Laptops and their batteries: Trying to get my Dell XPS battery to last

Topic

New Reply

Some time ago, I bought an Acer Swift 1 laptop that I happened to see on closeout for a great price. It became my go-everywhere laptop, and it excelled in that role… mostly.

The problem I had with the Acer was mainly about its soldered-on RAM. It came with 4GB, and was not upgradeable. It did better than it had any right to with Linux and my experimental swap settings, but it was still less than ideal. It also had a weak CPU, an Atom-derived Pentium N4200 (now that Intel’s former flagship branding has been reapplied to its lowest-end fare).

I bought the Dell XPS 13 (9310) to replace it. There is a lot to like about the Dell, but one thing I do not like is that its battery did not have a very long service life (talking about the time before the battery is in need of replacement, not how long it goes on a charge). Generally, a laptop battery is “worn out” in terms of cycle count when it hits 80% of its original capacity… and my Dell did that in under a year.

At first, this seemed like a battery quality issue. My Acer battery had lasted for about two years and only dropped to 90 something percent of original, including a bunch of runs down to zero charge (very bad for it). The Dell, though, despite being a “premium” XPS, didn’t even come close.

I’d had issues with the battery life in my Dell G3 too. Even though it is a gaming laptop that was seldom used on battery (and never while gaming), its battery began a nosedive in reported remaining capacity shortly after I bought it.

“Dell must be using cheap batteries,” I thought.

I remember reading some time ago about some firmware change that was meant to extend a given laptop’s battery service life. My conclusion that this was a band aid solution. It had to be a defect in the battery or in the laptop somewhere. Either the battery was failing to last as long as specified by the battery OEM, or else the laptop is not doing a reasonable job of taking care of that battery. I doubted that dysfunction at that level could really be fixed with a firmware upgrade.

I know a lot more about Li-ion batteries now, and I can see I was off base with that one.

You may have heard all of the hullabaloo about Tesla’s electric cars and their methods of preserving the battery capacity. They limit a battery to a lowish level of charge by default, but if you pay them more money, you can use the “full” battery capacity.

That rubbed a lot of people the wrong way. How can they charge you more money to allow you to charge a battery you paid for all the way up?

I am not sure how they do it, but if there is some kind of warranty for battery longevity, I completely get it. Charging your battery to 100% will cut its life down quite significantly, and if Tesla has to replace the battery at their own expense, they have to be able to recoup the losses. If they don’t cover the battery service life, then yeah, it should be user selectable.

Charging a battery to 100% is stressful for that battery, but leaving it at 100% for days, weeks, or years on end is much worse. When it is under that kind of stress, it is hypersensitive to temperature. Higher temps are bad for Li-ion batteries all over the place, but if you combine higher temps and 100% charge, it just devastates the battery. The ‘Battery University” site describes the horror of the situation: a battery kept at 100% charge and 40 C temperature (quite typical inside a laptop) will lose 35% of its rated capacity in one year.  If you cycle the battery, then charge it back to 100%, it gets worse. If you let the battery get too low, it gets worse again. If you charge it or discharge it rapidly, it gets worse. All of that getting worse is permanent!

Now I think I get why my laptop batteries are dying so quickly. The Acer was a very low power computer that did not even need a cooling fan. The Dells, though, get very hot during use. Perhaps they are roasting the battery!

To know how bad that is, I have to know how much charge the Dell is putting on its batteries.  There is no hard and fast point where the battery is definitely fully charged, and anything beyond that is clearly bad. The more you charge a Li-polymer-ion battery, the higher its voltage grows (up to a point). This level of charge is reflected in the voltage in each cell of the battery.

A fully discharged Li-ion-polymer (I will just call it Li-Po) battery has a voltage of about 2.9V. A fully charged Li-Po is around 4.2 volts. Keeping the charge below that will extend the battery’s service life significantly, and the effect is intensified if it gets hot.

So I go and check my Dell’s fully charged battery voltage (as reported by Linux). It reports something like 8.6 volts. It is a four cell battery, so apparently they are wired in pairs (in series). That comes to 4.3 volts per cell, and is above the generally-accepted  ceiling for laptop batteries at 4.2V.

Now the short life is beginning to really make sense. It is a powerful, slim, light laptop that gets really hot, and if that was not bad enough, it severely overcharges the battery. If they were trying to kill off Li-Po batteries, this is how they would do it.

It is a rather cynical strategy. Set all of your units sold to overcharge their batteries, giving them a longer battery run-time than they would otherwise have. Let the critics get their laptops and review them. The overcharged laptops have decent battery run time over the week or so the reviewer has the unit, so they get a good review. People buy them based on the review. Money is made.

