Forgot your password?
typodupeerror
Power Hardware

Breakthrough Promises Smartphones that Use Half the Power 110

Posted by samzenpus
from the save-the-juice dept.
Dupple writes in with news about a discovery that should extend the life of your battery in the near future. "Powering cellular base stations around the world will cost $36 billion this year—chewing through nearly 1 percent of all global electricity production. Much of this is wasted by a grossly inefficient piece of hardware: the power amplifier, a gadget that turns electricity into radio signals. The versions of amplifiers within smartphones suffer similar problems. If you've noticed your phone getting warm and rapidly draining the battery when streaming video or sending large files, blame the power amplifiers. As with the versions in base stations, these chips waste more than 65 percent of their energy—and that's why you sometimes need to charge your phone twice a day. It's currently a lab-bench technology, but if it proves itself in commercialization, which is expected to start in 2013—first targeting LTE base stations—the technology could slash base station energy use by half. Likewise, a chip-scale version of the technology, still in development, could double the battery life of smartphones."
This discussion has been archived. No new comments can be posted.

Breakthrough Promises Smartphones that Use Half the Power

Comments Filter:
  • by mattr (78516) <mattr&telebody,com> on Thursday November 01, 2012 @02:21AM (#41839371) Homepage Journal

    Why do I not believe that 1% of global electrical production goes to powering wireless base stations.

    • by aXis100 (690904)

      Yeah, I dont believe them either.

    • by Decker-Mage (782424) <jack_of_shadows@yahoo.com> on Thursday November 01, 2012 @03:04AM (#41839541)

      Probably due to the fact that all of IT consumes about 1% of all power globally. And notice in that statistic "about" which, if it comes above 0.0000...01% somehow gets magically gets rounded up (apparently using ceil (APL) function rather than a real rounding function). If they really want to save power generated capacity, they really should look at replacing all those power bricks out there with something remotely efficient before thinking about the power consumption drawn from an, also admittedly, inefficient battery, on the way to the power amps.

      Matters not much, methinks, as no one is going to take advantage of the new designs until (1) they are incorporated into "stock" parts and (2) they are cheaper than the designs they are replacing. Almost forgot, and no one is still running a fire sale on the old chips.

      Articles like these, long on promise, short on economics, or long on threat, and short on the same thing, economics, piss me off.

      • by drinkypoo (153816)

        Why do I not believe that 1% of global electrical production goes to powering wireless base stations.

        Probably due to the fact that all of IT consumes about 1% of all power globally

        [citation needed] [newscientist.com] - yours is the only estimate I have seen which is so low.

      • The initial market will be in the developing world, where 640,000 diesel-powered generators are used to power base stations, chewing through $15 billion worth of fuel per year.

        This quote from the company gives the economics away. The monetary savings from the new base station amplifier's efficiency likely do not offset the increased purchase cost -- for stations wired to the grid. So they're selling it as a way of decreasing cost and re-fuel frequency for off-grid stations.

    • by siddesu (698447)
      More importantly, why don't I believe that halving what the phone radio uses will double battery life?
      • On more modern / powerful phones with larger screens, it is not the radio that is the dominant problem anymore. It is the screen.

        If you are "Streaming video" for any length of time with a 4.5" screen phone, it is the screen using most of the battery, not the radio. Screen battery use is also bad because it even affects you in airplane mode.

        This seems to be solving yesterday's battery problem - but any gain is good I guess.

        • In addition to this, another key part of improving transmission has been improvements in DSP processing power to lift usable signals out of the noise. Unless you can find a unique way to improve the efficiency of the algorithm they use, you're going to have to wait for a die shrink to improve power consumption of that portion of the cellular chipset.

        • by tlhIngan (30335)

          On more modern / powerful phones with larger screens, it is not the radio that is the dominant problem anymore. It is the screen.

          If you are "Streaming video" for any length of time with a 4.5" screen phone, it is the screen using most of the battery, not the radio. Screen battery use is also bad because it even affects you in airplane mode.

