Shared publicly  - 
 
WHY is Vbus HOT on a USB-C receptacle/cable so dangerous? (or, WHY are 'degenerate A-A' cables bad?)
tl;dr: PELIGRO! You can fry your USB ports, charger, motherboard or device.

Benson has discussed this but never quite explained HOW it causes physical damage on #USB #TypeC . (It's a topic people who are proficient with using jumper cables would grok.) It is one of his least-commented FAQ entries despite being one of the most important.

FAST PUBLISH: Because of alarming feedback from users during the 'USB-C Charger Shootout' I am rushing this article. Please forgive errors, omissions, or lack of flow. (Corrections, comments, or clarifications are welcomed.)

This article has been split into two for clarity. See the companion article expaining HOW TO CHECK here:
https://plus.google.com/102612254593917101378/posts/dgLwqvACbJm

Vbus HOT on USB-C Receptacle and Degenerate A-A cords - Destroyer of Motherboards

tl;dr: "I was told there would be no math." Just look at the table image.
https://en.wikipedia.org/wiki/USB#Power-use_topology
http://electronics.stackexchange.com/questions/23943/what-would-happen-if-i-connect-two-different-dc-voltage-sources-in-parallel

The issue arises because the "5v" the charger generates isn't quite exactly the same as the "5v" of anything else. When connected, the two try to push voltage backwards across each other/short cicuit. When you do that to a a regulated power supply (like a computer), it tends to "let out the magic smoke".
https://en.wikipedia.org/wiki/Magic_smoke

This is where the math begins. Second image I've included a diagram of something very important to this discussion: a Thevenin Equivalent Circuit. (If your eyes are rolling back in your head, bear with me. I'll explain in a moment.)
https://www.ibiblio.org/kuphaldt/electricCircuits/DC/DC_10.html#xtocid230279
https://www.ibiblio.org/kuphaldt/electricCircuits/DC/DC_10.html#xtocid230272

Basically, it is an approximation of how a charger reacts. The more you load it, the lower the voltage goes. (In reality it doesn't go down in a straight line -- it's more of a knee-shape -- but close enough.) Note how this TEC can roughly simulate USB ports on a motherboard: usually you have one 5v source for a bank of 4 USB ports that share a single controller chip. (In reality each port's power is independently controlled, but from my tests some powered hubs do not.)

In the third image, I've shown what happens electrically when (a) you connect an Vbus HOT device/cable to a computer and (b) what happens with a *compliant* USB-C device/cable. (This can be a C-C cable, an A-to-C, or a captive-cable-charger.)

https://plus.google.com/+BensonLeung/posts/hF6b228zGvy

Normally, as Benson explained, compliant USB-C devices can tell this is happening and simply don't connect the voltage lines. (In the case of captive cable charger, you are relying on the computer's USB-C port. Note: this means a computer's USB-C ports can be unsafe/Vbus HOT too! Dun dun dun....)

-- At the top of the third image I used TEC's to approximate the current that would go through your motherboard if connected. ("Mesh current".)

-- At the bottom of the third image I show what happens when everything is USB-C compliant. *Absolutely nothing!* Great! That's the way things are supposed to work.

I did an example mesh current analysis on the table in the first page and you can see how messy it got. I omit the 3-loop version here with ground offsets for sanity. (Plus there are more unknowns.)

The key takeaway:

Any ONE Vbus HOT device in the chain can opens up a giant can of worms that can potentially cause damage. This is why unlike voltage sag, Benson gives 'A-A degenerate cables' the lowest possible rating. (Short of that one cable that blew up his Chromebook Pixel... I'm pretty sure he had that one stuffed and mounted.)
http://www.amazon.com/review/R2H1CUPXWPC972

Personally, I propose Vbus HOT cables/adapters should be entered in the same category as A-A cables for the exact same reasons.

Likewise, if using USB-C compliant devices, there is (almost!) nothing you can do that will intentionally cause a dangerous failure. (Except maybe plugging in USB-C cables with your hands wet, according to Apple.)
https://goo.gl/kA6vkD

Any manufacturer eschewing these precautions the name of "compatibility" should be met with the greatest scepticism. Safety experts may argue these adapters should not exist in nature at all... I say let them exist, but store them alongside the sharp pointy objects, toxic chemicals, and ETHERkiller cords.
http://www.fiftythree.org/etherkiller/
http://hackaday.com/2015/10/10/the-usb-killer-version-2-0/

#USBC
11
2
Achiya Elyasaf's profile photoNathan K.'s profile photoOndřej Pokorný's profile photo
13 comments
 
+Achiya Elyasaf​​​ Is that charger Vbus HOT? Have you tested it? I have not.

Your phone should be fine in the Type-A receptacle using a legal A-to-C cord. Type-A receptacles (holes) are always Vbus HOT. This is OK. 6P will be limited to 1.5A using an A-to-C cord.

Type-C receptacles (holes) should NOT be Vbus HOT. If Type-C is Vbus HOT, the danger comes because you can plug an A-to-C cord into itself and possibly damage the charger or cable.
 
According to benson (https://plus.google.com/+BensonLeung/posts/HJifLTzKmgo one of the last comments) it is VBUS hot (though I didn't find any official report).

So it is dangerous to plug an A-to-C cable from the type-A receptacle to the type-C receptacle.