And then, six months down the road for an actual user, the battery is already nearing its “death” point (according to the industry standard… it is still usable long after it reaches 80%, bit it won’t last as long, of course).  The Dell warranty does not cover battery wear and tear. Get out your credit card and buy another costly Dell-branded battery, so it can be overcharged and cooked to death too!

By contrast with this. my Xenia gaming laptop’s battery is still at 100% health. It is charged to 100%, and has the same kind of existence as the previous (Dell) gaming laptop had, yet it hasn’t had the nosedive in capacity that the Dell gaming laptop did. So I investigate a bit closer, and the fully charged Xenia battery reports 12.03V. My guess is that it is a six cell battery, with the cells in pairs once again, so each cell would be at 4.01 volts.

That is way, way lower than 4.3V for a Li-Po battery. It does not sound like much, but consider this: if I charge my Dell battery to ~4V per cell, it will be at no more than 58% charged. In order to get ~53 W/h out of this battery, Dell has to super-overcharge it. At the charge level of my Xenia, this battery that Dell calls 53 W/h would be no more than 31 watt hours. It’s not going to look very favorable in reviews that compare it to Lenovos and other laptops with a paltry 31 W/h! But what if that 31 can be turned to 53 at no additional cost to Dell, but with significant cost to the users over time, after you already have their money?

Like I said, cynical. I’ve never seen any of the reviews mention this. I’ve never seen it mentioned on a forum before. This is the kind of thing that could easily (and apparently has) slipped under the radar. I guess people just buy a new battery and call it a day (repeating as necessary). I get that… I just ordered a third battery for the thing myself.

I can (and I have) reduce the full-charged point on the battery via the UEFI settings. I can undo Dell’s overcharging. But then I get a much shorter battery run time than I thought I was getting, and what is printed on the side of the battery, when I read the reviews and bought the Dell.

Those are my choices. I can have either the battery capacity I was promised, or I can have a battery service life on par with other laptops, but not both.

I wonder if other laptop makers do this. I may have to start perusing the demo laptops at the big box stores and checking their battery voltages. Those new Acer Swift Xs are looking appealing…  and my previous Acer was much better than my Dells as far as battery longevity.

My opinion of Dell continues to crumble.

Dell XPS 13/9310, i5-1135G7/16GB, KDE Neon 6.2
XPG Xenia 15, i7-9750H/32GB & GTX1660ti, Kubuntu 24.04
Acer Swift Go 14, i5-1335U/16GB, Kubuntu 24.04 (and Win 11)

6 users thanked author for this post.

windbg, fridges, wavy, Elly, F A Kramer, jburk07

Reply | Quote

Replies

Ascaris wrote:

Charging a battery to 100% is stressful for that battery, but leaving it at 100% for days, weeks, or years on end is much worse. When it is under that kind of stress, it is hypersensitive to temperature. Higher temps are bad for Li-ion batteries all over the place, but if you combine higher temps and 100% charge, it just devastates the battery.

My Lenovo Y530’s battery is always on charge and at 100%.
The previous battery was swollen and replaced a couple of months ago after 4.5 years.

Reply | Quote

Charging a battery to 100% is stressful for that battery

I’ve read recently that, at least for cell phones, there is now built in software protection for the “continuous charge at 100%” issue.  If it’s true, it seems unlikely at least some newer laptops would have something similar.

 

Desktop mobo Asus TUF X299 Mark 1, CPU: Intel Core i7-7820X Skylake-X 8-Core 3.6 GHz, RAM: 32GB, GPU: Nvidia GTX 1050 Ti 4GB. Display: Four 27" 1080p screens 2 over 2 quad.

Reply | Quote

It’s not that the battery is continuously charged when it reaches 100%… older devices used to do that, which is why the advice was to never leave any rechargeable device on the charger for longer than was necessary, as the continuous charging would harm a fully charged battery.

On any modern phone, laptop, or tablet, the device’s charging logic will stop the charging when it reaches a specified point (usually 100%). Just having the battery at a static state of 100% charge, even if it is not being charged at that point, causes signficant capacity loss unless the battery is kept quite cold.

