Shared publicly  - 
This is a formal notice of correction for one of my previous analyses, and a public caution. The Huawei Nexus 6P charger has been removed from listing of recommendations until further notice.

tl;dr: Some stock Huawei OEM chargers included with the Nexus 6P go Vbus HOT when used with eMarked cables. Not all 6P chargers appear to be affected. Only in very limited circumstances (such as mine) will this defect cause damage to devices or accessories. If you only use a (1)non-emarked (2)USB-C 2.0 cable, such as the one included with the phone, and (3)only with the 6P, you will not encounter this issue.

It is currently the official policy of the Google Store that since the charger "works fine" with the cable included in the box, they do not consider this a charger defect or damage caused by using "third party" eMarked or USB3.1 cords a warranty-able item.

*Note:* Google's statement is in parity with the statement OnePlus gave regarding their noncompliant USB-C cables that would cause damage to devices.

Until the affected chargers can be identified, I urge caution to only use the USB 2.0 cable included with the phone itself. Any/all of the below cables will expose this bug with the charger:

* [USB-IF Certified] Plugable USB-C to USB-C Cable (3.3 ft / 1 m, USB 3.1 Gen 2 Type-C, 3A)
* Belkin USB-IF Certified 3.1 USB-C (USB Type C) to USB Type C Cable, 3 Feet / 0.9 Meters
* Belkin USB-C to USB-C Cable (USB 3.1)
* 0.5m (1.6 ft) Thunderbolt 3 (40Gbps) USB-C Cable - Thunderbolt and USB Compatible (any length)
* JUCX01 USB 3.1 Type-C to Type-C Coaxial Cable (20v/100w capable)
(Any USB3.1 USB-C cord, according to the spec)
(Any Thunderbolt 3 cord, according to the spec)
(Any USB-IF certified cord, according to the spec)

The matter is complicated by the fact that EMCA USB-IF certified cords are "safer", and going forward all USB-C cables will be eMarked as a result of previous bad cables. (Catch 22.)

Out of the interests of public safety, I am disclosing this after having previously notified the manufacturer and received an official response. Technical details are withheld until the situation can be addressed better.
#USB #TypeC #USBC  
Fida Mehran's profile photoKai Sheng Ooi's profile photoNathan K.'s profile photo
+Nathan K.​, I know this not may be related to this. But just meant to ask. Does the Monoprice still stand a good charger for Nexus 6p?
+Fida Mehran​ Yes, electrically it seems fine to me aside from the Type-A ports sharing 2.4a. However I haven't found any other charger that has a flat voltage output curve like the 6P one. I'd still suggest the Google 22.5W charger first and foremost.

Another note, I spoke with Benson at length today. Based on his anecdotes, and the clear need and safety implications for unbiased charger analyses, I've decided to resume testing cables/chargers from Amazon, even if at my own expense. This new format and users deserve better than what bad manufacturers are doing to it.

I'll just eat the costs for now if items are defective/noncompliant. I've ordered the SoneIC 6P-clone charger and plan to see if it is affected. If not, that will replace the Huawei charger.
+Nathan K. Two Thumbs up, brother. Keep up the good work. It's good to know that you and +Benson Leung​ are looking after us at your own expenses. This leads to believe there's still goodness in this world.
+Nathan K. When testing for Vbus HOT using the LED method on the UK 6P charger together with the stock cable, the LED lights up. According to your post, the stock USB 2.0 cable shouldn't exhibit such a behavior, right?
+Kai Sheng Ooi I am extremely surprised - this means the problem is a lot worse than I thought! I thought it only happened for the 6P on eMarked cables. Are you saying it is happening with a non-eMarked A-to-C?

Are you using the A-to-C that came with the phone? Can you please post a picture of the bottom of your UK 6P charger (showing the serial number and maker)?

