Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
×
Bug Hardware

Some Reversible USB-C Cables/Adapters Could Cause Irreversible Damage 136

TheRealHocusLocus writes: Three Decembers ago I lauded the impending death of the trapezoid. Celebration of the rectangle might be premature however, because in the rush-to-market an appalling number of chargers, cables and legacy adapters have been discovered to be non-compliant. There have been performance issues with bad USB implementation all along, but now — with improved conductors USB-C offers to negotiate up to 3A in addition the 900ma base, so use of a non-compliant adapter may result in damage. Google engineer and hero Benson Leung has been waging a one-man compliance campaign of Amazon reviews to warn of dodgy devices and praise the good. Reddit user bmcclure937 offers a spreadsheet summary of the reviews. It's a jungle out there, don't get fried.
This discussion has been archived. No new comments can be posted.

Some Reversible USB-C Cables/Adapters Could Cause Irreversible Damage

Comments Filter:
  • Stupid design (Score:5, Insightful)

    by Anonymous Coward on Friday February 05, 2016 @05:56PM (#51450179)

    If you're relying on the cable having enough intelligence to prevent the two devices from hurting each other, you've already messed things up.

    • by bondsbw ( 888959 ) on Friday February 05, 2016 @06:30PM (#51450451)

      Some Reversible USB-C Cables/Adapters Could Cause Irreversible Damage

      The irony that USB is finally reversible, yet the damage it causes is not...

      • Re:Stupid design (Score:5, Informative)

        by x0ra ( 1249540 ) on Friday February 05, 2016 @06:42PM (#51450489)
        TFA's title is mislieading. The cable in question was wrongly wired. If you connect GND to V+, and V+ to GND, bad things *will* happen even with USB 1.0.
        • So reversed wires cause irreversible damage?
          • You know how many times I've seen EEs blow up hardware hooking things up backwards?

            In the software world I see EEs cast datatypes every which way breaking the type safety. EEs that I know just don't understand that people will try to hook things up backwards and they should be protected from it. Compiler protect us when we aren't abusing the type system.

            We mix up type C connectors all the time now, so clearly the type safety is broken. We casted it into a host connector.

            • by mikael ( 484 )

              That's funny. Give them a C++ library with classes with constructors, algebraic operators, and they'll still want to create their own foundation classes that have the variables but no operators, requiring everyone else to have to work around that logic.

            • You know how many times I've seen EEs blow up hardware hooking things up backwards?

              This is why we invented bridge rectifiers

          • by Richy_T ( 111409 )

            Have done this myself wiring up a USB port with the pins the wrong way around and smoking a USB drive. Fortunately, nothing of value was lost.

          • Ever wonder why no USB ports work in computers of the electrical engineering buildings in universities?

            Very few computer USB ports are resistant to abuse caused by short circuit, reverse power, or worse.

            • Ever wonder why no USB ports work in computers of the electrical engineering buildings in universities?

              Very few computer USB ports are resistant to abuse caused by short circuit, reverse power, or worse.

              the ones around my house (mainly phone chargers) have a finite half-life, as far as i can tell just from the mechanics of plugging in and out. as do the cables. on both the big USB end and the little microUSB end

        • by hawguy ( 1600213 )

          TFA's title is mislieading. The cable in question was wrongly wired. If you connect GND to V+, and V+ to GND, bad things *will* happen even with USB 1.0.

          I thought USB 1.0 spec'ed over-current production for all power pins? (which is usually done using self-resetting polyfuses)

          • by x0ra ( 1249540 )
            we're not talking about over current, we're talking about simple wrong polarity.
            • by hawguy ( 1600213 )

              we're not talking about over current, we're talking about simple wrong polarity.

              If reversed polarity doesn't lead to over current, then what happens that makes it so bad?

              • Re:Stupid design (Score:4, Interesting)

                by Khyber ( 864651 ) <techkitsune@gmail.com> on Saturday February 06, 2016 @03:09AM (#51451989) Homepage Journal

                Current doesn't kill silicon, voltage does. Example, you take an LED. It's a red one that runs at 2V. You can probably dump 3-4x that voltage through it without a resistor, and it won't care as long as the polarity is correct and it has adequate heat sinking. Now, this same LED has a reverse breakdown voltage. Many LEDs now days have native protection about double their nominal operative voltage. So for this LED, it can take upwards of ~4V reverse polarity. You give it 5V or higher in reverse, you will destroy the p-n junction.

                This knowledge is what is used to design LED arrays which can run natively off wall power without any power driver circuitry.

                • by hawguy ( 1600213 )

                  Current doesn't kill silicon, voltage does. Example, you take an LED. It's a red one that runs at 2V. You can probably dump 3-4x that voltage through it without a resistor, and it won't care as long as the polarity is correct and it has adequate heat sinking. Now, this same LED has a reverse breakdown voltage. Many LEDs now days have native protection about double their nominal operative voltage. So for this LED, it can take upwards of ~4V reverse polarity. You give it 5V or higher in reverse, you will destroy the p-n junction.

