The World Needs a Better Smart Phone Battery Test. And You Can Help Build It.
It’s official. Sprint’s HTC EVO 4G is the fastest selling phone yet for the No. 3 carrier. And, according to many reviewers and early adopters of this powerful device, it’s also the fastest when it comes to running out of juice. Guess what? It’s not as bad as you might think. Using our own battery test, the EVO 4G surfed the web continuously for 4 hours and 21 minutes (over 3G). That runtime actually falls between the Motorola Droid (4:01) and the myTouch 3G Slide (4:40). So how does our test work, and more important, what do you think we should do to improve it?
Earlier this week Apple touted the longer endurance of its upcoming iPhone 4 , which is rated to provide 2 hours more talk time than its predecessor, and to last an extra hour when surfing the web over 3G and Wi-Fi. According to Apple’s Battery Performance Test Information Page, the company conducts its evaluations over Wi-Fi and 3G using dedicated web and mail servers, “browsing snapshot versions of 20 popular web pages, and receiving mail once per hour.”
The fact that Apple measures battery life based on Internet usage is actually quite refreshing, because most of the carriers are still quoting talk time. (Note to service providers: No one talks anymore.) Our test is different than Apple’s, but it’s not necessarily better—yet.
Here’s how Laptopmag.com’s Smart Phone Battery Test currently works:
- Similar to our notebook battery test, we set the screen brightness to 40 percent (which we have to guesstimate on many handsets).
- We turn Wi-Fi off and prevent the display from locking or dimming.
- The handset visits one of 60 popular websites every 30 seconds over 3G using a web-based app, which logs the amount of time the device has been surfing. (Any smart phone with a WebKit-based browser should work.)
- When the phone dies, the last log entry recorded is used to determine the smart phone’s endurance.
- All tests are conducted in the same location in our New York City office, near a window, although at different times of the day. Due to time constraints, we currently run the test once, though there are plans to increase the number of runs and average them.
Before I get to how we could improve our test, I want to stress that it’s an important first step to be able to compare smart phone battery life across multiple phones and platforms using something other than anecdotal evidence. I also realize that web surfing is only one measure of battery life, but it’s something we know that all smart phone users do and will continue to do more often.
Here are some of things I’m pondering and would love to get your feedback on.
- Should we increase the time interval between webpage loads from 30 seconds to 60 seconds, or longer? What most accurately reflects real-world smart phone usage?
- Do you think the phones should be set to 100 percent brightness, or whatever the default setting is out of the box? (We would prefer to keep things equal.)
- The only active program will be the browser, but should we turn off notifications or tweak the e-mail settings to turn off such things as push e-mail delivery? Or should we just leave everything else as is?
- Would you want to see both Wi-Fi surfing and 3G/4G times? We would likely only have time for two test runs in total for each phone.
Other things I’ve been thinking about in regards to our test include whether phones that support Flash are put at a disadvantage because they render pages, or portions of pages, that other devices can’t display. At some point I’d also like to add a multitasking round, where we play music in the background while conducting the surfing test, perhaps.
By working with LAPTOP Online Editorial Director Avram Piltch , who created this test, my ultimate goal is to give consumers a better estimate as to which smart phones will get them through the day without having to find an outlet. And once we’ve made some improvements, we’d like to share our test with handset makers and carriers so that the industry can potentially move toward a more realistic measurement of usage time.
Now it’s your turn to help us get there.
Editor-in-chief Mark Spoonauer directs LAPTOP’s online and print editorial content and has been covering mobile and wireless technology for over a decade. Each week Mark’s SpoonFed column provides his insights and analysis of the biggest mobile trends and news. You can also follow him on Twitter.
Follow Mark Spoonauer on Twitter,Google+. Follow LAPTOPMAG on Twitter, Google+ or Facebook.



Jun 11, 2010 02:12 PM EDT by 











June 11th, 2010 at 6:27 pm
“…but should we turn off notifications or tweak the e-mail settings to turn off such things as push e-mail delivery?”
Part of me says no. I don’t know about everyone else, but I use my phone for texting and email. So I’d like to know how long a phone would run with all that stuff on. But then I thought about the test. Normal usage isn’t visiting a web page once every 30 seconds either.
I chuckled at Apple’s test – “receiving mail once per hour”. Really? How many emails do they send to it. I don’t know how much power receiving email uses so I don’t know how much of a difference it would make to check it more than once an hour.
I guess as long as the test is the same, then at least we can get a relative comparison between phones.
June 14th, 2010 at 10:38 pm
I think 60 seconds is a more accurate reload time for the smartphone user. For some phones, a portion of that will be used to download the page, but I think that’s probably the average amount of itme i spend on a page, obviously depending how much content there is.
With regard to WiFi – while I don’t think this will be useful in terms of usage (how many situations are you in where you can get WiFi, but don’t have access to a power outlet?), it would be good for reference eliminating any disadvantages a phone might have due to stronger or weaker cell signal in the test area. That’s one of the only variables you can’t control in this test – unless you go to a location where each phone gets the same dBm reading from the tower – which is unlikely.
July 8th, 2010 at 11:03 pm
Brightness should defiantly be 100
April 1st, 2011 at 2:39 am
You would need to have the default settings in place as for when you take the phone out of the box for the first time. For many people they would not adjust any settings, so testing with tweaked settings would not show a real world setting. By all means then tweak ti to see how long a battery can last.