Can you send a picture of the LED lit using your setup, if possible? You need to use a 330-1k ohm resistor on one leg of the LED so it doesn't burn up while you are focusing the camera to take the picture.
+Kai Sheng Ooi​​ The LED only being dimly lit may be a good sign, I hope. +Benson Leung​​ recently informed me of some changes to the USB spec in a related section. The definition of "0v" (or "vSafe0v") is actually "0v to 0.8v". (Table 7-24 Common Source/Sink Electrical Parameters, USB-PD R2.0 v1.2)

Depending on the LED you are using, its forward voltage drop, and the charger's leakage current, you may be seeing this. (At least, that's what I hope.)

Let me try the same experiment with my 6P chargers (I had gotten a couple for testing purposes) and share the data back.
+Kai Sheng Ooi I think the best way currently to test for Vbus HOT chargers is to plug it in backwards into a USB-A hub. Leakage current alone won't have enough power to turn an entire hub on. (I'm tempted to remove the LED method based on your data.)
+Nathan K. Thanks for your explanation! Does this make the charger any more unsafe, or it'll be fine if I stick with USB 2.0 non-eMarked cables?
+Kai Sheng Ooi This can lead to damage in some situations -- those situations are just very difficult to demonstrate. Unlike previous demonstrations I've made where the situation is plain as day -- such as plugging in a laptop then a phone -- the situations with "leakage voltage" and "vSafe0v being violated" are much more nuanced. (The Nexus 6P charger spontaneously turning on with eMarked cables is a much worse problem.)

This is specifically addressed in [ Connecting Sources and Sinks]. Please note "shall not be harmed" doesn't mean it won't be harmed -- merely it shouldn't be if following the rules. Since the N6P charger already breaks the rules, the sentence becomes tautological.

"In the cases where no function results, neither port shall be harmed by this connection. The user has to independently realize the invalid combination and take appropriate action to resolve. While these two invalid combinations mimic traditional USB where host-to-host and device-to-device connections are not intended to work, the non-keyed USB Type-C solution does not prevent the user from attempting such interconnects. VBUS and VCONN shall not be applied by a Source (host) in these cases."

The Nexus 5X and Nexus 6P chargers already violate this (with all and eMarked cables, respectively).

Moving on, according to the exact letter of the spec, this leakage voltage of 4.80v is another, separate violation. Please see [ Exiting from AttachWait.SRC State].

"The port shall transition to Attached.SRC when VBUS is at vSafe0V and the SRC.Rd state is detected on exactly one of the CC1 or CC2 pins for at least tCCDebounce."

So by definition, if the N6P charger was following the spec, it should NEVER turn on, because Vbus is always above "vSafe0v" (0 <-> 0.8v, leakage voltage is 4.80v.)

See what I mean? The manufacturer will claim it poses a "theoretical" problem, and likely will do nothing to correct it. But it clearly presents unhandled situations where the charger CAN cause damage if used with ANOTHER device that IS following the rules, in limited circumstances.

So therefore I'd have to answer your question by saying: the Nexus 6P charger is unsafe to use. It is not USB-C compliant, and as such I cannot recommend its use, or intermingling with any other USB-C compliant devices or cables. I suggest purchasing a charger that is USB-IF Certified, UL listed, and follows proper safeties.

(However -- practically -- such a charger does not exist yet. USB-IF Certification of chargers just began. The best closest ones I've tested are the Anker PowerPort+ 5 and the dodocool 30W USB-PD2.0 [5/9/15/20v]. Neither are UL/USB-IF Certified. For USB-PD1.0 [5/12/15/20v], the Innergie 45W is UL certified and great option.)

If you want to maintain your warranty, however, you HAVE to use the unsafe OEM charger with your Nexus phone. If you use a "third party charger" the manufacturer will attempt to void your warranty coverage. In the US we have rules saying manufacturers can't do this unless they prove the third-party accessory caused the damage. But companies widely ignore this requirement and will void warranties anyway. So I leave the final decision to you.
Add a comment...