Battery depletes quickly on Thinkpad E580

Hello all,

I have noticed that running youtube, the Thinkpad only last for 60 minutes.
Just going idle with wifi on and network on it starts with 2h30 but It quicly looses juice and finally I has only 2h of up time at max.

I did probe the laptop: probe

What other info can be useful?

Edit with power consumpsion:
IDLE:
Code:
$ acpiconf -i 0
Design capacity:    45000 mWh
Last full capacity:    33540 mWh
Technology:        secondary (rechargeable)
Battery Swappable Capability:    Non-swappable
Design voltage:        11100 mV
Capacity (warn):    1677 mWh
Capacity (low):        200 mWh
Cycle Count:        428
Mesurement Accuracy:    95 %
Max Average Interval:    1000 ms
Min Average Interval:    500 ms
Low/warn granularity:    -1 mWh
Warn/full granularity:    -1 mWh
Model number:        01AV445
Serial number:         1553
Type:            LiP
OEM info:        LGC
State:            discharging
Remaining capacity:    34%
Remaining time:        0:40
Present rate:        17471 mW
Present voltage:    10798 mV

Youtube video running in background:

Code:
$ acpiconf -i 0
Design capacity:    45000 mWh
Last full capacity:    33540 mWh
Technology:        secondary (rechargeable)
Battery Swappable Capability:    Non-swappable
Design voltage:        11100 mV
Capacity (warn):    1677 mWh
Capacity (low):        200 mWh
Cycle Count:        428
Mesurement Accuracy:    95 %
Max Average Interval:    1000 ms
Min Average Interval:    500 ms
Low/warn granularity:    -1 mWh
Warn/full granularity:    -1 mWh
Model number:        01AV445
Serial number:         1553
Type:            LiP
OEM info:        LGC
State:            discharging
Remaining capacity:    28%
Remaining time:        0:30
Present rate:        18887 mW
Present voltage:    10617 mV

Kind regards,

Charly
 
You probably need to enable and configure powerd(8). I'm betting the laptop is just always on "full" blast, never throttling down when it can.
 
I already configured powerd in rc.conf:

Code:
powerd_enable="YES"
powerd_flags="-a hiadaptive -b adaptive"
performance_cx_lowest="Cmax"
economy_cx_lowest="Cmax"

Then it probably means wear on the battery or not the same battery quality as on the T series.

Could it help to have the drivers loaded for the AMD discrete graphics? I can not select the Intel GPU or the AMD one so I only loaded the Intel GPU drivers at boot.
 
Some laptops have a battery utility in the bios which can show useful stats about your battery.
I would start there and make sure its up to snuff.
Your problem could be there depending on battery age. As they age they lose strength.
 
Design capacity: 45000 mWh Last full capacity: 33540 mWh
Right there you can see some age.
The battery as new was 45000MWh and now laptop charging circuit will only charge it to 33540MWh

Cycle Count: 428
So your battery is roughly half way done. 1000 cycles and they become useless fast. (Depending on cells inside the battery pack). OEM packs can be much worse.
 
I can not select the Intel GPU or the AMD one so I only loaded the Intel GPU drivers at boot.
This Optimus stuff really draws down a battery fast on my Dells with it.
I also have an ACPI bug with them that it runs the fan full blast.
Maybe 30 min battery life.

So look around for IBM ACPI details. There may be a driver that needs loading.

What version of FreeBSD are you using?
 
Can you chuck in a Linux live CD or boot Windows and see if you experience similar?

It might be aging hardware more than FreeBSD doing something wrong.

As Phishfry suggested, a # kldload acpi_ibm might help for Thinkpad specific power management.
 
Well it makes sense with the battery cycle count, laptop from 2018 I think.
I have the acpi_ibm loaded already.

We can conclude that it is aging hardware like kpedersen suggested.

Thank for the info's and suggestions.

ps: never thought that freeBSD was the culprit, just me not configuring it correctly.
 
kpedersen suggestion is worth following up. Linux live memstick. Just for reference.

Like I mentioned our ACPI isn't always perfect and you are in the low spot with Optimus.

