Qualcomm Quick Charge 3.0 adjusts voltages for faster refuels

Qualcomm's slow trickle of information related to the Snapdragon 820 SoC continues. Today, we're learning about Quick Charge 3.0, which the company says is its "fastest, most efficient charging technology to date."

Quick Charge 3.0 uses an algorithm to determine the power level required to charge the battery at any given time, which Qualcomm calls Intelligent Negotiation for Optimum Voltage (INOV). The voltage can be selected in 200mV increments from 3.6 to 20 volts, compared to a max of 15 volts in Quick Charge 2.0 devices and chargers, such as the Moto X Pure Edition and Moto's TurboPower wall charger. The company says these changes are good for up to 38% faster charge times compared to Quick Charge 2.0. 

Similar to Quick Charge 2.0, the wider range of voltages in Quick Charge 3.0 requires new chargers for users to realize its full potential. Older chargers will still be able to charge Quick Charge 3.0 devices, just not as quickly. Qualcomm says the new spec is compatible with any USB physical interface, though, including the now-common micro USB and reversible Type-C connectors.

The Snapdragon 820 will, of course, support Quick Charge 3.0. Several other upcoming SoCs will include the spec, as well—the Snapdragon 620, 618, 617, and 430 will all have it. Qualcomm says to expect Quick Charge 3.0 in "next year’s round of smartphones and tablets."

Ben Funk

Sega nerd and guitar lover

Comments closed
    • Chrispy_
    • 4 years ago

    Fast-charging Li-Ion batteries is bad for their long-term health. That’s certainly the case with dumb Li-Ion 18650 cells and packs and I’m well-versed in all the underlying physics, chemistry and degradation of said cells over time, as well as how degradation is minimised and how not to have your 18650’s explode, vent or catch fire. I’ve done this research as I have a vested interest in not having my face blown off by battery packs on my bike frame and bars.

    Am I wrong to assume that Li-Ion cell knowledge can be applied to microprocessor-controlled Li-Ion packs in smartphones and tablets?

    As far as I can tell, no 2-year-old battery is anywhere near as good as it was when it left the factory, so the last thing you want to do is shorten the life unncessarily.

      • ludi
      • 4 years ago

      Battery tech isn’t sitting still. For example, there have been studies in how to modify the electrodes or the charging cycle to avoid the damaging changes in electrode volume that are associated with failure:

      [url<]https://www6.slac.stanford.edu/news/2014-09-14-study-sheds-new-light-why-batteries-go-bad.aspx[/url<] There has also been a lot of recent research into alternative battery designs and materials: [url<]http://www.sciencedaily.com/releases/2014/10/141013090449.htm[/url<] [url<]http://phys.org/news/2015-04-ultra-fast-aluminum-battery-safe-alternative.html[/url<]

        • Chrispy_
        • 4 years ago

        I know, but studies and research into future tech doesn’t change that these fast chargers are working on bog-standard Li-Ion packs, which don’t like being fast-charged.

          • siberx
          • 4 years ago

          Quantify “don’t like being fast-charged”. How fast is fast, how new is the cell chemistry, is it a prismatic or pouch cell, etc…

          Just saying that fast-charging shortens life isn’t a useful statement, because it matters [i<]how much[/i<] the life is shortened for a [i<]given[/i<] charge rate (which may or may not be considered fast depending on the cells and circumstances). One cell of a given type may be able to charge at 2C and achieve 1000 recharge cycles, while another cell might only reach 1000 if charged at 0.5C (cell chemical purity/quality plays a factor here). Depending on cell size and geometry, the currents and wattages involved in those charge rates will also differ - so a given current might correspond to a much faster relative charge in a smaller phone battery than a bigger one in a tablet. Just to think about some numbers though, most lithium ion cells stay reasonably healthy at a charge rate of 1C. In a modern smartphone, this corresponds to a charging current up to 3A in many larger phones (3000mAh battery) and a peak charging wattage of 12.6W assuming zero loss. The higher capacity the batteries get the more amps can be pushed into them safely for a given chemistry. Given the above, it's not reasonable to make a blanket statement that fast-charging at the rates enabled by the above standards are necessarily unhealthy for the cells in question. It depends on the packs in question and how they're managed.