          Not to mention that cellular radios are extremely power efficient because they have to be. The power amp is only used for transmission, and only powered up the instant th

  • What's going to happen is, about ten seconds after this technology is deployed, the cell phone companies will halve the number of cell towers using the argument that the new tech makes them unnecessary. Your battery is again drained as weaker signals mean more amplification is necessary.
    • Re:Nah. (Score:5, Informative)

      by Neil Boekend (1854906) on Thursday November 01, 2012 @02:49AM (#41839485)
      Nope, because it doesn't increase the range of the antenna. The same limits to transmitted power apply, no matter what technology you use to transmit it.
      Besides, in heavily populated area's the number of antennas has nothing to do with transmit power, but with maximum throughput.
  • by complete loony (663508) <Jeremy.Lakeman@gma i l . c om> on Thursday November 01, 2012 @02:29AM (#41839409)
    If you're just going to pick a few sentences out of the article, you should at least talk about "who" and "what". All we've been left with in the summary is "problem description" and "hype"
  • by zippo01 (688802) on Thursday November 01, 2012 @02:36AM (#41839437)
    But what will keep me warm on those long lonely nights while I watching inappropriate things on my phone?
  • by Anonymous Coward

    IBM Watson group announce Watson has solved Smartphone battery problems! Using the magic of connected-world technology, Watson now developed batteries that think-smarter.

    Soon Watson will be moved to self-driving cars, where it will be given the far more difficult task of following a white line around an empty track at superfast speeds! The connected technology painted stripe we leverage allows our world beating Watson to go around faster* than the competitors!

    * You may not benchmark it and no stopwatches wi

  • This improves the standby and talk time, and may be network power consumption. Most of the other stuff the apps use - like CPU, GPU, sensors will not be any different because of this. So to claim double of battery life is exaggeration. It may double the standby time and probably improve the talk time by a considerable percentage.

  • by aXis100 (690904) on Thursday November 01, 2012 @02:53AM (#41839499)

    Pretty sure most of the power used is not in the radio - before "smartphones" we had phones with similar battery capacities achieving much longer standby times AND talktimes. Even if you turn off a smartphone's Mobile data and stick to Wifi (with only 30mW transmit required), battery life still isn't great.

    I think it's got a lot lot more to do with:
    - Big, bright displays
    - Multicore, gigahertz CPU's regularly kept busy with background apps
    - Far more sensors embedded in the unit to power - GPS, accelerometers, etc.

    • I agree with your general point, but non-smartphones do not have overall longer talk times than smartphones. They do have long standby times though.

    • by icebrain (944107) on Thursday November 01, 2012 @05:24AM (#41840135)

      I think it's got a lot lot more to do with:
      - Big, bright displays
      - Multicore, gigahertz CPU's regularly kept busy with background apps
      - Far more sensors embedded in the unit to power - GPS, accelerometers, etc.

      Plus, the whole obsession with "the phone must be THIN!!!1!"
      If the manufacturers quit worrying about trying to fit the phone into the form factor of an index card, there would be enough thickness for a reasonable battery.

      • by Aranykai (1053846)

        Pretty much every android device out there is now rated above 7 hours talk time. I don't know what planet you come from, but that is easily double what smart phones did 4 years ago when they weren't dominant of the cell market.

        Motorola actually did what you are suggesting in their Droid Razr model, offering one with a 2.5x capacity battery called 'Maxx" or some such. Very few people sprang the extra money for the extra battery because 8 hours talk time was already sufficient. Most people aren't power users

        • by icebrain (944107)

          I only got a smartphone two-something years ago. Prior to that, I had a dumb phone with extended battery; after three and a half years, I still got 3-4 days between charges. When new, I charged it once a week, and that was with heavy talk usage. So even today's smartphones seem to have short lives compared to that.

          I have an extended battery on my current phone (rooted Samsung droid charge), and I still have to charge it every day, at least during the week (but that's also because I now work in a large me

          • by plover (150551)

            Agreed. Phone thickness is not the dimension I have a problem with. Length, width, and mass are my most important constraints. The iPhone 3 and 4 is about my ideal size - the iPhone 5 is a bit too long for convenience and comfort. However, the iPhone 4 would certainly benefit from a larger viewable screen area - thinner borders, and using more screen area near the earpiece and home button.

            What I'd really rather have than a more massive phone would be a replaceable battery. I'd happily keep a spare batt

    • by Spacelord (27899) on Thursday November 01, 2012 @05:54AM (#41840269)

      It really is mostly the displays. On Android phones you can see what is using the battery, and it's almost always 60-70% the display.

      As for those multicore CPUs, modern smartphone operating systems are remarkably good at keeping them clocked down when they're not needed. As a matter of fact, if I leave my Galaxy Nexus unattended (i.e. don't use the display), there hardly is any battery drain. I wouldn't be surprised if it would last a whole week that way.

      • by Aranykai (1053846)

        Beg to differ, On my last two devices, display usually ranked 2nd or 3rd to cell standby and phone calls. Only on occasions where I had an extended game session or a movie did it ever break 30% usage.