I mentioned battery in the BIOS. What you see is a conditioner or discharge cycle that resets the battery and you gain some MWh back. Not all laptops have it.

So its worth seeing if BIOS has anything to help refresh battery.
 
Phishfry is right: Your battery simply is not the newest one anymore.

I was working as an battery engineer for a couple years, specialized on power electronics such as protection ciruits and fuel gauges for lithium-ion batteries.
Comparing
Last full capacity: 33540 mWh
with
Design capacity: 45000 mWh
shows the real capacity C is down to app. 75%.
A roughly estimated rule of thumb within battery engineering declares 65..60% as the end of life for a battery.

Also 478 cylces shows the battery is anything but new anymore.
I'm talking full cycles; cannot say, what the acpi shows of the batterie's internal fuel gauge controller (mostly TI's are used, e.g. such as BQ40Z80, or Maxim) if it also counts partial cycles. 1k would a bit much for a battery used under real load conditions, but app. 600...800 almost is what you may estimate as a cycle life time for a good brand's LiIon-battery in a well engineered battery pack for those kind of load types like computer applications.
LGC is not bad, but it's neither Sony, Panasonic, Sanyo or Samsung which may be named as best quality brands for lithium-ion battery cells.

Unfortunately the older the battery the faster it ages.
But face it:
The times are still good for a laptop.
2.5h are the normal average design usage time under low to medium load conditions for a average laptop computers.
Playing videos, especially in higher resolution, plus browser with today's browsers and multimedia websites are more kind of above medium if not heavy load 🧐
(A good sign to estimate load conditions is to listen if the cooler fans shift gears.)

P.S.: Don't trust given fuel gauge values to the exact mAh/mWh. Battery fuel gauging is a complex art for itself, and anything else than a precise science.
10% measurement error is very good, 5% is exemplary precisely (Even if vendors advertise with 1% or 0,5% - but to explain that, I'd need several pages :p.)

[smart-ass="ON"]
Just because a computer tells you 12 post decimal positions does not mean even the positions before the decimal point needs to be exact! 🤓
[/smart-ass="OFF"]
 
I was working as an battery engineer for a couple years, specialized on power electronics such as protection ciruits and fuel gauges for lithium-ion batteries.
Your experience might help here. In your opinion how far are we from a battery that doesn't degrade? Is it all a trade off of expense to effectiveness or are they artificially set for planned obsolescence?

[smart-ass="ON"]
Just because a computer tells you 12 post decimal positions does not mean even the positions before the decimal point needs to be exact! 🤓
[/smart-ass="OFF"]
Actually this is similar to my pet peeve, when a i.e scientific article gives an approx distance in meters (such as 1000) and then another article takes the liberty to convert it to feet (for the English readers) only to end up with some overly precise figure (3280.84) even though the data was based originally on a less accurate approximation. :D
 
Your experience might help here.
Thanks. But frankly instead of batteries from over 15 years ago beside mobile phone and power tools today's portable computer batteries are of the best engineered anyway (at least if you don't buy any cheap total cr#p, of course.)
In them there are built in the most comprehensive and best precise battery electronics available (SMB/I2C -> all data and statistics available via ACPI or else... you simply cannot have more.) And as I said, the fuel gauges are with 1% or better as datasheet, 5% in reality best possible anyway.

artificially set for planned obsolescence?
I cannot preclude it's not done by one brand or the other, but to do this by the battery would be way more engineering effort and less controllably and cause more unsatisfeied customers as to do it by make the software "age", such as raising memory/CPU consumption, not supporting drivers anymore, or what have you.
Over 15...20 years ago the batteries were simply badly engineered. Some may also remember the first cell phones we had in the 90s with those real junk batteries...
Those days are over.
You need to see, it's always a compromise: How long should the battery last for one cycle - How much weight the customer tolerates? Something like a modern version of the IBM portable PC may not hit the main market today anymore 😁
How much is the customer willing to pay for (keep in mind that LiIon battery cells cost app. 3..5$ each @ >1M lots, therefor besides CPU and RAM they are the most expensive components for mass production.) How long is the average time of usage until a customer is buying the new model anyway (2...3 years)?

