Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Bad USB-C cable destroys laptop

plus.google.com

Post reply on HN
Print view View original

81–90 of 146 posts

Re: Bad USB-C cable destroys laptop

#81

Post by kstenerud »

bsimpson wrote: post #33
AdmiralAsshat wrote: post #12

HN has previously had a story on the author, a Google engineer who tests USB-C cables on Amazon to make sure they're spec-compliant. He does some amazing work: https://news.ycombinator.com/item?id=10508494 I don't have any devices right now that use USB-C, but whenever I see a deal pop up for one, I usually check the Amazon listing to see if this guy has reviewed/certified the cable yet. I recommend others do the sam…

I bought one of these recently, based on Benson's endorsement, for my Nexus 5X: http://smile.amazon.com/dp/B018F8BGLA It's worth noting that the Nexus 5x doesn't do USB 3.1, so even if you get a fancy Super Speed cable like this one, you are still suck with slow charging from your computer's USB A port.

The Nexus 5X supports the new 15W fast charging USB-C standard . It takes about 15-20m for my phone to go from empty to full from the wall charger, and considerably longer from a USB A port.

Re: Bad USB-C cable destroys laptop

#82

Post by knughit »

gpvos wrote: post #74

Seems to be a kind of bait-and-switch operation, see this G+ comment by Lance Nanek: >>It's pretty common on amazon for some good items to be shipped to get some good ratings, then for the seller to replace them with cheap, almost non-functional, shoddy replacements. It's happened to me before with bicycle equipment (pushing the button on a rear light pushed right through the casing, etc.). So it's entirely possible…

These are the wonders of globalization. USA consumer protection laws don't apply to foreign manufacturers using noname brands.

Re: Bad USB-C cable destroys laptop

#83

Post by Perixoog »

cornellwright wrote: post #4

I wonder if this one cable was miswired or it's something consistently wrong with the product. Either way that something like that could make it out of the factory demonstrates terrible or no QA/QC. I've never used the Twinkie but it looks pretty cool. This shows that it may be worth adding some simple protection to the circuit to prevent simple miswires from damaging it or the host PC.

>I wonder if this one cable was miswired or it's something consistently wrong with the product.

He bought more to find out: https://plus.google.com/+BensonLeung/posts/EBGMagC46fN

Re: Bad USB-C cable destroys laptop

#85

Post by klodolph »

mindslight wrote: post #73
klodolph wrote: post #72

Earlier quoted context omitted.

Crowbar isn't a diode, and it doesn't protect against reversed polarity, and it's a bad idea on the input because you're relying on the other device to limit current properly, when it's the other device which got the voltage wrong in the first place.

It acts like a literal crowbar, so why not call it a crowbar? There are likely already plenty of diodes across the rails (the body diodes of the chips), so without a fuse the power is already limited by just the source. IMHO you might as well try and protect the circuit from foreseeable power-limited situations like frayed cables and the OP.

I'm saying a crowbar isn't a diode, and you're saying that you should call a crowbar a crowbar? Color me confused. Usually a crowbar is a circuit with a triac or SCR in it designed to protect against overvoltage.

This isn't about frayed cables. Frayed cables don't do this. This is about a cable which is wired the opposite of the way it should be. A minimal protection circuit is probably going to cost around $0.20 in quantity (Si2323 on Mouser). For a $500 Chrome Pixel C, the protection circuit is worth the effort if more than 1 in 2,500 ($500 / $0.20) laptops would fail due to charger cables with reversed polarity.

If you think that fewer than 1 in 2,500 laptops will fail this way, then don't bother.

Re: Bad USB-C cable destroys laptop

#86

Post by klodolph »

guan wrote: post #11

Reverse polarity protection is often not done with diodes for the reasons explained. It is more common to use a p-channel MOSFET (or ideal diode controllers when there are multiple power sources) when more than a trivial amount of power is used by the design. They are not cheap when properly sized, but are not ruinously expensive either and are not too large for most designs.

Doing the math, a reasonable PMOS might be the Si2323, although it can only handle ~5A. Costs $0.20 in quantity. Therefore, on a $500 laptop, don't bother if it doesn't save you more than 1 in 2,500 laptops. That is, assuming that you are only trying to save the company money, not make a reliable product. PR can always blame the faulty cable / charger.

