Battery draining with external power while under load?

[RndmNmbr]

I received a PBP a couple of weeks back, and generally I really like it. (Though I wish the code for the keyboard/clickpad controller was opensourced!  The trackpad lag makes the PBP “feel” a lot slower than it really is!)

I’ve been doing some larger software package builds (with make -j 6) with the PBP and have noticed that when I have heavy constant load on the CPUs, the internal battery begins to drain even when the unit is externally powered via the the 5V 3A barrel connector.  Is this normal or do I need to try and track down some other issue here?  

I apologize if this has already been covered; I did a couple of searches and didn’t find any information that really matched what I’m interested in.  I’ve seen some comments on this type of behavior when powering external USB devices, but not with CPU load.

If this is normal, I’m interested in tweaking some of the charger circuit setpoints and seeing if it improves things. (With a new power adapter of course!)  Looking at the schematic I see some potential spots that may be of interest.  However, without layout art and a complete BOM, it’s hard to say how much margin there is to increase the draw without risking CCA damage.

[Syonyk]

That's a thing, yes. Turn your screen brightness down, or maybe only do a make -j2 and let the little cores idle. They don't have much L2 cache to share, so I'm not sure you gain a ton with them.

But if I start a kernel build with a fully charged battery and my screen at full brightness, I'm down to 50-60% battery by the time the build is done. I think it's just the nature of the power circuitry on this system - it can draw more than the charger can provide.

I mentally toss it in the "Yeah, beta hardware..." quirks category and figure that in a few revisions, USB-C Power Delivery will be used to solve the problem.

[RndmNmbr]

Thanks. 

I was seeing a similar ~30-40% battery drop while building OpenCV off of the internal eMMC.

You have interesting point regarding the A53 cluster.  Now I’m curious enough to try an experiment with -j 6 and -j 2 and see what the build time differences look like.  

[JasonG-FR]

I've experienced the same thing and it seems to be related to the temperature of the laptop.

While the PineBook is charging and under a heavy load, it can easily reach over 50°C. I noticed that not only the battery decreased under the load but then struggled to increase once idle. I checked the charger and it was cold, so it definitely wasn't outputting 3A. The battery started to charge after unplugging and plugging the cable back in.

Maybe it's some kind of thermal protection?

[granstr0m]

Mine is draining even without any heavy load which is a bit frustrating. 
If you totally drain the battery the laptop becomes unstable and unusable until you reached about 10% battery.
And charging from zero to 100% takes me about 2-3 days of non use connected to a poweroutlet which is insane.

I love this machine so far, got Kali working from the mmc. Keyboard is nice, the screen is super for this price. Touchpad is a disaster but I use a mouse. I would love to buy a whole set of these for pen testing purposes in my company BUT....

The issues I have with the battery is a bit of a showstopper. Is this expected behaviour or can I do something about it???

[Syonyk]

That's definitely not normal.

What are you charging from? Have you tried USB-C charging if you're on barrel plug, or the opposite?

[JasonG-FR]

Yeah that's definitely not normal, mine charges from 20% to 80% in less than 4 hours while using it lightly. If it takes several days to fully charge it when it's turned off then maybe there is something wrong with the power supply you're using?

Are you using the power supply that came with it from Pine64 or a USB-C phone charger? The PineBook Pro needs something capable of delivering 3A at 5V which is quite uncommon. It doesn't use Quick Charge either, I've tried to power it with a QC charger and the voltage was still 5V.

How long does your battery usually last when full?

I'll try to power mine with lower voltages than 5V to see if a how much it affects the current that the PineBook Pro can use to charge.

[Vinay]

Yeah, received my PBP this weekend and overall quite happy with it. Battery was completely full when I received it and life seemed quite good actually even when using it. It would easily make it through the day though when it had about 30% left I decided to charge it using the USB to barrel cable. Not sure what current my USB socket is able to deliver, it is one of these that delivers both the regular 220V (euro socket) as well as two USB outlets. Initially it charged nicely but near 80% it starts to level off. When I started Redshift (an open source kind of f.lux, something to tone down the blue from the screen for evening work) the charge actually dropped. Even with the lid closed, not doing anything at all. Stopping Redshift helped. Though it still doesn't charge beyond 88% or so using that cable. I made a screenshot with gnome-power-statistics of the charge state, for those interested. If so, please tell how to upload that picture. I'll try using a separate USB charger someday which I know that is able to deliver the 3A current and see how that works out. For now I'll quit charging with that cable through my USB socket and use the supplied charger instead, I'll report back how it works out.

