Two weeks ago Qualcomm introduced Quick Charge 4 with USB-PD compatibility. SemiAccurate already covered a lot of the Google/quick charging/USB-PD flap a while ago but there is a lot of new information now.
You might recall that in Google’s latest Android N compatibility definition guide seemed to indicate that quickcharging technologies that were not USB-PD on a USB-C interface was a bad thing. Some even read the guide as saying non-PD based schemes would soon be outlawed. As SemiAccurate explained, the truth is a lot less dramatic, it is just the way the industry way of nudging cost-sensitive OEMs to take up a technology that has merit in the long-term.
In mid-October Qualcomm had an event where they strongly suggested the next Quick Charge variant would be based on USB-PD. With the new revision it became official and QC4 is a superset of USB-PD. Any conspiracy theory about technical shortcomings of quickcharging technologies from Qualcomm will now hopefully be put permanently to bed, but this is the internet so look forward to 17 more debunkings in the next few weeks.
On that sad note we will move on to what Qualcomm’s QC4 actually brings to the table in a technical sense. The first part is that it now has USB-PD as it’s base meaning the two new QC4 chips can speak USB-PD. Those would be the SMB 1380 and 1381 companion chargers that put the dual in dual charge. The 1380 is a 6A part and the 1381 can only handle 5A, both for the battery charging side, not the adapter to phone side, that is handled by the PMIC.
This brings us to a point that most people don’t understand, the chain of events that make up the process of charging a phone. There are two fairly separate parts to the process, the link from the wall adapter/power source to the phone, and the phone to the battery. Quickcharge deals mainly with the link from the power source to the phone itself, or more precisely to the PMIC. This is what Qualcomm’s Quick Charge covers, not the PMIC to the battery link.
The PMIC to the battery charging side of the process is what most people think Qualcomm’s QCx does, but that isn’t really part of the mix. Most also believe that USB-PD also has an affect on the battery charging, it doesn’t. USB-PD and QC1-4 only cover the link from the wall socket or at least the adapter’s USB port to the PMIC itself. In more simple terms quickcharging technologies cover how fast the energy gets into the phone itself, not what or how the phone deals with the influx.
Four years and four iterations
Sort of. This is where another technology from Qualcomm comes in, one called Dual Charge. Starting with QC2, DC was introduced and with each following iteration it gained a plus symbol. Today’s QC4 is officially Dual Charge++, or at least it is an option for QC4. No credit if you figured out why the SMB 1380 and 1381 are called companion chargers, they are the dual part of dual charge. In essence they are a slave chip to the PMIC that halves the duty of charging the battery between two pieces of silicon.
Why? Two only need half the amperage each and so they run cooler. Cooler is more efficient. Better yet two physically distinct chargers can be physically separated to spread this lower heat from dissipation out even further. Cooler individually and with half the heat in two locations means lower skin temps for the device, a really good thing. In short the DCx idea is a good one with some pretty additive efficiency gains. We won’t go through the virtuous cycle of waterfall effects, just keep in mind that there are a lot of them.
That leads us to another technology in Quick Charge that came in with QC3 and is now up to v3 in QC4, INOV. Intelligent Negotiation for Optimum Voltage does what it sounds like and can signal the wall adapter to supply only the energy needed to efficiently charge the battery. If INOV does it’s job correctly there will be less wattage to bleed off as heat or overcharge the battery with. QC3 could step the voltage from 3.6 to 20v in 200mV increments, QC4 can likely do better but no specific steps were mentioned.
Like INOV and DC, Battery Saver Technologies and HVDCP all have been incremented, another + for the former and 1.0 for the latter. Those are minor changes, the biggest bangs are in the aforementioned USB-PD support and added cable quality detection. USB-PD has some rudimentary cable quality detection and more importantly an ID chip to specify the capabilities of the connection. As we all know these foolproof mechanisms take about 37 seconds to counterfeit so Qualcomm’s added cable quality detection is a welcome change at the wattage levels for USB-PD and QC4.
This brings us to an interesting point, the relationship between USB-PD and QC4. Qualcomm describes QC4 as a superset of USB-PD, a choice of wording SemiAccurate thinks is very apt. QC4 can talk USB-PD but it can step voltages up and down in a vastly more granular fashion and likely faster too. Better yet all QCx certified products must be UL tested as well so problems are less likely than with random USB-PD or alternative quickcharging schemes.
That brings us to performance and Qualcomm’s promise is that QC4 will be >2.5x faster than the crusty old QC1 from 2013. Compared to QC3 it is 20% faster, 30% more efficient, and up to 5C cooler, all with the ‘up to’ caveat of course. These are all fairly credible claims for the new generation but slideware numbers are a tad, well, not what SemiAccurate is all about. So here is a picture from Qualcomm’s 4G/5G Summit a month ago to give you a little more detail.
More technical detail is always better
That should give you a better idea of the benefits of QC4 and Dual Charge++, over 1W saved which according to the slide would result in a significant skin temp decline, more efficiency, longer battery life, and faster charging. All of this is now a superset of USB-PD so all the worries about Google cutting things out just went up in smoke. QC4 should be better in every way than QC3 and once the Snapdragon 835 devices bearing it come out, we can try it for ourselves.S|A
Latest posts by Charlie Demerjian (see all)
- HyperX ships it’s 60 millionth enthusiast memory module - Oct 15, 2018
- Bittware/Nallatech water cools 300W of Xilinx FPGA - Oct 12, 2018
- More on Intel’s 10nm process problems - Sep 17, 2018
- Intel puts out another 14nm 2020 server platform - Sep 11, 2018
- Why Can’t Intel Supply Enough 14nm Xeons? - Sep 10, 2018