                  This knowledge is what is used to design LED arrays which can run natively off wall power without any power driver circuitry.

                  V+ and GND are power supply rails, are you claiming that an external device can overdrive the computer (or USB chipset's) power supply without sending enough excess current through it that would trip the fuse?

                  • by Khyber ( 864651 )

                    Yes, fuses will blow either by over-voltage in such a situation, as you need to overcome the voltage from that power supply in order to start doing damage.

              • If reversed polarity doesn't lead to over current, then what happens that makes it so bad?

                Really? After all the times that Will Wheaton or Jordi decided to reverse the polarity, you have to ask?

            • we're not talking about over current, we're talking about simple wrong polarity.

              you see, gay marriage will ruin America after all.

  • by sims 2 ( 994794 ) on Friday February 05, 2016 @05:57PM (#51450189)

    Timothy you've mentioned this before..
    http://hardware.slashdot.org/s... [slashdot.org]

    • by p0p0 ( 1841106 )
      Gotta get that sweet sweet ad revenue.
    • To be fair to Timothy, up until now Benson had never received a cable that was so incorrectly built that it killed his testing equipment and a laptop. To find a cable that bad, that's news.
  • by Anonymous Coward

    http://hardware.slashdot.org/story/15/11/05/1959216/google-engineer-warns-against-perils-of-buying-cheap-third-party-usb-c-cables

    • by phantomfive ( 622387 ) on Friday February 05, 2016 @06:13PM (#51450327) Journal
      This is a new development. Benson Leung found a cable so bad it destroyed his analyzing equipment, and he says he won't be able to do reviews anymore because of it. The cable was actually missing wires internally, among other things.
      • by ArtForz ( 1239798 ) on Friday February 05, 2016 @06:21PM (#51450403)

        The missing wires weren't the really bad part - the result of that would've "merely" been that it wouldn't work as a USB3 SuperSpeed cable and only connect in USB2 High Speed mode.
        What really set that one apart was that it had VCC and GND swapped on one end.

        • You have to wonder how a cable like that gets released.....didn't they plug it in even once?
          Stuff like that makes me start to believe stories about people who got electrocuted while talking on a charging phone.......
          • The surprising thing in all of this is a lack of regulation.

            Surely there ought to be a body like the FCC that issues a certification for electrical components sold by a major American online retailer like Amazon before these things even get listed.

            • That sounds like an invitation for corruption. A few large companies manage to take control of the FCC (or whatever), who then sets regulations that only large companies can follow. Or a similar problem.
              It is enough to check if the cable has been certified by Underwriters Laboratories or something similar.
            • by Mike610544 ( 578872 ) on Friday February 05, 2016 @09:24PM (#51451125)
              There is an organization [usb.org] that certifies USB stuff. If a cable has a little sticker with that certification logo, it's probably good.

              Not sure how much they do to prevent unauthorized use though.
          • I wouldn't be surprised if the people that manufacture them have no idea what they are for.

          • by mikael ( 484 )

            The Atari 800 used to have 9 pin sockets for the game controllers. These matched RS232 connectors. We replaced the cheap plastic moulded connectors with metal RS232 connectors. Only downside was that touching the metal case against the lower pins of the socket would reset the machine.

            That mistake was duplicated with the desktop PC where the EGA connector was also 9 pin. You could end up plugging a monitor into the RS232 port.

            Then there's SCART, where the cable was attached to the socket diagonally, constant

        • I'm sorry, but his "testing equipment" isn't all that great if it can't handle that. I mean, that's pretty much the simplest problem there can be right? Entire pins of the cable swapped. Ethernet testers are designed to catch this exact thing. Slap a diode on the frigging test equipment or something.

          • I'm sorry, but his "testing equipment" isn't all that great if it can't handle that.

            As someone else mentioned, it wasn't all that great, just a chromebook and a traffic sniffer.
            Still, USB 3 is capable of 5amps of current, so it can fry stuff fairly easily.......

      • by x0ra ( 1249540 )
        Come... "lab equipment" here is a dongle and a chromebook, not a $10k setup...
  • Test your equipment (Score:5, Informative)

    by jgotts ( 2785 ) <jgotts@gmaCOLAil.com minus caffeine> on Friday February 05, 2016 @06:17PM (#51450375)

    When it comes to USB, test your equipment, even if you haven't upgraded to Type C yet.

    I've personally discovered two counterfeit or substandard (depending upon your personal definitions of the terms) USB charging cables.

    What I use to test is a Samsung Galaxy Tab 3 10-inch tablet. This tablet wants approximately 0.7-0.8 amps at 5 volts, but it will charge in a degraded mode if the charging cable isn't up to snuff, or if it's plugged into a desktop or laptop (which normally only supply 0.5 amps).