            • Chrispy_
            • 4 years ago

            Well, LCO is the common chemistry for mobile devices and that’s safety-capped at 1C maximum, charging it at 1C is ludicrious – risking fire through overheating and undue stress. The protection circuit cutoff is 1C and most manufacturers recommend 0.8C as a maximum safe charge rate. Unprotected cells that are charged over 1C yet do not hit thermal runaway at about 150C are severly damaged because the self-heating rate is proportional to the temperature. It’s usually best to discard them as their cycle life can drop by 75-90%

            The cycle life of LCO is governed by depth of discharge and temperature, and temperature is the killer that reduces life to below 500 cycles, in theory a high-spec LCO cell can survive 500 discharges down to 2.5V as long as it never charges/discharges above 1C

            Yes, there are exotic battery chemistries that can be fast charged well in excess of 1C:

            LMO – can tolerate 3C charges but is has lower specific energy and shorter cycle life, so useless for smartphones.
            LTO – can be fast charged at up to 5C but very low specific energy ; even more useless for smartphones than LMO
            NCA – good LCO alternative at 0.7C charge but can be fast-charged without incurring damage only when new (first 50-75 cycles)

            This is all public domain, not guesswork or speculation. At best, fast charge is only applicable to 70% SOC, meaning that there’s increased likelihood of a deep discharge, [i<]which shortens the lifespan of the battery.[/i<] Whichever angle you look at it from (thermal and internal corrosion stresses, or typical discharges 30% deeper than with a 100% SOC-charged battery) you are shortening the life of the battery. Now, how much shorter is important, because some battery life degradation is worth it for the convenience of fast charging, but both LCO and NCA are in the 500-cycle range and that's only 18-months if you assume people charge overnight, and that's before fast-charge degradation has reduced that 500 to maybe 400 or fewer cycles.

    • siberx
    • 4 years ago

    That’s actually a pretty clever idea. Typically, unless a device builds in a buck converter it is otherwise limited in terms of its charging capabilities by the power losses associated with dropping from supply voltage (such as 5V) to whatever the target charging voltage is (4.2V for single-cell Li-Ion, for example).

    Since these losses are proportional to current (charging rate) dissipating the waste power becomes more problematic the faster you charge. INOV allows the device to select a supply voltage that provides a sufficient delta from the charging voltage to provide reliable regulation while keeping ohmic losses to a minimum, allowing higher charging speeds with lower hardware cost in the device.

    To give an example of the numbers involved, a device trying to use a 5V/2.1A power supply to charge a single-cell Li-Ion at maximum current will have to dump 1.7W of waste heat, which would not be atypical for a whole device’s typical heat output except this is concentrated in a single tiny voltage regulator chip somewhere on the board.

      • ludi
      • 4 years ago

      Nearly all devices use switching converters these days. Being able to reduce voltage to battery levels for trickle charging is a small benefit, but what this is resolving is the fact that USB connections are limited by the maximum allowable current through the power pin. If the limit is 2.1A and the voltage is 5V, then the maximum power that can ever be supplied through the connector is 10.5W. If the voltage can be raised to 20V, then potentially more than 40W can be supplied on request.

        • siberx
        • 4 years ago

        Higher voltages obviously allow higher power to reach the device, but QuickCharge 2.0 [i<]already[/i<] allowed a device to negotiate 5V/9V/12V charging voltages (and thus correspondingly higher wattage). If it was simply a power delivery concern, why wouldn't Qualcomm have simply extended QC3.0 to support a 20V profile instead of permitting fine-grained 200mV selection of any voltage in that range? It's certainly more costly to implement adapters that support that kind of fine voltage selection and if we're talking switch-mode converters, it doesn't make a huge difference to efficiency if your source is 7.4V vs 7.6V for example.

          • ludi
          • 4 years ago

          Perhaps because once the logic has been added to negotiate a different voltage with the host device, it’s trivially easy to add signalling for stepping the voltage. Also, switching converters are not 100% efficient, and that heat does get dissipated into the device, so allowing the load to self-optimize the VI curve can still minimize losses, even if they are smaller losses than would be had with an equivalent linear regulator.

          Particularly if the load can pull a maximum of 40W or so at 20V, where an efficiency difference of 1% results in a 0.4W change in waste heat.

    • UnfriendlyFire
    • 4 years ago

    Hopefully we don’t get any battery fires from this.

    • Duct Tape Dude
    • 4 years ago

    I don’t like proprietary standards, but luckily Qualcomm is so ubiquitous it doesn’t really matter. It’ll be nice if they make this an open specification.

Pin It on Pinterest

Share This