My Galaxy Note20 Ultra is a 2020 flagship smartphone I actually bought as “old stock” back in November 2024. I’m still on the fence about whether to trade in its big screen, S Pen, awesome camera, and expandable storage for a newer phone within my budget.

Unfortunately, this phone is starting to show its age, with a single charge only lasting two-thirds of a day on its stock software. I did try to rescue it by flashing it with ArtisanROM, which dramatically improved battery life, but it left me wondering whether the battery itself had given up.

Would replacing it buy this phone some more time, or was I wasting money throwing it at an aging flagship? Samsung’s stock software declared the battery was “Good,” without telling me how much capacity actually remained. So I tested it four different ways before deciding whether to replace the battery or the phone.

Samsung said the battery was “Good,” but what does that even mean?

The built-in verdict was reassuring until I looked at the numbers underneath

The stock diagnosis and customer support app “Samsung Members” gave me an easy but not overly helpful result at first. Its battery diagnostic turned “Normal” and “Life: Good.” I ran it again after installing it on ArtisanROM and got back the same ambiguous verdict.

That was briefly reassuring until I looked at the line beneath it. “Capacity: 4500 mAh (typical)” describes the Note20’s original battery specification, but it was very unlikely it still held that much.

The battery information page in Settings gave me two numbers that didn’t really line up:

1. The battery health score was 91%. This sounded a little too good to be true for a 2020 phone, especially after the poor performance I experienced on the stock software.
2. The battery cycle count was 1,444. That’s a good ~400 full-cycle (100% to 0%) charges above what is typically considered end of life for a flagship phone’s battery.

The discrepancies between these dates were, in part, explained by the phone’s weird history. The battery was manufactured on the 4th of July 2020, while the first-used date was the 4th of November 2024. That’s a four-year gap, and it does match because the phone was bought as an old stock item around September 2024.

Regardless of the results Samsung gave me, neither could actually tell me what the long shelf life or those 1,444 cycles had actually done to this battery.

Here, I was deciding between getting a replacement battery and a new phone. The “Good” result and the reported 91% were an ideal starting point, but I needed to watch the phone discharge and charge before trusting either number with that decision.

An hour of video showed me what life with this battery would look like

It lost charge steadily, but the remaining time forecasts were all over the place

Having the battery just sitting on the desk is one metric, but I wanted something closer to real-life use. I installed Battery Monitor from F-Droid, unplugged the phone at 44%, and let it play what I assumed was a very informative, hour-long nature documentary while I went and ate dinner.

Its history logged 43% at 8:02 pm and 34% at 8:56 pm, a nine-point drop in 54 minutes. The level stepped down one percent every few minutes. There weren’t any real sudden plunges or shutdowns.

This information was a bit more useful than simply calling the battery “Healthy.” Aging lithium-ion cells can lose capacity and, when worn down, struggle to maintain their voltage under load and start having performance problems. Samsung itself lists behavior like abrupt drain and unexpected shutdowns among the warning signs.

My video wasn’t exactly what I’d call a stress test, and I didn’t run it into the more revealing low-battery range. Still, sustained playback at mid-range charge behaved relatively normally.

The runtime forecasts were, unfortunately, really inconsistent. Battery Monitor later predicted 7 hours and 34 minutes remaining at 32%, while Samsung confidently predicted 15 hours and 31 minutes at 31%. Both were forecasting future use and not measuring battery health.

Samsung’s battery page also attributed 12.3% to YouTube over its displayed usage period. That figure shouldn’t be confused with the 9% lost in my timed test.

It was relieving, especially given the same test would have dropped the battery by ~20–30 on the stock software. However, I wasn’t ready to call the battery healthy just yet. ArtisanROM likely reduced the power the phone uses during ordinary tasks. The battery hadn’t magically regained lost capacity.

This test showed me the phone was working well enough to keep using — at least with the custom ROM. A proper measured test, though, would tell me if the battery needed replacing.

Developer(s)

Tomasz Świstak

Platform(s)

Android 8.0 or newer

Battery Monitor tracks charge level, temperature, voltage, and battery history. Its discharge log helped me see how steadily my phone lost power during the video test.

One full charge gave me a less reassuring estimate

AccuBattery saw 76% health where the phone reported 91%

I installed AccuBattery from the Google Play Store because it demanded a substantial charging session before it would produce test results. At 28%, I set its design capacity to 4,500 mAh and, being impatient, connected my USB-C PPS charger.

The app showed live current and voltage, then tracked the climb to 100%. All up, it took 1 hour and 11 minutes, which was significantly less painful than watching the phone discharge.

Across those 72 percentage points, AccuBattery recorded 2,469 mAh added. Its estimated full capacity therefore was 3,429, since 2,469 divided by 0.72 is 3,429.11. Compared with the original 4,500 mAh specification, the battery has fallen a long way to a measly 76% health.

The phone’s battery information page claimed it was 91% healthy, and this was a 15-percentage-point difference that couldn’t just be ignored.

This was the first result that made me start considering replacing not only the battery, but also the phone. It was clear that Samsung’s “Good” verdict was rather generous, but I wasn’t quite ready to declare it wrong because one test hardly qualifies as conclusive.

AccuBattery’s documentation states that it needs a 60% charge for its health calculations, and that multiple sessions improve its accuracy. My 72-point charge definitely qualified, but later charges could change the estimate completely.

The app’s “0.97 cycles” of wear was another issue. That’s a description of its wear estimates for this session, not the phone’s whopping 1,444 lifetime cycles. For now, 76% gave me a reason to keep measuring before finally deciding to scour Amazon.com for a replacement.

I am keeping the flagship, but I will keep testing its battery

A replacement battery makes more sense to consider before replacing a phone I still like

Four checks later, I think Samsung’s “Good” was far too generous to be useful. AccuBattery estimated 76% health against 91% given by Samsung’s battery information. The issue is that the lower figure only came from a single charging session. I’ll need to keep measuring over the next few charges to see just how far off Samsung’s estimate was.

I’ve decided to replace the battery anyway. After 1,444 heroic full charge cycles, the lower estimate gives me enough reason to spend the money on a new cell. Outside of curiosity, I don’t really need to know its exact remaining capacity to make that choice.

Even though I know the Note20 Ultra won’t last forever, I still love using it. Its screen, camera, S Pen, and expandable storage, plus the improved performance running on ArtisanROM, are all features that I would have to check carefully for a potential replacement in 2026.

A new flagship of this caliber costs a lot more than the Note20 did back in 2024, while a phone within my budget range would mean giving up the things I love about this one. That’s a lot to sacrifice when a new battery just might give this awesome phone a few more good years.