      • If you leave your Galaxy Nexus unattended and plugged into USB to charge with the Facebook app running in the background, I'd be surprised if it lasted a whole day. The first time I installed the FB app, I didn't even use it and my phone started struggling for power (slowly losing charge WHILE CHARGING). Looked in data usage, found that in 5 minutes the FB app had pulled 4 times my total data consumption for the past MONTH from all other apps. Deleted that shitfest.

        In 1 hour I lost more than 5% battery

  • by TheInternetGuy (2006682) on Thursday November 01, 2012 @02:59AM (#41839523)

    Breakthrough Promises Smartphones that Use Half the Power

    Seems inefficient, wouldn't it be better if they used all the power?

  • Its about time we use our phone battery power for productive work instead of wasting it in a grossly inefficient piece of hardware: the power amplifier.
  • Class C (Score:5, Insightful)

    by GrahamCox (741991) on Thursday November 01, 2012 @03:19AM (#41839611) Homepage
    Class C RF power amplifiers can be ~90% efficient, because they drive a tuned load. That's been known for most of the 20th century. Is the problem that these need to be wideband amps? Perhaps there is a clever way to reconcile those needs, though I'm not seeing it.
    • Might be related to those phones supporting four to six different mobile network technologies operating in many different bands.
    • by Anonymous Coward

      The measure that matters is "energy per coded bit". In that regard, Class C amplifiers suck because they can only send one bit at a time. With clever coding (eg. QAM), linear amplifiers can send many bits at once.

      Linear amplifiers waste as much energy as heat as they create in RF. Engineers have dealt with that for many years. They did it for satellites way before there were cell phones.

      The various cell phone protocols are designed with power conservation (battery life) in mind. I would be surprised if

    • Re:Class C (Score:5, Informative)

      by Gordonjcp (186804) on Thursday November 01, 2012 @05:27AM (#41840155) Homepage

      No, the problem is that they need to be linear. You can use class C for FM (and therefore GMSK) because you're running at full carrier power continuously. For 3G, you need a linear amp because QAM has potentially got a variable carrier level. There are tricks you can do to get round this (envelope restoration) which could be what TFA is on about, but it's slashdotted.

      • by Andy Dodd (701)

        Yup. As a point of reference:

        A typical FM station's power amplifier is Class C and can achieve 70-90% efficiency. BTW, the new digital "HD Radio" is a nightmare for these guys, as while the OFDM signal used only needs 1-10% of the RF transmit power to achieve the same range, with typical amp technology back in 2003 or so, that meant 15-20% efficiency at best.

        For UMTS signals, which have a pretty high peak-to-average ratio, if a PA achieved 17-20% efficiency, that was REALLY good back then. There were a f

      • by MattskEE (925706)

        Based on the article it sounds a lot like envelope tracking:

        The new advance is essentially a blazingly fast electronic gearbox. It chooses among different voltages that can be sent across the transistor, and selects the one that minimizes power consumption, and it does this as many as 20 million times per second. The company calls the technology asymmetric multilevel outphasing.

        But looking back at one of their papers: http://www-mtl.mit.edu/~jldawson/Dawson_digest2009.pdf [mit.edu] it is not exactly envelope tracking

      • by slew (2918)

        The basic idea they have is somewhat novel, but requires some understanding on how modern output stage power amplifiers are designed.

        As mentioned by many, a "C" class amplifier can be designed so it's pretty linear over a certain range, but with variable-envelope modulation schemes used in modern communications, they waste quite a bit of power.

        On the other hand, classic pwm (pulse-width modulation) switch mode power amplifers (aka D-class amplifiers), allow trading pusle-width modulation for amplitude modul

    • Re:Class C (Score:5, Informative)

      by smpoole7 (1467717) on Thursday November 01, 2012 @06:57AM (#41840567) Homepage

      > Class C RF power amplifiers can be ~90% efficient, because they drive a tuned load. That's been known for most of the 20th century. Is the problem that these need to be wideband amps?

      You're on the right track, but the answer is a bit complicated. (The article, by the way, sounds like a PR piece for someone expecting to patent a technology that, by the same arguments we use against software patents, probably shouldn't be patentable, because it's an obvious rearrangement of existing technology. I shall elucidate.)

      1. Yes, it's difficult to run wideband amps in class C. Class C works best with a single frequency at a constant level.

      (To illustrate: my wideband HD transmitter, for example, must be re-biased to class AB. I can switch it to "pure" class-C FM mode and it puts out 2-3 times the power as when it's in HD mode.)

      2. As a general rule: designing an efficient amplifier becomes more difficult the higher the frequency. Wireless phones run at high frequencies.