Best advice I can give you in this point is a truism you can put together yourself:
If you really depend on working much on battery power, you need to set this on top priority judgement for your next purchase.
High quality brands are seldom fool you with their specs - at least in this point. You just need to keep in mind that those maximum values of course are mostly given for a state when your computer nearly sleeps, not "bulding world" in the background while playing the most current 3D shooter at maximum resolution on VB... :p
Also take care, config your system and use it as low energy consumptious as possible.

E.g. for my part I know from my own experience I don't really use my laptop running on battery much. I actually need the battery only for small power interruptions such happens when I travel by train - capacitors would be sufficient for my style of use :cool: But I wanted hardware power so I purposely chose a "Gaming" Laptop (ASUS TUF FX505DY), which on battery runs out on capacity about 30 minutes. No problem for me, since I know what I need.

You also may keep in mind that there are also additionally choices, like buying extra battery packs you can carry with you charged (maybe too expensive, and not exchangeble while the system's running), or there may also external high capacity batterie power supplies you may plug in instead of the AC adapter.

In your opinion how far are we from a battery that doesn't degrade?
Well, nothing can never degrade. Every real system has a limited lifetime.
But for over app. 10 years there are already battery cells available that are really very longlife: FePO4
Those last at least ten times longer than common LiIon cells - BUT....
Point is, their capacity is way less (app. 40...50%), because those are high current cells (app. 10...20 times higher currents), build for more power, for drives actually.

With batteries is always high capacity versus high power - you cannot have both. Well, of course you also can get more power from a high capacity cell as it's rated in their specs (>200...400A no problem - if you need it for less than 0,01s 😁), but then you pay for it with a loss of cycle life time.

Within mobile computers (notebooks, smartphones, etc.) you prefer high capacity cells (>3000...8000mAh @ 0,5...2A [high current cells: 1200...<3500mAh @ 5...20A])
You could build a laptop computer with FePO4 cells... but that would at least double the volume - and weight - for the battery pack just to get the same capacity (use time per charge), but the battery would last easily 5...8 years, if not longer.
But you also could increase the lifetime of a LiIon battery pack by increasing its capacity (twice C ~= twice lifetime).

But then we are back to the same question again:
Who would buy such a big club?
respective in times when companies and users wanted new hardware every 2..3 years anyway...

Sorry for that long post, but I hope I could answer the one or the other question.
 
I am in batteries engineering for over 15 years. I also can tell you within the specialists magazines app. every half year "THE magic" battery cell (chemics) is announced, revolutionizing the complete battery market, extreme capacity, very high currents, no self discharge at all, extremely long life, absolutely safe.... no disadvantages at all, advantages only.... and then, you'll never hear from them again.
The only cell (chemics) I observed that was actually realized within my 15 years in this field is FePO4
And those eclosed also not to be the holy grail solving all our problems with one solution. 😁

As always in life it's to chose the right compromise between advantages and disadvantages for your very purpose.

The possibilites are widened by LiIon and FePO4 extremely, of course. And it's very good that those possibilities are used.
I may be (hopefully ;)) completely wrong, but in my eyes with LiIon we reached the end of what is physically possible with galvanic energy storage. There will be improvements, of course, but do not expect a real great revolutionary leaps in batteries anymore.

You see, there are two crucial - physical - points about this:
The pairing of elements with as much difference in electronegativity as possible, resulting in the voltage.
And the amount of surface area you may pack into a specific volume - how many molcules take part in the electrochemical reactions (stored energy).

With the first there still is some potential - in theory.
The idea of lithium based rechargeble batteries is way over hundred years old. The reason why we have those cells available for mass production only for twenty years is the capability of handling lithium for such a purpose. For higher differences in electronegativity we are not talking about lithium anymore, which in its elemental form is very highly inflammable by any contact of oxygen at normal conditions - we are talking about metal burning, > 2000°C. Then we are talking fluorine and such "funny" elements.

