Quick Charge users don’t have to worry about USB compatibility.
Benson Leung, a Google engineer, has been testing USB Type-C cables across the market to verify they meet the USB Type-C spec. He’s succeeded in getting some poor quality items off Amazon, which is very helpful to consumers.
In November he said Qualcomm’s Quick Charge (QC) technology couldn’t coexist with USB Type-C on the same connector because the latter violates the former’s spec:
While he’s certainly correct that QC violates the USB spec, the situation is lot more nuanced than that, and some of what he said is unfortunately incorrect.
I replied directly to that post, and some of the content of those replies follows.
Let’s get a few things straight about QC first, and then review Benson’s assertions using them:
Unlike the status quo for USB†, OEMs actually have to go through a UL (a worldwide saftey certification organization) certification process for QC, detailed in this Test & Certification Application (PDF). Per page 3 shown below, devices must maintain “the plug-and-play ease of use of USB connectors” to pass certification. This implies that there is no data transfer use case in which a QC device would behave any differently from a non-QC one and certainly none in which the QC device’s behavior would be unpredictable.
For all non-QC use cases, QC devices behave like any other normal USB device. Per page 3 above, QC certified devices must support “Conventional USB charging,” defined as “1A at 5V” (though I’m not sure how they get 15W from that. P = IV, so power should be 1A * 5V = 5W).
UPDATE: I contacted Qualcomm about this and got this useless reply:
Any information other than what is listed on our website (URL listed below for your reference) is Proprietary to Licensees.
Unfortunately we are unable to assist with your inquiry. However, we keep a list of compatible devices at Qualcomm.com/quickcharge. Hopefully this has the answer to your question.
Otherwise, we recommend you follow-up with a vendor that carries this product and seek their feedback on your technical questions.
Please note, Qualcomm is the technology provider, not a manufacturer of consumer products and therefore we are unable to answer your product specific question. We hope this direction helps.
Thank you for your inquiry,
Qualcomm Technologies Inc.
FWIW, in my experience QC (2.0) chargers slow charge non-QC devices and don’t (can’t?) fast charge them.
It follows from points 1 and 2 above that QC uses USB data lines for power only when connected directly to an A/C adapter, which has no use for a data line anyway.
While I can’t find any publicly available definition of the QC spec, it’s based on a patented technology called HVDCP (High Voltage Dedicated Charging Port). The “Dedicated Charging Port” detail name implies the tech works for ports that do nothing else but provide power and doesn’t work for data ports. Here’s the patent (which I haven’t read).
Fairly detailed implementation info can be found by searching “HVDCP” on Google. Here’s a very detailed product preview (PDF) by ON Semiconductor for their NCP4371 HVDCP controller.
Points 1 to 3 disprove Benson’s claim that the “Type-A port can’t be used to communicate to your PC at the same time you fast charge” via QC, as QC devices would fail the certification requirements if that were the case. In addition, implementing QC on a PC Type-A port would be disingenuous as it would disable data transfer for QC devices, thus killing most of the Type-A port’s functionality. No PC OEM in their right mind would do that.
The HTC 10, LG G5, and all other Quick Charge 3.0 and lower devices will work just like every other Type-C compliant device for every use case except when connected to a QC charger, in which case they behave like a QC device.
You’re probably wondering how we got to the current charging standards mess. I’ll probably explain that in a later post.
Despite the Verizon Samsung Galaxy S5 (SM-G900V) CPU (Qualcomm Snapdragon 801) being listed as Quick Charge 2.0 capable on Qualcomm’s site and claims to that effect on various blogs, the phone doesn’t support the charging standard in testing.
FAT32 and exFAT formatting were done using Storage -> Disk Management in Computer Management on Windows 8.1 Update 1 Professional 64-bit, while ext4 formatting was done using Linux Mint 17’s USB Stick Formatter.
The results are in, and they’re not that great:
Basically the S5 refuses to mount anything that isn’t USB 2.0 AND FAT32 or exFAT.** If you have any comments, counterexamples, or ideas please be sure to share.
*I didn’t buy a USB 3.0 OTG cable as I couldn’t find one online that was guaranteed to even fit the S5. I’ve never encountered that issue with USB before, so I have no idea if the fault is with the cable OEM or Samsung.
Also, part of my motivation for getting an OTG cable is to facilitate Sneakernet file transfers while out of the house. Since very few phones support micro USB 3.0, I decided to go the 2.0 route.