From Battery University’s article on preserving Lithium-based batteries:

State of charge: 100% (~4.2v per cell)

Temperature           Amount of capacity loss

0 C                         6% of original capacity lost in one year (Battery health 94%)

25 C                       20% of original capacity lost in one year (Battery health 80%)

40 C                      35% of original capacity lost in one year (Battery health 65%)

60 C                      40% of original capacity lost in three months (Battery health 60%)

It is not the act of charging a battery to 100% that causes this, nor the total number of charging cycles. It’s the amount of time spent at 100% charge that results in this loss. As you can see, if you keep the battery cold, at the freezing point of water, it only loses 6% of its original capacity within one year if it is charged to 100%. But if you keep it near room temperature, it will lose 20 percent of its original capacity in a year if it is kept fully charged, without any additional cycling or charging (other than that required to keep the battery at 100%, which is very minor in Li-ion batteries).

It is not practical to keep a phone or laptop at 0 C, so the better way to preserve battery capacity would be to not charge it to 100% and leave it there. If you charge it to 100% and take it off the charger and begin using it, you will be back below 100% before too long. Some damage will have occurred from the charging cycle, and a little more from the time spent at 100% charge, but it will not be at 100% as long as a laptop that is pretty much plugged in all the time.

For the purposes of defining a battery’s lifespan in cycle counts, a battery is considered “worn out” when it reaches 80% of its original capacity. If you let the battery charge to 100%, the default level in most cases, and if you keep it there (as with a laptop that is mostly used on AC power), the battery will reach this EOL before the first year is out, even if the cycle count never reaches 1.

Normal cycle count is variously defined as 300 to 1000 cycles over the life of the battery, but it is far from being the only cause of battery degradation. When I began reading about this, I was perplexed about how my battery, which I doubt has anything close to 100 cycles, let alone 300, is at its EOL point.

On the XPS, Dell’s charging logic allows it to charge up to ~110% of normal by default. An extra 10% like this will roughly cut the battery’s life span in half, according to the same site above, compared to charging it to 100%.That would be 150 cycles on the lower end of the range, but I am sure it was nowhere close to that. It was kept at 100% nearly all the time, though.

Given this, it is perhaps remarkable that my two year old Dell battery “only” degraded to 80% after two years (with a second battery in rotation keeping the usage at about half what it would have otherwise been). I have kept it fully charged most of the time, and I don’t keep the laptop at 0 C. It began to degrade pretty quickly, and that was when I went to the alternate OEM battery. When that alternate reached 20% loss, I swapped the original battery back in… and when that one hit 20% loss, I bought a new aftermarket battery, on the hopes that it might be a good quality battery without the expensive name. “OEM” Dell batteries ordered anywhere except Dell or an authorized seller (if they even have any for batteries) are, as far as I am concerned, all counterfeits or factory seconds.

There is a thriving gray market in the tech sector for these branded items that failed QA (but that still work, at least superficially) and should have been recycled.

Back when I used to (mostly try to) sell PCs for a system integrator in the 90s, I wondered how our competition always managed to undercut on prices that had essentially no margin. Even at the “cost” price, we were being undercut by a hundred or two dollars on a PC. The owner of the shop was a smart guy with the same connections to China that everyone else had… so how did everyone else get so much better prices on things?

I learned then of the gray market for factory seconds. I was expressing my chagrin over this situation to one of the acquaintances I made, and when I mentioned the bit about selling computers made out of factory seconds that failed QA, he sort of just gave a little shrug and said, “So? Everyone does that.”

No, everyone doesn’t. We never did.

How can you compete with that without getting down into the mud and being a sneaky crook too? If you tell people that the competition does it, even if you don’t “name names,” you look bad for badmouthing the competition, and people probably won’t believe it is true. That would tend to drive customers to the “victim” of your “smears,” not the other way round. But if you don’t mention it, your QA assured parts look exactly like the QA failed parts from someone else, at a price premium, in a market where “cheaper! cheaper! CHEAPER!” was what it was all about.

It’s hard to convince people you get what you pay for when your wares are literally assembled from commodity parts that look exactly like the same-branded parts from someone else. It did mean that they would have a lot of their computers come back for repairs under warranty, but they saved so much that they were willing to just slap in another factory second if the first one failed. Eventually they would get you to the end of the warranty period… or just go out of business and come back with another name, which we used to see quite a lot too.

 

Dell XPS 13/9310, i5-1135G7/16GB, KDE Neon 6.2
XPG Xenia 15, i7-9750H/32GB & GTX1660ti, Kubuntu 24.04
Acer Swift Go 14, i5-1335U/16GB, Kubuntu 24.04 (and Win 11)

Reply | Quote

My Samsung S22 Ultra has a setting to limit the charge to 85%. Of course I have it set and I get over 2 days of usage normally. I usually charge it when it gets around 20%.

May the Forces of good computing be with you!