The second point - the area of the surface, how much chemical stuff is paired over the electrolyte - decides how much capacity can be stored within a volume.
Herein lies the contrast of capacity (Ah/Wh) versus the robustness of a cell. The more surface the more capacity, but also the more touchy, damageable the cell becomes.

Until 2007...2009 the trend of cell's manufacturers was to reach even higher cell's capacities. Sanyo even had a 18650 (standard laptop cell) with outstanding 14Ah.
But it never came available on the market. Why? Because that cell was way too dangerous to put it in the sticky hands of little chimps. 😁
Because there was a significant raise of accidents by LiIon battery packs worlwide at that time (Some may remember the case of the sudden smoking laptop in a lecture room (Japan or USA?), smartphones catching fire in user's trousers, cargo planes catching fire and much more. Not funny at all!
And it always was brought back by media to the origin of the well known cell's manufacturers, even they were not responsible for most of the accidents. So they decided to put priority on safety, not to say foolproofness, instead of more and more powerful cells.
They actually reduced capacity again just to ensure the cells will not catch fire or even explode when treatet badly by silly monkeys.

In two cases I know the cell's manufacturers actually had a quality issue. Sony had brought cells with polluted chemicals making them self-inflammable in consumer products on the market. And in another case Matsushita (Panasonic) "only" flared off their own cell production plant in Japan (very big loss for the whole batteries market for a couple of years.) But those were exceptions, and both companies reacted quickly and intransigently, and also started to audit all of their customers to increase safety.

For nearly all accidents caused by lithium batteries the companies are responsible who carlessly cobble-whacked the packs (I cannot call that "engineered", even not "assembled", maybe "lovelessly nailed together" at most - this also ain't not funny, really, it's actually bizarre what some companies fire on the market and burden their customers with [Always excluding mobile computer companies and (quality brand) power toolers in this point! They do know their business. There may some minor security issue with a polymer cell in a single smartphone, but the really dangerous times are long over. They know what they risk. They know what they are doing. And they doing it right. No serious issue here.]
I am talking about all the companies who deliver the battery packs for any other company who decided to produce something battery powered. They not seldom actually ignore mandatory security measurements such as protection circuits or UN transportation test - costs money, makes effort, kills time schedules, dispels customers, they buy where are no such fancy, useless stuff is sold, so skip it! I do not tell made up jokes! I could finger companies by name. That's what really happens, thousands of times every day. You just don't read that much in media, because it's not interesting if not somebody's balls are toasted by a smartphone. Also not few engineers really believe they could do protection circuitry with MCUs (It's a bad idea. Don't do it! But that's another story one could bang his head against the wall and not to be told here.)

However, back to our real topic:
So what we see today is - as I think, electrochemical, galvanic energy-storage - batteries - are already nearly at their highest peak possible.
There will be some improvements of app. 10... maybe 50%. But as I said, don't get your hopes up and expect great, revolutionary leaps anymore.
I don't see it.

So, bottom line:
Batteries are opening possiblities, no question.
But you also have not only to decide, which type of battery you use for what application (lead, ltihium, FePO4, whatever...), you also always have to carefully think of, if it's even a good idea to use a battery anyhow and how.
Especially when you respect the bad effiency factor a battery has
(max. 80% for a high quality new lithium ion. The battery of the start of this thread may have something about 60% if not even less around 40% [it's the quotient of how much Ah you discharged and Ah charged. You may already have observed that charging lasts longer, and discharge [usage] is shorter, the older the battery becomes. Not only the capacity is reduced by age {internal resistance rises because of molecules are not taking part at the electrochemical reactions anymore} but also the efficiency factor drops rapidly)
you need to keep in mind that a battery by nature it's a bypass technologie, small jumper storage, or for some kind of niche only, but nothing for main purpose such as battery only powered cars - what in my eyes is a complete nonsense bogus. 😵 But, that's also another topic not to be discussed here in this thread. (I hope, I did not really killed it, and charly got his answer for hir original question 😇)

Enough battery engineering lessons for today.
I don't only write ([too] much) - I also read you and learn by you :)

