Slashdot Mirror
T-Mobile Smartphones Outlast Competitors' Identical Models
An anonymous reader writes Laptop Mag battery tested the leading phones on all four major U.S. carriers and found that the same models on T-Mobile typically last 1 to 3 hours longer on a charge. This trend is not new, but has continued for over 3 years of testing.
The article says While we don’t know for certain why T-Mobile phones last longer on a charge, there are some strong possibilities. T-Mobile’s network could be more efficient at sending and receiving data because of the bands it uses, or maybe there are far fewer customers on its LTE network, easing the strain. Another possibility is that T-Mobile tends to pre-load less bloatware on its flagship devices relative to the other carriers.
AT&T is firmly in second place in the battery life findings presented, with Verizon and Sprint jockeying for last of the four carriers measured. It woud be interesting to see a similar test battery for phones in marginal reception areas; searching for service seems to deplete my battery faster than talking does.
I thought a phone was using maximum RF power when it was looking for a tower to talk to?
John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
They only tested Android smartphones. So we don't know if this is something specific to Android or not. Not like anybody uses iPhones, huh?
My bet - different CDRX settings, fast dormancy, idle timers. Is probably a better engineered network.
It would be interesting to know if an unlocked AT&T phone moved to T-mobile's network suddenly lasts longer.
And this is with t-mobiles software installed. With a clean phone, the T-mobile "my account" software is the highest usage bit of software on the phone. Disabling it was worth hours of runtime.
You would have to be crazy to be sane in this world. -Nero
My satoshi on this.
Let's reconvene at an appropriate time to proclaim winners.
Prove anything by multiplying Huge Number times Tiny Number
I'd like to see what this test looks like with all the phones involved running the same software load. i.e. No Verizon crapware. Just scout out a handset available on all 4 carriers, install Cyanogenmod on one and leave a second one stock. Then we should get a more accurate picture of what's going on here.
The Amarri pray for god, the Caldari pray for profit. the Gallente pray for peace, but the Minmatar pray their ships hol
according to the Colbert Report, they're great for clearing up constipation!
http://thecolbertreport.cc.com/videos/p4t1a2/vibrant-constipation-pill
Ok, so I did an anecdotal apples-to-oranges comparison between different phones on different networks and noticed that my phone was different, so that means that the apples-to-apples comparison the researchers did must be wrong, even though they explicitly controlled for that factor.
I think this is what you said. Let me know if I mistranslated.
Test with iPhones. No pre-loaded carrier bloatware, same exact OS across all carriers.
All iPhones sold in the U.S. can be used in most other countries - they are multiband.
But, for example, I have a Verizon iPhone. If I want to move to T-Mobile I have to get a new phone. I think T-Mobile's frequency for data is one not used elsewhere in the world, even though T-Mobile is a GSM provider. I think you can move between AT&T & T-Mobile freely (both GSM). I believe over time this will iron out more...
It's a mess to be sure.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
For another data point: I got a Dell Streak from eBay a few years ago that was factory unlocked. My understanding is that meant it had radio bands at that time for AT+T (HSPA/HSDPA+ I think?). I got the cheap T-Mobile prepay plan, $30/month for 100 minutes talk, unlimited text, 4GB data at "3G" speeds, then throttled back to 2G for any over that limit (it was going to be an experiment in heavy Android usage as mostly a secondary data phone vs my Verizon main phone then, a Windows Mobile Imagio).
At that time, T-Mobile was just starting to buy and phase in those bands that they had bought from AT&T, but I mostly just got Edge network, so slow (and no danger of hitting 4GB in a month...). As they started to build out the network for switching the faster HSPDA+ from whatever bands they had been using to the AT&T bands my phone started showing that "H" on the top status bar in my travels around the area more than the "E" (Edge), and it was/is a lot faster, WHEN it is strong. However, at my house and workplace it is constantly fluctuating between the H and the E, indicating that I am in a "boundary" zone between the 2 types of signals at those places, and that runs my battery down FAST. If I lock the phone setting to "GSM" only, instead of the default, "GSM/WCDMA auto", then the battery lasts much longer, but then so does the time needed for data transmissions.
I have since upgraded from the WM 6.5 Imagio to several VzW Android's in turn, Moto Bionic, Samsung Note 2, Moto Razr HD (Note was just too big to tote around on my belt), and their LTE has not really been much faster, if any, than T-Mo's HSDPA+ when it is a strong signal, BUT I get VzW's LTE in a lot more places (including home and work), so the batteries have usually lasted just fine, although I almost always have the phones plugged in with USB power in the car, at work, and at home.
Still, the times I roam around unplugged show the VzW phones hold up fine as long as the signal is reasonably strong, which has not been the case for the T-Mo Streak. I would be interested to know how much effort was made in the testing to stay in strong signal zones for the respective carriers with their phones - it makes a HUGE difference in my experience.
FWIW
AT&T & T-mobile were GSM in the states, and the rest were CDMA... but could be wrong. You're at least partially right about the bands though. As a visitor to the US, I pretty much stick to AT&T as they use the same bands as most of the world (i.e. ones my phone can do). T-Mobile I think uses one regular band, but also a rare non-standardish one.
There's N channels for each radio technology: 1XRTT, 3G, EVDO-RevA and RevB, LTE, etc.
The phone gets informed by the carrier which channel it is on, and depending on the channel, it will bring up the antenna more, or less often, to receive things like SMS, PTT, that should come in a timely manner. There are many strategies to keep the traffic channel up, or to trip and dip into the network less frequently.
You also do not have any control of which traffic channel you will be on, as that's pushed down to you depending on congestion, and signal strength, etc.