RG

PowerShell & VBA Rule!
Computer Specs

Reply | Quote

RetiredGeek wrote:

My Samsung S22 Ultra has a setting to limit the charge to 85%. Of course I have it set and I get over 2 days of usage normally. I usually charge it when it gets around 20%.

Apple’s iPhone and watch are automatically charged to 80% (changeable by users).

Reply | Quote

While I broadly agree with the advice you give I think the narrative that Dell is overcharging batteries is unfounded.  Dell is likely using a different Li-Po battery chemistry than your other laptop.  Specifically, Dell may be using an NCA battery which in a lot of cases do have a specified 4.3v max charge voltage (example).  Additionally your other laptop may have been using a chemistry with a lower max voltage (4.1v for LMO as an example).  So I believe the narrative that Dell is purposely over-charging the laptop is unfounded.  Additionally the benefit of the overcharging you claim is quite small, capacity vs voltage is certainly not linear and in that region of charge there is a logarithmic relationship b/t voltage and charge (take a look at the Nernst equation if you want more information).

Reply | Quote

Sorry it took me so long to notice and read your reply… I hope you will see this as a guest.

It is possible that you are right about the difference in chemistry affecting the maximum nominal charge. I am no battery expert, and all I have to go by is my own limited experience and the references that I’ve read while researching the topic. There are a lot of them, and they all cite the same nominal cell voltages that I have repeated here.

The reports on the capacity vs. voltage are based on what I saw reported by the battery itself, so I have to presume it is correct.

Ultimately, as they say (when they get the idiom right!) is that the proof of the pudding is in the eating, and I am not accusing Dell of chicanery based solely on voltage readings. I noticed horrible battery service life first, then attempted to find a reason for it.

My laptop with lower voltage (the Xenia) is 16 months old by now, and its battery still sits at 100% of original capacity, as it did months ago when I wrote the original post. By the time my Dell G3 (also a 15.6 inch gaming laptop) was that old, it had already lost enough capacity that I ended up replacing it. Both of them were used the same way (nearly 100% of the time on AC power, charged to 100%, with periodic AC-only gaming sessions that generate a lot of heat).

The G3 came with a 135w power brick that was too small for its intended purpose (gaming), and I replaced it with a 180W Dell power brick that has been flawless. The Xenia came with a 230w power brick that was also flawless. If anything, the Xenia is dissipating more heat than the Dell, but its battery is like new, where the Dell’s was well past the 80% threshold already, and plummeting like a rock. .

I had a similar experience with several XPS batteries, though with a very different usage pattern. I used the XPS on battery quite a lot, but for every hour on battery, there were several on AC power, where I let it charge to 100% and maintain itself as the firmware dictates with the stock settings. Despite a relatively low number of cycles, the batteries were hitting the 80% mark in well under a year. Two OEM batteries (direct from Dell, so I know they were genuine) and one aftermarket, same story for all three.

I have not posted a follow-up, but I have some more info to report.

Since that post, I bought another Dell battery, sold as used but with no more than 3% reduction in capacity. I consider any “Dell” battery sold as new on eBay or Amazon to be almost certainly couterfeit, but I thought the used-good battery route could be less likely to draw the fakers. It did not come from Dell, so it could be a clever counterfeit, but I compared it side by side with the original, and everything is identical, right down to the Dell part number stickers and QR codes. The one I got actually read 100%, and it still does, after four months of daily use that is probably harder (deeper cycles) than with the other three batteries. I realize four months is not much, but by now the other Dell batteries were in the upper 80s in terms of remaining capacity.

The difference is that I set the max charge level to 85% in the UEFI.

If the overcharging is not a valid explanation, then there is some other thing at work making two Dell computers use up four batteries in a short time, where the Xenia has not done that… and also happens to report lower voltage levels as 100%. Given how average state of charge has a minor effect on battery longevity at low temperatures but makes these batteries exceptionally sensitive to higher temps, as you would normally find in a modern laptop, it stands to reason that the state of charge is why my Dell batteries died so fast. It is not the number of cycles or the temperature of the laptop alone, as the examples of the Xenia and the XPS with the used good batteries show, The only difference has been the voltage to which the batteries have been charged.

The way Dell configured my laptops and the charging parameters led to very short battery life… that much is indisputable. If there is a better explanation than overcharging, I would be most appreciative to find out what that is.

 

Dell XPS 13/9310, i5-1135G7/16GB, KDE Neon 6.2
XPG Xenia 15, i7-9750H/32GB & GTX1660ti, Kubuntu 24.04
Acer Swift Go 14, i5-1335U/16GB, Kubuntu 24.04 (and Win 11)

Reply | Quote