currently all values shown are reported directly as the OS reports them. I don't know the exact algorithm they use to calculate it, but it involves a periodic update based on recent history that probably involves more than just how long it took to drain. Maybe doing the full drain once triggers the routine to update that value... I don't know, but I'm hoping this little experiment will reveal something if a few of us do it and watch for changes in Health in the day or two thereafter (and then send me the export file).
- 12-13-2011, 08:59 AM #17
So my drain-test last night produced the following results:
Health before: 91%
Recharge count: 28, went to 29 while draining during the test
Duration (100% to shutoff): 9 hr 58 min (with the last 30 min at 0%)
Average power: 1.862W (calculated over the drain-test period in the export file)
Recharging now... will report again if I actually see any update to the health reading. Given that my own math says health is 93% based on this, it could change, but I suspect it's close enough that it won't. Also I think they may include a bit of a safety margin in the numbers (thus why we often sit at 0% for a while) so they may artificially lower the health value a bit to account for that.
By my math: 1.862W over 9.967h is 18.558Wh. That's 18.558/19.98 or 92.9% health. Probably not enough difference to be considered significant, and I'd want to repeat the test a few times before judging how reliable/repeatable it is anyway.
- 12-13-2011, 09:17 AM #18
Does the log stay or does it get deleted after the PB shuts down?
Following your 10 step program now. I'm at 1.94 watts and have been evertime I look at it. The graph shows some spikes in power comsumtion up to around 2. Its been 8.5 hours and I'm at 5% battery left.
- 12-13-2011, 11:54 AM #19
The log is maintained in the app's sandbox data folder, indefinitely, unless you've set it up to auto-delete any of that information at startup. Check the top-swipe menu settings in Battery Guru to see. By default it will automatically purge data older than 30 days but you can change the settings easily, including disabling auto-purge entirely.
I should note that my OS version for the above test is 184.108.40.20670. I'm running the same test now on a 220.127.116.1169 unit.
I also saw little spikes, as seen in the screenshot (small bumps on the black line in the graph). No idea what they are yet, other than maybe some periodic housekeeping the OS does. Looks like they're once per hour, almost like clockwork.
- 12-13-2011, 12:28 PM #20
I never looked at that log before. What's interesting is my health went from 86% to 88% in the past 2 Months. From 86 to 87 from what looks to be a long time on the charger after it hit 100%. Neat to see you usages patterns all layed out like that.
My graph looks almost identical to yours. I think I will keep that app running more often. if you want to take a look at the log let me know where to send it.
Thanks again for a great app.
- 12-13-2011, 08:48 PM #21
So on my second unit here, which is actually my original from May, now running 18.104.22.16869, I got this:
Health before: 89%
Recharge count: 154, went to 155 while draining during the test
Duration (100% to shutoff): 9 hr 19 min (with the last 32 min at 0%)
Average power: 1.977W (calculated over the drain-test period in the export file)
That means it consumed 18.42Wh during the test, which means health is really about 92.2% (9.32 * 1.977 / 19.98). Will be interesting to see if a 3% difference is enough for this to adjust upwards after a day or two.
Also interesting to have about a half hour "reserve" on both units, which are fairly worn-in by now. It could be RIM's deliberately tweaked the settings to target about 30 minutes of reserve capacity, to avoid people getting annoyed by running out just as it hits 0%. Or it could be a total fluke... need more people reporting numbers to have a clue.
- 12-15-2011, 12:50 PM #22
Two more data points, on new units from the promotion:
Health before: 97%
Recharge count: 0, went to 1 during test
Duration: 11 hr 13 min (last 2 min on 0%)
Average power: 1.665W
Health after: 95% (went to 96% during draining, then was 95% by restart)
This (1.665 * (11+13/60) / 19.98) is 93.5% health, so not far off and it obviously was adjusting in the right direction during the test. Surprising to me that the power is so much lower than on the other units I did before.
Also a 32GB:
Health before: 97%
Recharge count: 6, went to 7 during test
Duration: 11 hr 5 min (last 28 min on 0%)
Average power: 1.684W
Health after: 91% (went to 90% during draining)
This is (1.684 * (11 + 5/60) / 19.98) a health of 93.4%, so it overshot in adjusting during this test. This unit was also surprisingly much lower power than the first two I did. No idea what's different about them, unless there's something flawed with my test (which is the point of doing this exercise... to try to ensure the test is robust).
Neither unit has been used for anything much yet... in fact other than the initial update of the OS, and sideloading Battery Guru, neither has been used for anything. I'd be surprised if that explained it ...
- 12-15-2011, 01:13 PM #24
As best we can tell, the charger will basically bypass the battery once it's fully charged, and provide all the power directly to the PlayBook for operation.
That implies that it's perfectly safe for the battery if you just leave the thing plugged into the charger all the time, as it is not charging the battery the entire time.
Generally speaking, that would probably be considered "good" for the battery.
That said, the literature (depending on what you read) suggests that Lithium-Ion batteries may have a longer life if you don't have a full charge on them all the time. It's unclear, because other literature suggests that the reduction in life comes more from the number of charge/discharge cycles that you do. That means it's possible that having the battery sitting at 100% (or ever-so-slowly draining) while you have the thing on the charger all the time may reduce the life relative to if you could have it sit at, say, 40% under the same conditions. Doing that is simply not an option, however, unless you've rooted your PlayBook and can figure out how to muck with the charging algoriths.
I'd say until someone conclusively shows it's just one or the other effect that matters, assume it's both, but ignore the whole issue. The difference is going to be negligible anyway... the only way to maximize the life of your battery is to not use the product, leaving the battery around 20-40% and sitting on the shelf, recharging it to that range every half year or so as it needs it.
That's just silly though, so ignore everything here, including your own question, and sleep better. Just use the product with no concerns that your mode of use is doing anything particularly bad to the battery, and maybe keep an eye on things using Battery Guru. ;-)
- 12-15-2011, 01:28 PM #25
Using the BB while charging should be ok, as most of the current should go to running the device rather than to the battery, making it less of an issue how intelligent is the charging circuitry. But, I would not leave the BB plugged in while storing it for long periods (i.e., days) for a variety of reasons.