Computer Barrery Math not adding up

[Nnokalad]

My laptop, (Early 2014 macbook air) has recently started running out of battery faster than normal. So I did a few tests on it.

I ran a few tests, trying to limit my computers use of hardware,
eg. CPU maxed at 2% usage, low strenght backlights, no network signals.

The average time it took my laptop for a "complete" discharge was around 6.6 hours, and the average dischare rate was around -4.780W.

My battery has a full charge capacity of 44.750 Wh.

Wouldn't it make sense that my laptop at a average discharge rate of -4.780W should last around 9 hours?
What is causing this almost 40% difference ?
Would it come from heat?
Throughout the test the laptop barley went above 35C , so im not sure.

Thanks!

 

[itsmoked]

Heat is ultimately where all the battery goes.

A problem you are probably having is that as a battery's capacity leaves due to age it's internal resistance goes up. That means more and more of the battery is consumed IN the battery on the way to trying to get out to your load. Reducing the load to see how long the battery will last is misleading to you since you want the battery's capacity at the load you need to use it at.

I'm not clear how you're measuring the discharge rate. How did you get -4.78W?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[IRstuff]

I think it's because the OP knows that a realistic loading would make it hot, and make it worse. Heat and excessively faster charging are two known things that degrade battery life.

Your 2% CPU loading is highly unrealistic, so the battery probably lasts no more than about 1 hr under more normal conditions.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529

 

[Nnokalad]

Aghh... This is my second time posting this, the first one didn't work properly, and then I tried again, but ended up deleting it.

I have tried to make the reply easier to read, so Iv'e made my main questions bald.

A problem you are probably having is that as a battery's capacity leaves due to age it's internal resistance goes up. That means more and more of the battery is consumed IN the battery on the way to trying to get out to your load.

So pretty much my battery has degraded and it has an increased internal resistance. According to the test above, It draws 40% more power than "required" because of its high resistance. Therefore when I see a discharge rate of -10W it is actually drawing around 14W from the battery, but the extra 4W is dispated into heat. This was the cause of my inaccuracy.

Although this may not be entirley correct, can you confirm that I have a basic idea of what is going on?

Reducing the load to see how long the battery will last is misleading to you since you want the battery's capacity at the load you need to use it at.

Im not too conearned about my batteries "realworld" performance, I just used thoes, as I already recorded them during a previous experiment. I know a batteries performance on a small workload wont be equal to one of a "normal" workload, if that is what you are trying to say.

As for how I aquired the discharge rate, I have downloaded a displayes it using the sensors inbuilt on my battery. You can find it here:

https://www.hwinfo.com/

If you think that -4.7W is too little for a laptop, just note that I configured it to EXTREME power saving mode, taking ridiculous precautions to ensure I got the most battery effiency.

Heat and excessively faster charging are two known things that degrade battery life.

What do you mean by excessively faster charging?

Do you simply mean like charging it at a different speed?
If so, would that still apply if I have the same charger as others who are experiencing significantly better battery life?

Your 2% CPU loading is highly unrealistic, so the battery probably lasts no more than about 1 hr under more normal conditions.

Well the battery lasts around 6 hours for my average daily use. The CPU tends to be the biggest culprit in power consumption of computers, however I dont use alot of CPU bound prgrams. If I was to stress all parts of the CPU, it would last about 1.5 hours, and thats assuming all other hardware components remain at 0% utilisation.

Please dont be afraid to correct anything I have said.

 

[IRstuff]

It's been recently reported that the faster charging of Li batteries in smartphones, etc., tends to degrade battery life, since they get very hot during rapid charging. Therefore, to maximize battery life in your application, you should make sure that the charger charges over as long a time period as possible to minimize the charging stress on the battery.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529

 

[itsmoked]

Confirming your first statement above: It is correct.

Most laptops are now very miserly with their CPU power consumption. I have a very powerful high end gaming laptop with a quad-core CPU and it actually has a speed/power knob I can use to directly alter the CPU's "activity". What I've found during times in the field where I have to conserve is that turning that knob to the slowest results in only a small savings in the battery run time.

What really makes a HUGE difference is the screen backlight. When I want to extend the run time by an hour or two I dim my screen's backlight by one single dimness step. I normally run at 3 steps (out of about 10). If I cut that to 2 steps I add an hour if I cut it to one step I add almost another hour.

What this tells me is that the maker (Samsung in this case) has done a pretty good job of minimizing the CPU's power consumption already, so those four settings make little difference. What they can't do anything about is the back light. That's fully up to the user. The backlight uses more than the CPU. Same with cell phones the absolute biggest power draw is the backlight.

Thanks for the link to the battery measurement SW.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Nnokalad]

Thanks so much for the information, it REALLY helped!

I also was reading that a li-ion battery, will be unusable after a COMPLETE discharge, and computers will shutdown before they reach this stage.

Using the same HW monitoring program, I can see that my battery has a 51Wh "designed capacity".
My "full charge capacity" is roughly 45Wh. There is a difference of 6Wh between the two.

Is there a way to measure what percentage of the gap in capacity is caused from, natural capacity imdedence of the battery, and what percentage is still working but can not be used to ensure my battery never reaches zero. Or is the 51Wh capacity not including the space to ensure I never reach 0, and thus my battery dies.

I hope I was clear enough. If not, let me know and ill try an explain it better.

Also, internal resistance lowers the voltage right?
And are you sure that the internal resistance can impede the charge duration by 40%?
Not to doubt you, it just seems like a huge amount.

 

[IRstuff]

The internal resistance is only part of the problem. Battery University is usually a good place to go to find information:

http://batteryuniversity.com/learn/article/how_to_prolong_lithium_based_batteries

In any case, while there's no memory effect as with NiCad, that doesn't mean that Li batteries don't have their own life limits. While there's some desire on the part of manufacturers to improve performance, that is presumably tempered by the realization that improved longevity reduces future revenues.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529