      (To illustrate: class D switching amps have made it possible for your teenage son to have 1,000 watts of audio in his Nissan Sentra. But you must use a switching frequency that's much higher than the signal -- easy to do with audio, not so easy with even just an 800-900 MHz wireless signal.)

      3. Read the fine print and look beyond the smoke.

      We just went with Modulation Dependent Carrier Level (MDCL) on our 50,000 watt AM, and it has indeed cut our power bill substantially, which is strikingly analogous to what these people are proposing. But this is highly dependent on modulation (i.e., what we're "playing" at any given moment, whether music or voice) and other factors.

      In this case, if they're obtaining the higher efficiency by "smoothly" switching between standby and active modes, one wonders how efficient it will be during rush hour, when everyone is on the phone, calling their significant other to have dinner ready when they get home. :)

      • (The article, by the way, sounds like a PR piece for someone expecting to patent a technology that, by the same arguments we use against software patents, probably shouldn't be patentable, because it's an obvious rearrangement of existing technology. I shall elucidate.)

        There's a scam going on where people sell these big capacitor banks as "Power Savers" to reduce your electricity bill. They work on solid principles. Such things are used in steel mills where big driver loads run by heavy motors have a low power factor. Correcting the power factor greatly improves actual operating efficiency. Some of these mills shut down operations when the power saver fails because it's more expensive to operate without power factor correction than it is to idle the plant. In a resi

        • by smpoole7 (1467717)

          > If these people bring in shit that drops cell phone tower electricity usage by a significant factor and it actually works, it's not obvious

          But the phase correction that you're talking about is mostly used with big reactive loads. Yes, you can save a ton of money in that case. But except for the HVAC and (possibly) the UPS units that back up the power, that doesn't really apply to cell sites. Besides, the application discussed in the article is for the transmitters only. Has nothing to do with these ot

          • by Andy Dodd (701)

            The basic concept of modulating the supply voltage is obvious... Getting it to actually WORK properly without excessive distortion of the output signal is *not*.

            You might be able to build what they're talking about with chips from DK or Mouser - but the actual techniques for using those ICs in concert with each other without making the output signal garbage are not obvious.

            Seriously - you're going to need a $80k-120k spectrum analyzer and a $30-50k signal generator just to be able to properly test the syst

          • that doesn't really apply to cell sites

            Which is the point. All that shit doesn't apply to cell towers, but cutting back their power usage would be worth a lot of money. Therefor since we can imply great demand for an imaginary product--anything you can shove on-site at expensive private-owned major infrastructure and drop your continuous costs by enough to offset the cost of the magical widget is de-facto "in great demand"--we can imply that if that product does not exist, it is not obvious. If you know how to do it but it's extremely expensi

        • by dj245 (732906)
          There's a scam going on where people sell these big capacitor banks as "Power Savers" to reduce your electricity bill. They work on solid principles. Such things are used in steel mills where big driver loads run by heavy motors have a low power factor. Correcting the power factor greatly improves actual operating efficiency. Some of these mills shut down operations when the power saver fails because it's more expensive to operate without power factor correction than it is to idle the plant.

          This is onl
          • The point stands. If there's a way to save a business thousands a month per location in operating costs on electricity, and it's not been implemented, it's not obvious. With example.
        • In a residential setting, however, the general whole-house power factor corrector isn't helpful: most loads aren't the type that need tuning

          That, and I've heard that residential customers don't have to pay the power company for reactive load (e.g., imaginary power), just resistive load (e.g., real power). Industrial customers have to pay for both, which is why they build those capacitor banks.

    • Class A amps are the least effecient. Class AB is common for high fidielity, but still run in a transistors linear reigon generating heat.

      The high power car stereo industry has been using switch mode PWM transistors to drive high power sub bass for a long time. This is not new tech. It is new tot he cell phone industry.
      Class D is switched Mode PWM.
      Class G&H are variable Switched mode power supplies that vary the voltage provided to the amplifiers to increase effeciency.

      These classes do not have the f

  • by Anonymous Coward

    ...is here: http://dspace.mit.edu/bitstream/handle/1721.1/62479/712059703.pdf?sequence=1

  • A Gadget? (Score:5, Insightful)

    by jamesl (106902) on Thursday November 01, 2012 @05:14AM (#41840081)

    ... the power amplifier, a gadget that turns electricity into radio signals.

    Any article that calls an important piece of technology a "gadget" is neither serious nor credible.