They work the exact same way, so I suspect they are dipping into the traffic channel less often (as well as getting fewer updates from the network) for T-Mobile than for VzW.
Settings... Battery
Sorted by %usage including cell standby, wifi, voice calls, apps, etc.
There doesn't need to be a mystery.
From the article “make sure that it’s receiving at least 3 bars of service” The words “at least” worry me here. That seems to imply some had 3 bars and some had 5. Since signal strength is often tied to network speed and how much power the radio needs to communicate with the towers this alone makes the results suspect. A carrier with 5 bars is going to have a huge advantage over one with 3. Maybe they misspoke and really they all had the same number of bars... even then I'd think they'd have to run a speed test as well since those bars are phone service bars not 3G/4G/LTE/whatever bars.
>Didn't RTFA so don't know they compared the same four phones on each of the four networks, with sixteen phones tested in all.
I think this is what you said. Let me know if I mistranslated.
This is not really related to T-Mobile, but I do know that the battery life on my wife's HTC Vivid **DOUBLED** when I installed CM11 on it.
It used to last 7-8 hours and now she can comfortably go all day and not have to carry a power pack around with her everywhere she goes.
Different versions of the same model phone use different radio chips.
I just bought my wife an S4 Mini, with a choice of at least 5 different models that only really differ in the radio chip - I9190, I9192, I9195T, I9195L and I9197.
They're all S4 Mini's, one without LTE or NFC, one with dual SIM, the others are all LTE with different frequency bands.
You're seriously arguing about this because they didn't use the exact same physical units across all four carriers for each of the four phones? Come on.
Jesus Christ they never run out of things to fuck up.
Interesting observation... I've generally had the opposite experience, at least comparing AT&T against T-Mobile devices.
I took an unlocked Nokia 1520 from AT&T, to T-Mobile, to Consumer Cellular.
Originally, on AT&T, I could not make it through a complete day on a single charge. Took it off the charger at 4:30AM. Battery was dead by 3pm.
Took phone to t-mobile. Off charger at 4:30AM, phone still had a quarter charge left at 10:30PM when I plugged it back in.
Now on Consumer Cellular. Same phone. AT&T is the service provider to Consumer Cellular. Battery is not making it through the day again.
Usage patterns are similar through all three carriers. I did not do any rigorous scientific tests on this. This is observational usage data.
I'm wondering if WiFi calling has anything to do with it. My T-Mobile Note 3 lasts all day (with more than a half battery to spare) while at work on the WifI network, it appears to shut off the LTE radio altogether. LTE/GSM signals in my building are crap, and if it's not on WiFi, the battery runs out quickly, usually before the end of the day.
Did the study include the effects of calling over wifi?
I have t-mobile and connect to wifi networks at home and work for my phone connection and my charge lasts a lot longer than
when I'm away from wifi networks I can use.
As far as I know, I think t-mobile is the only carrier to implement calling over wifi.
(What, RTFA and check if that's mentioned? Of course not...)
I switched from T-Mobile to P-Tel which uses T-mobile's network and lost about 2/3 of my battery life on my iPhone. Nothing else changed...
as far as I know, cell phones and smartphones won't even start up without a SIM card
I'm told smartphones with no SIM work on Wi-Fi to roughly the same extent as an iPod touch or Wi-Fi-only tablet.
tl;dr
T-Mobile has Wifi calling. It turns off the Cell radio and uses WIFI for calls. This is one likely reason Calls are made through a router 50 feet away instead of a tower a mile away.
T-Mobile has wifi-calling, so when it's using Wifi, it's not using cellular, so it could be that.
Bloatware on my Verizn phone means when a phone call comes in, or a page, and I view it, the browser is killed off due to lack of RAM. When I switch back, the browser re-downloads the page.
Verizon is pathetic. There's a "kill off unnecessary running programs" feature under task manager that they had to be dragged kicking and screaming to use, and it kills off just one thing, too. Which increases used RAM by 2 mb instead of decreasing, and that program shortly auto-restarts anyway.
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
T-Mobile and AT&T are ahead of Verizon and Sprint in battery life because of their use of GSM. GSM is a widely used global standard while CDMA is mostly found in North America, so power efficiency of the GSM radio probably gets more attention. GSM reception is also less computationally intensive; the spread spectrum methods used by CDMA take more work to decode. This all has some impact even if you never talk on the phone, because it is using power to maintain its connection on the network.
Talk time is a more complex equation. Both standards use power control, so battery life is lower in weak signal areas. A CDMA radio has more stringent signal linearity requirements - it must maintain constant power and constant signal delay across the entire spread spectrum - that make it less efficient than a GSM radio.
Then there is data power consumption. That is a complex calculus of amounts of data, bands and modes used for data communication, and network congestion. AT&T and T-Mobile have a mix of LTE, HSPA+, and HSDPA, and heaven forbid, EDGE in remote areas. (T-Mobile had HSPA+ 42 in some areas though they no longer offer that speed in areas with LTE coverage; AT&T topped out at HSPA+ 21.) Verizon and Sprint mix LTE with 1xEV-DO and 1xRTT; their non-LTE networks are much slower, so any data transmission on those networks takes longer and uses more power.
...which usually come with a cheaper phone.
I have firsthand experience of this. I had a $30/mo. account; my next-door neighbor had a $60/mo. account. I could only get about half the signal she did, even in the exact same location.
I don't know how that impacts battery life, but I got about 10 hours on a cheap phone (at least, before its battery went poorly at just over a year old).
~REZ~ #43301. Who'd fake being me anyway?