However, if I only use a C-to-C cable to connect my Nexus 6P to the type-C receptacle, and an A-to-A cable to connect my old phone, say Nexus 5, to the type-A receptacle - then my phones should be fine as well.

Thanks!
 
+Achiya Elyasaf I like to think of it like a car manufacturer designing their car so there's no cover over the gas tank.

Yes, you are probably not going to have an "accident" if you are diligent. But because it doesn't have that safety, it indicates (1) the manufacturer cut safety corners, perhaps in other places too and (2) an "accident" WILL happen sooner or later, instead of being impossible to happen.

I damaged my tools on a Vbus hot charger only a few days ago. That caused $50-80 in damage. For a $15 charger. Not worth it.

Also, you are paying good money for a cut-rate product. If you pay for a USB-C charger, Vbus circuitry should be included for the money. Since the manufacturer is cutting costs by skipping it, you are being cheated, indirectly.
 
+Nathan K.​ Thank you very much for your answer.
1. I already bought the C2PTU charger, so I will ask tronsmart for a refund.
2. The one you mentioned has only one receptacle and I want to simultaneously charge my wife's Nexus 5. 
 
+Ondřej Pokorný The Belkin one does have only one port, yes, but it follows the USB-C and USB-PD standards and is legal.

The Tronsmart one, if that quote Benson posted is accurate (and I have no reason to doubt it, he just hasn't made a G+ post about it) is bad and potentially dangerous.

The watt rating doesn't matter if the charger itself is bad/violating safety protocols.

+Achiya Elyasaf The only way I would regard the C2PTU as "safe" is if it has had a USB C-C cable permanently glued/attached to the faceplate. Then it becomes THIS, which is compliant/safe:

http://www.amazon.com/review/RX4U3GW1TUBNH

If returning is absolutely not an option, use some epoxy and glue the outside (not inside!) of an inexpensive USB 2.0 C-C cable to the charger to prevent any "accidents".
 
+Nathan K.​ I actually recently ordered the one you've just linked, with integrated cable, verified by Benson. This is why I'm asking if the Belkin is really more powerful but didn't want to make things more complicated so I've asked about the one supposedly non-compliant. :)

Benson noted in the G+ post ( http://plus.google.com/+BensonLeung/posts/HJifLTzKmgo ) that the power output was reduced to 33W after the fix. That made me believe that the C2PTU has the same guts as the updated CCTA Benson 5-starred because the advertised parameters are the same in every way except it has the USB C port instead of the cable. But if +Benson Leung​ says it's vbus hot, I believe him. :) Maybe we're both right? Hadn't been the cable integrated, the 5starred charger would be illegal as well despite the same guts?
 
+Nathan K.​ Thanks, I will try to return the charger.
What did you mean by USB 2.0 C-C? I thought that C-C cables are always 3.1. Will it be unsafe to glue a 3.1 C-C cable? 
 
+Achiya Elyasaf I'm glad you noticed! No, not all Type-C cables have to be 3.1 (Gen1=5 or Gen2=10gbps). Some can only use USB 2.0 wiring! This is explained in Benson's post here:

https://plus.google.com/+BensonLeung/posts/HBWyVBmdVzw

Yes, due to the mistake Tronsmart made, using a 3.1 C-C cable will be unsafe. USB3.1 cables require good wiring. Tronsmart's wiring is bad. Using USB 2.0 C-C will "cover up" the bad wiring.

tl;dr: Basic 5/6-wire USB 2.0 C-C cables are good for 3A. Don't require anything fancy. Dumb copper wire. Good for 60W/3A.

Fancy 15-22-wire USB 3.1 C-C "full featured" cables. These have a chip called an eMarker. Requires different wiring. Also called "USB3.1Gen2" or "10gpbps superspeed" cables. These are also good for 60W/3A.

Super-ultra-fancy cable are called "100W/5A capable power delivery eMarked cables". These are USB3.1Gen2 cables tested extra well, have a special eMarker. Good for 100W/5A, but are expensive.

Also: "Thunderbolt 3" cables. Just marketing fluff. Different logo and special branding. (Almost) identical to "USB3.1Gen2 10gbps" cables. Also have an eMarker.  (Some "special" ones can do 40gbps.) If you can't find a USB cable, Thunderbolt 3 cable will work fine. But comes with Intel tax.
 
+Ondřej Pokorný  The Belkin supports USB-PD at 9v/3a=27W. At 5v/3a it does 15W. You will not get the benefit of USB-PD 9v unless your device is a USB-PD device that supports >15W charging.

The Tronsmart power wasn't "reduced". The internals are probably the same, they probably just fixed the labeling. 18W QC over Type-A + 18W QC over Type-C [BAD] = 36W. 18W QC over Type-A + 15W 5v/3a over Type-C [legal] = 33W. See the difference? It's honest advertising. Plus they got rid of the BAD voltage changing circuitry.

The reason the version with the cable works/is legal, and the one WITHOUT the cable is bad, is because having a fixed USB 2.0 C-C cable "covers up" their engineering mistake inside the charger.

This is why I am advising people to "make" their noncompliant Vbus HOT chargers corded by gluing in a USB 2.0 C-C cable. This also "covers up" the mistake. Keep in mind it has to be a USB 2.0 C-C cable, because a USB 3.1 C-C (with extra wires) will "uncover" the mistake again. I explain why in this post here:

https://plus.google.com/102612254593917101378/posts/7Cjq1yS2CHz
Add a comment...