Either way, of course I think it is still nice to be able to just carry that cable when going somewhere and be able to charge beyond 80%.

Quick question as we're here, what bash command returns the charging state. I'm used to using acpi -V (when using SuSE) but for now I'll stick with the supplied Manjaro and acpi does returrn something, but doesn't seem to read the charge state (instead it tells me I have a 0% charge). Is there an other command I should use? acpi wasn't installed by default so I get that Manjaro or Pine64 may prefer to use something else.

Thanks!

Update 1: Had it charging using the supplied charger since my original post. There was a (relatively) quick rise from 88% upwards but it now seems to have settled at 92%. I don't expect it to charge up to the full 100%. Question I now have, is this an issue for the battery health to not ever be charged fully? I think even with this battery charged up to 92% I'd probably be happy with the mileage it goes but I don't want it to cause excessive wear. I'm no expert in electronics (or anything computer related, probably) but I've read somewhere that it is actually the stages near full discharge and fully charged (where it constantly stops and starts charging) which cause most wear. But yeah, I would appreciate some expert views on this.

Update 2: With the lid closed and connected to the supplied charger, my PBP eventually charged to 100% ! We could probably still consider charging relatively slow but at least it does charge to the full 100%. Question remains, how much of an issue for the hardware is it to never/rarely charge to the full 100% (but stop between 80% and 90%)? Would this damage/wear the batter in any way or is it fine (or even better)? The practical adantage I can see is that I can use the much more portable USB to barrel cable and plug it into these (nowadays quite common) USB power outlets or powerbanks.

[JasonG-FR]

Quick question as we're here, what bash command returns the charging state. I'm used to using acpi -V (when using SuSE) but for now I'll stick with the supplied Manjaro and acpi does return something, but doesn't seem to read the charge state (instead it tells me I have a 0% charge). Is there an other command I should use? acpi wasn't installed by default so I get that Manjaro or Pine64 may prefer to use something else.

You can use this command to get the remaining capacity :

Code:

cat /sys/class/power_supply/cw2015-battery/capacity

Question I now have, is this an issue for the battery health to not ever be charged fully? I think even with this battery charged up to 92% I'd probably be happy with the mileage it goes but I don't want it to cause excessive wear. I'm no expert in electronics (or anything computer related, probably) but I've read somewhere that it is actually the stages near full discharge and fully charged (where it constantly stops and starts charging) which cause most wear. But yeah, I would appreciate some expert views on this.

A lithium cell suffers from being fully discharged and fully recharged, which means that you'll get more cycles from your battery if you keep its capacity between 5% to 80% than if you charge it from 5% to 100%. Actually it has been estimated that you gain around 200% of battery lifespan if you only charge it to 80% instead of 100%. This is why trickle charging (constantly charging / discharging near 100%) is so damaging to batteries. That being said, it's not a big issue to charge it to 100% once in a while (it can even help to re-calibrate the battery and improve the capacity reading) or when you actually need the extra 20%.

Here are some studies if you want to have expert views:

Factors that affect cycle-life and possible degradation mechanisms of a Li-ion cell based on LiCoO2, https://doi.org/10.1016/S0378-7753(02)00305-1
Study of life evaluation methods for Li-ion batteries for backup applications, https://doi.org/10.1016/S0378-7753(03)00208-8

[Vinay]

Oh, that's pretty good news then . This implies that I'll be just fine using that USB to barrel cable (https://store.pine64.org/product/1-meter...wer-cable/) most of the time. I'll only need to use that complete supplied charger when I expect to need a lot of battery capacity, which isn't often for me. There are more than enough occasions usually to charge and the applications I use are fairly light (like LyX etc). If it helps to extend my battery life then I'm all for it.