And now, my charging results. Bear in mind, these are totally unscientific, it's based on how fast I can charge the Switch via the USB-C PD port compared to charging it via the USB-A port delivering a 5V 2.4A charge.
Here are the ingredients:

The Xiaomi ZMI QB820 powerbank, the Nexus 6P, the USB-C to USB-C Apple cable approved by Benson Leung and Nathan K (model number MLL82AM/A (A1739), and the Switch itself.
What I did was discharge the Switch's internal battery, by connecting it to the Nexus 6P. The Switch now acts as a battery for the Nexus, with this notification coming up on the Nexus 6P.

The Switch was fresh out of the dock, at 100%. Drained its battery to 90%, but as I unplugged the Nexus 6P away from the Switch, the percentage drops to 89%, so I took it as 11% for the calculation later.
I then ran this scene in Zelda BOTW, with grass billowing in the wind, and Link's constantly beeping health warning and ambient music in the background. The display was set at maximum brightness and audio cranked up to maximum volume as well.

The Switch is then plugged into the USB-C PD port on the ZMI QB820 powerbank, and I started the timer on the Nexus 6P to measure the time it takes for the the powerbank to charge the Switch in this condition (max brightness, max volume, graphics intensive scene in Zelda BOTW with constant health warning beeping and ambient background music) to charge to 100% back again. It completed the full charge at 22 minutes 43 seconds. That's roughly 1% every 2 minutes.
From the article posted during the early days of Switch release and powerbank tests done with USB-A ports on regular powerbanks, it was stated that standard 5V 3A powerbank used in
this article by ArsTechnica only able to trickle-charge 1% every 6 minutes under unknown play condition (volume and/or brightness at what level, playing what game, and at which section of that game in?). This already beats expectation, but I had to do my own digging to be sure, so I tested it again, draining it again to 89% (for parity), and this time, charging it via the the QB820's 5V 2.4A USB-A port. After the same time (22 minutes 43 seconds) have elapse, it barely able to charge the Switch up to 91% from its previous 89%, a.k.a 2% within that span of 22 minutes.
Charging the Switch via third party USB-PD device has one quirk. You NEED to turn off your Switch, plug in your USB-C cable to the Switch, turn on your powerbank/external wall charger, before the Switch can negotiate a higher charging draw from the device that is charging it. I had previously plugged the Switch to the QB820 powerbank and the Tronsmart U5P USB-PD wall charger, and I did not get any "charging notification" at all and the Switch eventually slowly drains. I suspect the Switch did not recognise the USB-PD device that is attempting to charge it, and as such, only releases the safest charging profile available (I reckon it's 5V 0.9A charge profile, the default USB3.0 charge via PC USB port), which is too low to be recognised as a charge or fail to charge completely. Remember, reboot your Switch, plug in your third party USB-PD charger/powerbank.
These are the chain of devices that I currently have to make my Switch-playing time uninterruptible. Or, you can just carry the official Nintendo AC adapter with you and be done with.





The QB820, despite it carrying the Xiaomi's name associating itself mostly with being "affordable or cheap", feels the opposite of that. The outer shell is cold to the touch, confirming its metal escapsulating outer shell construction. Has 4 dots on the side to denote charge level left. And with the USB-PD Tronsmart U5P wall charger, I'm charging it rapidly as well. 20,000mAh in the size of a Nexus 6P. The only way this powerbank can be any more perfect is with more charge capacity. The powerbank itself doubles as a USB Hub when connected via USB-C, I don't know what other powerbank with USB-PD capable of charging a low power laptop that can do that!
Disclaimer: I'm not responsible for any blown Switches or any bodily harm that happens should you choose to emulate my setup here. I wish someone more knowledgable than me like Benson Leung or Nathan K can verify this setup as safe to use.
This post has been edited by stringfellow: Mar 18 2017, 05:01 AM