Hacker News new | ask | show | jobs
by randallsquared 3783 days ago
USB C is not a passive signal cable.
1 comments

Tomay-to/tomay-to.

Yes, USB C can also be used to provide power (which is also a form of signal), but again my point is that a cable like this is largely expected to be a passive component.

That a crap charger can do bad things would probably surprise no one, but hooking a good charger up to a good unit shouldn't be able to fry anything, ever.

Wasn't USB C supposed to make our lives easier? If we need to ensure all our USB C gadgets, cables and chargers are all not-exploding certified, in parts and in combination, I may just go back to plain old regular USB.

Hooking up a good charger up to a good unit with the ground and voltage wires reversed can fry things.

You can try it with your laptop. Cut the cable after the transformer, flip the ground and live wires and start it. It will get fried. Is this the fault of the specification or the person who intentionally flipped the wires?

The manufacturer was the person flipping the wires in this case.

In this case, it's DC over the USB-C cable. Reverse polarity protection is 1) easy and 2) cheap. That's the most striking part to me, and I almost thought the original post was a joke... because it seems insane there wouldn't be reverse polarity detection built-in to the laptop.
According to the engineer who tested the cable:

"Most devices with a well-specified connector will not have any reverse bias protection in them because both it shouldn't be needed and for the technical reasons of power loss and space used.

Reverse protection is usually done with diodes, the canonical "one-way valve" of electronics. Diodes have a voltage drop across them, usually 0.6V for standard and 0.4V for Schottky types. Using one of these to protect one rail means you'll have about 1W lost to heat when charging at 3A. They are also not small for the currents involved with high-speed charging, being about 7x6x2.5mm for the smallest ones I can find that can handle 3A.

Devices don't usually have too much in the lines of over current protection outside of something like a polyfuse because the device will dictate the current used; if everything is okay in the device it'll set the charging rate and it only needs the most basic of protections in the case that something goes grossly wrong with the device. Sources are what really need overcurrent protection as they don't "have a say" in how much current is drawn."

I mean, yeah, it makes sense. You can reasonably expect chargers/cables not to be quite this jacked up, and in turn, you save space in a space-constrained device. It's just a hard pill to swallow thinking that a stray cable could ruin my device like that.

(Power dissipation aside, I did manage to find some smaller Schottky power diodes that could easily handle the current. There's also the possibility of using a MOSFET, which would turn into Rds(on) * Iavg, so something like 0.15W for Rds(on)=50mOhm, which is much more manageable. I'm just a shade tree electronics guy, so I realize there was probably enough data to make a decision to forego said protection.)

A power cable would be a passive component.

I've had several of Apple's little power charging blocks melt and burn on me.

I had a CRT monitor explode when I was a kid.

Just because stuff is getting smaller doesn't mean it's any more or less liable to break. You use electronics, you're taking a risk, and if you use off-brand, non-official components with other brands, well, hell, even if you're using the official brand apparently, you're playing with, well, not fire, but electricity, which is almost a cousin of fire? Heh.

Pretty sure you could make a "plain old regular" usb cable that would do similar. Maybe you should go back to RS-232? You should also beware of just about any usb device you see, as it could be a http://kukuruku.co/hub/diy/usb-killer in disguise
Unfortunately for you, the exact same problem can happen with plain old regular USB; in fact, this cable had only the wires from plain old regular USB. Reversing the wires on a USB 2.0 cable would have the same effect.