Just always keep in mind:
There is no single solution that solves all problems.
Complex problems only can be solved by multiple solutions. (Law of nature)
If anyone announces the one-solution-solves-all-problems, it's always a markting trick. There is nothing, that has advantages only. If that appears to be, someboday only wants to sell you something, because people want to believe there will be a jack-of-all-trades (would be so easy), and most always buy the next one again, if it's offered 😁

But who I am telling this here?
Of course FreeBSD users are fully aware of that, and don't need to be told. :cool: 😀


P.P.S.: Attached a picture from my personal archive to emphasize:
I know what I am talking about. Lithium-Ion batteries needed to be handled with respect.
If engineering is done carefully and correctly, nothing much can happen really.
But if as in many cases "engineering" is just understood as "roughly estimate capacity, attach a wire with a plug at it and ask who can deliver fastest for the lowest price" only, than such could be the result very quickly - lucky if like in this case only a warehouse flared off without any damage to humans or animals.
 

Attachments

  • IMAGE_052.jpg
    IMAGE_052.jpg
    377 KB · Views: 61
I'm pretty sure you'll quickly find lots of stuff about it in the internet anywhere, but here are some
Tipps for a long battery usage time per charging load:

Above all, before you even start:
Define your needs exactly - what are you going to use your laptop/notebook for, and how much hours per day you really need it running on batteries only - what needs to be done while on batteries, what can wait until net power is available?

Based on that you could make a better purchase decision for your mobile computer.
Do you really need a 8Core-CPU @ 4,2GHz or would a 4Core @ 2,2GHz also be sufficient?
You may exchange the drive with a smaller, maybe slower SSD but having less power consumption.
Do you really need a >1...2TB disk in your laptop? A 128...256GB drive can easily more than sufficient for FreeBSD. You may save the large amounts of data elsewhere and make them available only if you actually need it.

Try to distinguish what you like, nice to have, from what you want, really need.
If you find you cannot or will not afford a measurement, then you don't really need it.

What is necessarily need to be done really on your mobile maschine, and which work you can source out on your desktop/tower/server or be done at another time?
Keep in mind that by their origin nature laptops/notebooks are ment to be a kind of a second machine, not the primary one. ('m aware of today many people only have a mobile computer. [if not yet known by now, you may be interested in so called docking stations; IBM's ThinkPads really had a clever solution (I'm not sure, but I think Lenovo still offers such.])

Save as much power as you can.

Use powersaving options supported by BIOS and OS ("Adjust CPU dynamically if supported", ErP, and what there is.)

Only run what you really, currently need, and switch anything off you don't need.

RGB lighting... does nothing productive, just looks and consumes current - switch it off!
Some may say:"but this don't cost much... RGBlighting is just 10..20mA"
Yeah, well, but it adds up! 5mA here, 20 there..... especially over time!
That's something you become sensitive of if you'r dealing with batteries on an engineering level:
"10 mA ain't much!"
"No, but 10 mA over 10h are 100mAh." And - oops - quickly 200..800mAh of your battery's capacity is wasted, you could have saved easily.
(It's a classical battery's engineering failure to accept more than µA stand-by consumption of the battery's electronics.
1000h are 41 days, a bit more than a month. Now you see, why many batteries are deep discharged (and seriously damaged, junk) in springtime.
Because the engineer thought: "5mA ain't much" and of course the customer did not fully charged the battery again before it was stored in autumn in the garage.)

It may not strike some, but did you see in charly's very first post the current wattage his ACPI reported?
Present rate: 17471 mW

Present rate: 18887 mW

There easily is a fluctuation of 1.4W just in this example.
Can you imagine how much longer the battery could last if a machine would be brought from 18W down to 12W?
Right. 33% longer. By an average usage time per battery charge of 3h this would gives you 4h.

(Lower consumption also make the battery's cycle life lasts longer.)

Bluetooth, WLAN and especially GPS are real juice suckers. Not on few computers they are switched on by default, although not used, and many users are suprised how longer one battery charge lasts by only switching those off, if not actually needed.