** I don’t own a Mac and so couldn’t try HFS Plus, sorry.
Recently a friend came into possession of an HP Compaq 6730b laptop: a 6+ year old relic running pre-SP1 Windows XP. After quickly determining that it wouldn’t support Windows 8.1, I decided to load it up with Linux Mint. Here’s how to do that:
Download the Cinnamon 64-bit (assuming you have a 64-bit CPU) ISO here. Ignore the other builds: a general rule of thumb with Linux is the more exotic you get, the more problems you run into. Keep it simple.
In Windows 8.1, open the folder containing the ISO file.
In the View tab, check File name extensions.
Change the .iso in the ISO filename to .img.
Plug a USB flash drive of capacity 2 GB or greater into your PC.
Run Win32 Disk Imager.
Click the file icon in the Image File section.
Navigate to the renamed .img file in Step 5.
In the Device menu, select the name of the drive in Step 6.
When that’s done, remove the USB flash drive from your PC and insert it into the PC onto which Mint is to be installed.
Reboot the install PC.
Set the USB drive as the boot device (how to do this varies from PC to PC). Mint should fire up from there and give you the option of installing it to the hard drive, which you should then do.
If your Wi-Fi doesn’t work after installation, try resetting the PC’s BIOS to factory default settings. That worked for me.
It’s fast. The old laptop boots to a usable desktop faster than my Core i7-3770 powered Dell XPS 8500.
Very low RAM usage.
Excellent legacy hardware support.
It installs quickly.
Peripherals (assuming they’re recognized by the OS) install and configure rapidly.
You can get up and running with word processing, email, web browsing, etc. in no time.
Easy to nonexistent learning curve for basic functionality.
Read NTFS storage out of the box.
No annoying, unmovable dock (unlike Ubuntu).
No need to worry about security software.* Unfortunately, this pro is nearly nullified by a con below.
Since most of the ways to hose the system involve command line use, you can hand it over to less experienced users without worrying about them borking it.
Very familiar to Windows users.
Aero Snap is better than it is in Windows.
Can play DVDs out of the box. (Ironically, Windows lost this feature in Windows 8+.)
Documentation is thin to nonexistent for anything more advanced that email, productivity, and web browsing.
Limited default software support compared to Ubuntu, positively woeful compared to Windows. While LibreOffice worked, I couldn’t get OpenOffice to install, for example.
As with many distros, solutions posted online often lack important details.
Samba just doesn’t work.
Doing anything more than the basics still requires using the terminal or manually editing config files (I’m a GUI guy).
Although there are no antivirus worries, there’s also no way to check for and/or automatically install updates; it has to be done manually. This produces a significant security risk, especially if you’re passing the machine on to non-technical users who’re unlikely to check for updates themselves.
The menu for adding/removing users doesn’t update immediately, so deleted user accounts show up until you close and reopen the menu. Confusing and annoying.
Dated UI that looks like a reimagined Windows XP.
(For Firefox Nightly users) Nightly builds are few and far between compared to Windows.
Little annoying things. For example, while Aero Snap Windows key combos work amazingly well, some basic Windows key combos with obvious analogs in Mint don’t work.
I’ll be honest: I grew up on Windows, and the way it does things just makes more sense to me. It’s not that Linux Mint is bad, it’s just that it would have to offer some significant advantage over Windows in daily use for me to switch. Unfortunately, it doesn’t, except in the following cases:
You absolutely need Linux for something that can’t be done on Windows.
You have older hardware that doesn’t meet Windows 8.1’s system requirements AND don’t want to buy a new PC. Seriously, if you can stand the cost of a new machine, just go that route instead and eBay the old one.
That said, it’s the best Linux desktop experience I’ve come across yet, and my friend seems pretty happy with it.
*This is mostly because there isn’t much Linux security software out there, not because Linux malware doesn’t exist.
I’ve had a few issues myself over the years, ranging from printers suddenly vanishing from the OS to mice that refuse to wake when the PC is resumed from sleep. I once returned a Belkin hub under warranty 3 times in a row because devices connected to it would disappear after the PC had been running for over a day or so.*
Fortunately, Microsoft has an entire blog devoted to USB and its attendant issues. Unfortunately, like most good things at Microsoft, it isn’t promoted at all. The blog is great for troubleshooting, but it’s also highly informative for power users who want to know how Windows handles USB under the hood.
*The root causes of the above issues are Link Power Management & Selective Suspend (and their equivalents on other desktop OSes). Both features allow the OS, the host device, and/or the client device to move to very low power states so as to save energy. Apparently, their implementation can vary at the OS, host, and client levels, resulting in odd behavior.