Help Battery app with correct percentage indicator?

[Buckeye Boy]

all these apps have battery life in increments of 10%. do any go by 1% ? for example, phone shows 82% battery life and not 80%...

 

[Deleted User]

Best I can tell, since the system only reports battery life in 10% increments, any app that does anything more granular is an estimate at best.

I am using the Battery Left widget. Unfortunately, my usage patterns don't seem to provide any kind of good calibration for it, so its granular % indicator is typically 20% or more less than what is actually left. I spoke with the creator of the widget and it's just because Battery Left's algorithms try to get a hold on the estimated TIME left based on my historic usage (i.e. that's going to be a rough and variable estimate), and the percent reflects the amount of that time, not the percent of the actual battery.

So... long story short, please post if you find a good one - because so far I haven't found one. I like the Battery Left widget and am pleased with the developer's responsiveness, it just doesn't do exactly what I want it to.

 

[malek9696]

Battery Left does. However, that app is not the most accurate. It monitors your battery life one time through, and if your battery happens to last longer or shorter than that first time through, Battery Left will be way off. I've had Battery Left tell me I had 0% left when the system said I had 50%.

 

[Deleted User]

Battery Left does. However, that app is not the most accurate. It monitors your battery life one time through, and if your battery happens to last longer or shorter than that first time through, Battery Left will be way off. I've had Battery Left tell me I had 0% left when the system said I had 50%.

You can set the app to continuously calibrate (there's a checkbox that says "calibrate until accurate" - if you clear that, it won't stop calibrating once it gets a single run-through), but unfortunately, it's still not particularly accurate for me. Unlike my first run-through, it no longer registers 30-40% off, but if it's even 10% off, I might as well use the stock estimates.

Right now, I'm using it as a nice looking battery widget that also has a temperature sensor, and having it show me the stock battery estimate.

 

[racinwarrior]

i have been thinking about this problem since launch day.

is it possible (im not a dev so i dont know) to take the voltage of the battery and relate that to a percentage of life left? you should not have to calibrate it ever since its reading voltage. if voltage is X1-X2 then you have Y% left

ie. 3.822 - 3.820 Volts is 82%

 

[ASC]

i have been thinking about this problem since launch day.

is it possible (im not a dev so i dont know) to take the voltage of the battery and relate that to a percentage of life left? you should not have to calibrate it ever since its reading voltage. if voltage is X1-X2 then you have Y% left

ie. 3.822 - 3.820 Volts is 82%

Did some quick research here: Battery State of Charge Determination

And found this:
Problems can occur with some cell chemistries however, particularly Lithium which exhibits only a very small change in voltage over most of the charge/discharge cycle. The following graph shows the discharge curve for a high capacity Lithium-ion cell. This is ideal for the battery application in that the cell voltage does not fall appreciably as the cell is discharged, but for the same reason, the actual cell voltage is not a good measure of the SOC (State of Charge) of the cell.

 

[racinwarrior]

Did some quick research here: Battery State of Charge Determination

And found this:
Problems can occur with some cell chemistries however, particularly Lithium which exhibits only a very small change in voltage over most of the charge/discharge cycle. The following graph shows the discharge curve for a high capacity Lithium-ion cell. This is ideal for the battery application in that the cell voltage does not fall appreciably as the cell is discharged, but for the same reason, the actual cell voltage is not a good measure of the SOC (State of Charge) of the cell.

hmmmm ok makes sense i guess.

however the default os must read some sort of variable to determine SOC. (maybe current)

we need to find out what variable its reading and change the resolution

 

[Napalm]

I would like the same thing for various testing purposes. However I believe that has long as the system only reports in 10%'s then that is as granular as you will get. Even rooted.

It could be a limitation on the system itself.

 

[ASC]

After reading the article a little more, which I think was meant for electric car batteries and not cell phone batteries, there are just way too many variables in the mathematical algorithms and calculations for it to be practical to get all that accurate for a cell phone.

Just some of the variables that need to be calculated are:
Battery state of health
battery temp
ambient temp
open circuit voltage
charge and discharge rates
cell ageing
self discharge
and on and on....

But I get what you're saying about taking the measurements of whatever they are reading in smaller increments to get the reading in 1% increments and not 10% increments.

LOL, Way beyond me....

 

[racinwarrior]

well i guess i should be able to figure this out. im an eletrical engineer after all. sometimes the simplest ways are the best ways. it seems to me that the os already has to account for all of these variables in order to report any battery percentage at all. i would think that a dev should have access to all of the system information needed in order to create some sort of calculation to determine the SOC in 1% increments.

then again it may be very hard to do and most likely never happen

 

[ASC]

I agree that they should be able to report the percentage in smaller increments, just add some more 000 place holders or something, after the decimal point and figure the extra data into the calculations.

But on the other hand, they don't necessarily "need" to account for all the variables to get an output, it just wont be a very accurate output which is probably one of the issues here.

In the article I read (Skimmed ) they were saying that to get "Gas Gauge" (Battery indicator) on an electric car to be accurate, just the battery monitoring system alone would need its own microprocessor that would be continually taking measurements of all the variables and continually calculating and analyzing the results. (Wouldn't want to be stuck in the middle of the Arizona desert in the middle of summer because your car told you it had 25% battery left then all of the sudden died )

But on the other hand.... If they spend the time and money to figure it out for electric cars, then the technology could eventually make its way over to more accurate cell phone battery indicators.

Doubt they hired a bunch of physicists to figure out the algorithm for the Cell phone software when they were writing the code for the battery indicator.

 

[racinwarrior]

i doubt it too

the problem is (very annoying to me) that it has been done before on countless phones. phones that like the x also run android. could it really be that hard to give us 1% increments. come on moto

 

[ASC]

You're right, If they did it on other android phones, then they should be able to do it on the X. Probably a Motoblur limitation rather than a Android limitation. And with that being said, hopefully it can be changed easily in the future.

 

[Martimus]

I guess it's a matter of personal preference... I, for one, see no value in having it burn up the battery that much quicker by making much more frequent checks of the battery level. It would be nice, however, if they gave us the option to choose for ourselves whether or not we wanted that level of granularity.

 

[iresq]

If it only sees in increments of 10, how does it report and shut down when battery life hits

 

[Napalm]

You're right, If they did it on other android phones, then they should be able to do it on the X. Probably a Motoblur limitation rather than a Android limitation. And with that being said, hopefully it can be changed easily in the future.

My point exactly. I doubt its a motoblur issue as much as its a hardware issue, IE the mainboard/processor doesn't sample or even attempt to report more than 10%'s

Course does anyone running a rooted DX have this issue?

I guess it's a matter of personal preference... I, for one, see no value in having it burn up the battery that much quicker by making much more frequent checks of the battery level. It would be nice, however, if they gave us the option to choose for ourselves whether or not we wanted that level of granularity.

Well granted for normal use I agree. For the most part I run a full day + without issue and don't care. However when comparing car chargers and application use on the DX I would like a little more info.

I mean really on a 20 minute drive, did I charge from 40 to 50% or did I charge from 37-54%. Don't know, would be nice to know so I know exactly how much I'm getting into the phone.

 

[deemedic]

I use Free Beer battery widget from the Market and I know it has given me readouts of 15 and 13% but I can't guarantee it's really down to 1% readings. Plus the meter looks like a glass of beer...can't help it I am easily amused. It's probably not nearly as technical of an app as you guys need but it works for me. My requirements were voltage and temp and it does that for me.