Get rid off all fancy, but useless stuff, or at least switch it off while you're running on battery.
window's animations of your desktop environment, animated menus,.... animated backgrounds (use a simple single-color black background),
...this all costs CPU power thus battery's juice.
Adjust monitor brightness,... (in that case Apple's new MacBook is exemplary. It automatically adjusts brighness and contrast, you don't even realize it's done if you change from indoor's shadow into bright sunlight outdoors - still very good readable monitor. Nice. (But still no reason to buy fruit 😁) [....something like this may also could be done on other systems, by using the camera as a sensor...? Maybe there already is such a demon; or maybe I'll find the time, to check it out....])

If you use your laptop as a terminal only just to log on a server you may ask if you even need a X-server anyway.
Especially within turnkey systems (I'm not talking about such as windows or ubuntu only, I'm also talking desktop environments such as KDE, too) there's lots off stuff built in for automatically fit into any situation and also just to make the system look cooler, only.

Especially under a very individually controllable OS like FreeBSD you may check out, what your system starts while booting (especially for power saving laptops an own compiled kernel could make sense.)
Throw anything over board, you don't need. Only start it, when you need it and switch it off immediately, when you done with it.
E.g. you may check out if or what disk/filesystem maintence services can be scheduled to when the laptop is plugged at net power again.

If you look hard enough you'll find lots of thing you don't really need or can be switched off/removed at least for battery mode.

Try to get your machine as light as possible. Strip your system down to the bone! But most of all, test yourself about any just optical or convinient but actual useless overweight you're carrying.
If you put some effort in it and dispense things, you don't really need, at least while under battery conditions, it's possible to have a laptop/notebook running on batteries for 6...8h, even more.
Try to train your system and yourself to have a special battery mode.
 
I'll read this information multiple times.
And see if I can indeed reduce some watts that my system is using.
Thanks for sharing all this knowledge.
 
We have quite a few e580s at work (no battery problems btw), but one thing to look at in the uefi is, forgive me i can't recall it's exact description, is something like "fast charging while off". This allows the laptop to charge a device (phone etc) via usb-c when off and it uses a lot of power even if not actually being used to charge.:eek:

Also make sure your bios is the latest because lenovo did do a bios update for battery life. At least as far as i recall.
 
We only have Lenovo, so I can't speak to HP. It's a common issue with these so I suspect it's also common with other laptops that provide the ability to charge a device while the laptop is off or in a sleep state.
The constant power usage to "poll" the USB-C will add up over 24 hours and then days until it will deplete the battery. This has been seen on many occasions.
So, I would check all laptops for this "feature" and disable it, if possible via the BIOS/EFI.
 
Just curiosity... what Desktop Environment are you using right now? Or are you using a Window Manager?
Edit: What software are you using? Excuse me if I get into your private things. Just check the consumption installing htop If you don't understand what it means, post a capture.
 
An in-depth article/tutorial on power saving …

Manual pages for the article by vermaden

Online at www.freebsd.org:
execsnoop(1) from sysutils/dtrace-toolkit:
  • nroff -man /usr/local/share/dtrace-toolkit/Man/man1m/execsnoop.1m | more
  • or ln -s /usr/local/share/dtrace-toolkit/Man/man1m /usr/local/share/dtrace-toolkit/Man/man1 (assuming no interference with other pages in section 1) and add a .conf file to /usr/local/etc/man.d/ comprising the line below
MANPATH /usr/local/share/dtrace-toolkit/Man

Not found:
  • acpi_call(4) – see below
  • acpich(4)
  • lscpu(8) – see above.



Code:
% pkg info --list acpi_call
acpi_call-1.0.1_1:
        /boot/modules/acpi_call.ko
        /usr/local/sbin/acpi_call
        /usr/local/share/licenses/acpi_call-1.0.1_1/BSD2CLAUSE
        /usr/local/share/licenses/acpi_call-1.0.1_1/LICENSE
        /usr/local/share/licenses/acpi_call-1.0.1_1/catalog.mk
%
  • no manual page.
FreeBSD bugs:
 
Back
Top