Re: Bad USB-C cable destroys laptop

#87

Post by klodolph »

nerfhammer wrote: post #48
lucaspiller wrote: post #27

Earlier quoted context omitted.

I was playing with a FTDI (a clone :D) and an Arduino and accidentally mixed up VCC and GND from an external power source. I only realised because Minicom was failing to connect to the device. Then I realised what I had done, I tried removing it from the USB port and putting it back in, but it wouldn't work. I then tried a mouse too, but that wasn't working either. Thinking I'd killed a USB port on my Mac (well there…

the usb ports on macs are supposed to have overcurrent protection. You actually get a dialog popup in the OS when it happens. I know because I've triggered it doing something with an arduino.

USB overcurrent protection is pretty sophisticated. Devices get 100 mA for free, then have to ask the host for more current. Well, it's a bit more complex nowadays, but that's the basics.

Re: Bad USB-C cable destroys laptop

#88

Post by wtallis »

knughit wrote: post #78
teraflop wrote: post #46

Earlier quoted context omitted.

> Each cable needs one of 3 resistors depending on them being a A/B-to-C cable, 1.5A C-to-C cable, or 3A C-to-C cable. No, for C-to-C connections, the resistance is provided by the host ("downstream-facing") port, not the cable. The resistor in an A-to-C cable is basically emulating the signal that would otherwise be provided by the host. There's no such thing as a "1.5A C-to-C cable"; all type-C cables are required…

A world with smart cables is a dumb world. I want smart devices/adaptors and cheap dumb cables.

USB cables aren't smart cables by any stretch of the imagination—not when we've got standards like Thunderbolt. A single resistor is not unreasonable complexity on top of the requirements to get the pinout correct.

Re: Bad USB-C cable destroys laptop

#89

Post by jethro_tell »

knughit wrote: post #78
teraflop wrote: post #46

Earlier quoted context omitted.

> Each cable needs one of 3 resistors depending on them being a A/B-to-C cable, 1.5A C-to-C cable, or 3A C-to-C cable. No, for C-to-C connections, the resistance is provided by the host ("downstream-facing") port, not the cable. The resistor in an A-to-C cable is basically emulating the signal that would otherwise be provided by the host. There's no such thing as a "1.5A C-to-C cable"; all type-C cables are required…

A world with smart cables is a dumb world. I want smart devices/adaptors and cheap dumb cables.

Like dumb parallel port cables? or db29? The good days amirite?

Re: Bad USB-C cable destroys laptop

#90

Post by mindslight »

klodolph wrote: post #85
mindslight wrote: post #73

Earlier quoted context omitted.

It acts like a literal crowbar, so why not call it a crowbar? There are likely already plenty of diodes across the rails (the body diodes of the chips), so without a fuse the power is already limited by just the source. IMHO you might as well try and protect the circuit from foreseeable power-limited situations like frayed cables and the OP.

I'm saying a crowbar isn't a diode, and you're saying that you should call a crowbar a crowbar? Color me confused. Usually a crowbar is a circuit with a triac or SCR in it designed to protect against overvoltage. This isn't about frayed cables. Frayed cables don't do this. This is about a cable which is wired the opposite of the way it should be. A minimal protection circuit is probably going to cost around $0.20 in…

Crowbar is a general concept - a short across a supply. A diode across the rails acts as such, so I referred to it as a crowbar. I'm familiar with SCR-based active crowbar circuits - what we're disagreeing about is narrow-prescriptive versus wide-descriptive usage of a term.

Most frayed cables are going to simply short or open, but the right combination of such would cross wires. But sure, they're going to be rare.

On a $500 device I would indeed expect a modicum of robustness against negligently designed cables and other weirdness. I mean, I'm not the only one that doesn't have a cable tester and buys cables without a connected-equipment warranty, right? Historically, electrical connections have been designed to deal with weird shit. Do you think you could kill an RS-232 port with any permutation of connections?

Of course the cost metric explains why devices are no longer robust. In fact the manufacturer only has to worry about in-warranty failures, and furthermore only the ones they can't blow off as "user error". Just because devices are designed for the 4 month (or whatever it's down to) upgrade cycle, doesn't mean that informed consumers shouldn't call out designers of such equipment as the cheapskates they are!

Post reply on HN