    • by multiplexo (27356)
      You kids these days with your gadgets. Back in my day we had thing-a-ma-jigs, and whatcha-mcallits and we were darned lucky to have even those!
    • by mcgrew (92797) *

      Any article that calls an important piece of technology a "gadget" is neither serious nor credible.

      Why? What's wrong with "gadget?" Do you have a problem with the British calling scientists "boffins"? There's no difference between "device" and "gadget"; they're synonyms. Do you trash articles that say "count" rather than "enumerate"?

  • Envelope tracking (Score:5, Informative)

    by Iconoc (2646179) on Thursday November 01, 2012 @05:30AM (#41840171)
    What this article never really manages to describe is Envelope Tracking (ET). This has been in development for several years. Look at the diagram in http://www.nujira.com/technology-pa-746.php [nujira.com] for a better description of the concept. This article describes the application of ET in the handset.
  • That's odd, every time I look at my phone's battery consumption it always tells me that my primary power drain is the screen.

    • The battery isn't controlled by the OS. Each time a process runs, the OS stops another process and schedules the waiting process onto the CPU, setting up all the page mappings and registers and then context switching into userspace at saved EIP. It knows how much CPU is used for what task because it schedules it. As for the battery, it roughly understand how much power various components are supposed to draw and makes an educated guess.

      My Galaxy Nexus tells me Phone Idle 50%, Cell Standby 22%, Screen 15

  • "The technology" turns out to be called "asymmetric multilevel outphasing". No wonder the submission was too embarrassed to include this tidbit. Small problem, though. Nerds don't omit.

    Empathy represses analytic thought, and vice versa [eurekalert.org]

    Even for nerds it turns out that dog paddling through life in the default network generates more discussion forum page views. Somewhere a kitten dies.

  • If you ever go into your settings to see battery usage, you'll notice it's not the radios or the amplifiers, it's the *display* that uses the vast majority of the battery power. Sure, we need to get more efficient radios, but let's really go after getting the display technology much more power efficient, while at the same time coming up with a much more efficient battery technology than lithium ion or lithium polymer. Graphene and other battery technologies in the works look promising.
    • 50% "Phone Idle," 22% "Cell Standby," 15% "Display," 8% "Bluetooth" in 10 hours. 30% radio versus 15% screen and the amount of data transferred is under 5 megs.
    • by DeadCatX2 (950953)

      Samsung Galaxy SIII. Running on battery for 26 hours, 10 minutes. 61% charge remaining.

      Cell Standby 59% (battery used by cell radio)
      Android System 20% (battery used by apps)
      Device idle 7%
      Screen 6% (time on 45m 32s)
      Exchange 4%
      Android OS 4%

      So in my case, for the past 24 hours, the radio has consumed ~10x as much battery power as the screen (which was on 3% of the day, a not insignificant amount)

  • I always thought it was the backlight that used the power? My phone can be on standby for 3 days, yet 1 hour of ebook reading in flight mode kills the battery.

  • If you've noticed your phone getting warm and rapidly draining the battery when streaming video or sending large files, blame the power amplifiers

    No. I'll blame the bits that actually make heat, such as the screen backlight, and the CPU and GPU when displaying video.

    If the amplifier running at 50% efficiency when you're running maximum 0.125W for HSDPA or 0.25watt for GSM is causing your phone to heat up then you're doing it wrong!

    • If the amplifier running at 50% efficiency when you're running maximum 0.125W for HSDPA or 0.25watt for GSM is causing your phone to heat up then you're doing it wrong!

      No, no... you're holding it wrong. :)

  • Except for most modern smart phones, screens take up a bulk of the battery's power.

  • This sounds like old technology. They have been doing what sounds like the same thing in audio for years.
  • I haven't had to charge my phone twice in one day in over 5 years, when I moved away from a piece of crap Motorola Razr.

    This might come as a huge shock: don't play games on your phone for hours at a time, and you don't burn the battery down.

  • Not sure where TFA gets the idea this will double the battery life on my phone since the "power usage" page shows the vast bulk (70% or more) goes to powering the display, with low double digits (or even single digits) powering the various radios.

  • On first read I wanted to say: "yes please" then noted that hand sets would be years out.
    So good thing my provider contract is so far out.

    Then my brother called from the storm zone in NJ. The cell coverage was
    fragile and he noted that most towers in his area were down for want of
    battery power and no mains to recharge them.

    We need ethecical rules to limit calls to a short period and text only.

    We need more durable towers. We need "storm modes" that conserve batteries automagically
    and log a location that can

  • to that of Toshi Station power converters? I think I could manager to pick up a few on my way home before dinner.

Maybe Computer Science should be in the College of Theology. -- R. S. Barton

Working...