    Every cable should begin by charging in the degraded mode when plugged into my laptop and then upgrade to normal charging mode when plugged into any of my half dozen or so 2 amp USB chargers. Among over a dozen cables, I detected two that were not up to snuff, and you'd be surprised at my results. One cable from the dollar store was garbage, but another, colored cable from the dollar store that had fancy LEDs was fine. Three 10 feet cables were fine. The other reject was an average-looking cable with an average feel. It did not appear to be substandard or counterfeit.

    If you want to get fancy you can get a device from banggood.com that measures current and voltage across the USB port. They cost about $3 shipped. That is how I determined that my tablet will draw approximately 0.7-0.8 amps. From that experience I'd be surprised if many devices actually draw a full 2 amps. It's nice to have a 2 amp supply, though, because it gives you a safety factor if your cables are somewhat substandard. Maybe the newest 2016 phones will draw close to 2 amps. Get the meter and find out!

    Based upon my experience, the best USB chargers are from Samsung and anything else that has a counterfeit-resistant UL sticker. And also based upon my experience, if you notice that a charging cable is getting warm, you should probably replace it because it's dissipating electricity as heat rather than conducting it.

    • How about it should work as described? I doubt most users would be testing equipment to be sure it matches spec. They would simply expect it to work, as it should.

      If enough cases of exploding computers happen, due to a proliferation of dodgy USB-C cables, I would believe it would quickly get the attention of law makers, rather than just the tech community and merchants?

      At the same time, enough exploding computers may simply drive people to high-end brands, if they care enough about their investment.

      • "....If enough cases of exploding computers happen, due to a proliferation of dodgy USB-C cables, I would believe it would quickly get the attention of law makers..."

        Lawmakers give attention to computer users?

  • by AmiMoJo ( 196126 ) on Friday February 05, 2016 @06:20PM (#51450395) Homepage Journal

    This isn't a USB 3 problem, it's just a general problem with stuff you plug in to computers.

    The issue was that the power and ground wires were swapped over. Even on USB 2 that only supplies 500mA it would most likely have killed something. And the same goes for every other port on the machine, including HDMI, Thunderbolt, FireWire, PS2 and eSATA+power.

    Any type of cable wired this way is liable to kill something.

    • Most USB2 hosts include a polyfuse. Not all, but most. The purpose of the polyfuse is to prevent a shorted device from damaging the host. That doesn't work so well with a 3A port, because even normal charging current is quite enough to set a substandard cable smouldering.

      • Most USB2 hosts include a current limited load switch. When an overcurrent is detected, it turns off the switch and signals the host. This allows the host to display a warning to the user. Just using a PTC resettable fuse does not allow for any user feedback. Also, those PTC fuses are not very accurate and take some time to blow. The load switches are less impacted by changes in the ambient temperature and are much faster to react to an overcurrent event.
        • by AmiMoJo ( 196126 )

          The current limiting helps when you have a short or other over-current situation, but not so much when the port sees a negative voltage. That's basically what happened, the +5V line saw -5V instead. Being 10V lower than it was designed for killed it, most likely by blowing the FETs in the current limited supply.

    • by sexconker ( 1179573 ) on Friday February 05, 2016 @06:39PM (#51450473)

      You should never be able to permanently damage a grounded device by connecting something to its ground line. The ground line should be safe for all sane voltages (all bets are obviously off with lightning). On laptops and phones and shit this is a problem because they aren't grounded and have to dump out to the case or something. Even when plugged in, many laptops aren't grounded.

      Modern motherboards advertise physical USB port protection for such bullshit precisely because it's becoming a more common issue with shitty USB devices and cables. It's not a USB 3 problem it's a USB and shitty implementations problem, but that's what happens when you commoditize so hard that all anyone ever does is buy the cheapest item listed that ships with Prime.

      • that's what happens when you commoditize so hard that all anyone ever does is buy the cheapest item listed that ships with Prime.

        Clearly, the solution then is to only buy Monster cables. ;P

      • by dfsmith ( 960400 ) on Friday February 05, 2016 @07:24PM (#51450649) Homepage Journal

        It's a shame that USB went with Vbus/Gnd rather than V+/V-. The latter would allow the negative side to drift away from ground (which is typically kept on the shield in a zero-current path). They had to devote a small section to voltage drops it in the latest spec (3.6.10.1 in USB_PS_R2_0 V1.1: bottom line is 375mV of Gnd rise and 625mV of Vbus drop).

        The Ethernet guys realized this would be a problem so they made their signals transformer isolated. But USB is designed for short runs and cheap interfaces, not 100m runs across different electrical grids.

        Note, however, that this case was not related to the ground potential at all. Vbus and Gnd were reversed, exceeding the Vgnd_drop limit by over 9V (2400%), or possibly 40V (>10000%) if the device had managed to negotiate the new voltage limits before it died.

        • The Ethernet guys realized this would be a problem so they made their signals transformer isolated.

          The ethernet guys designed an electrical system that required differential signalling back when the easiest way of doing so was a pulse transformer. The voltage had nothing to do with it. The pulse transformer is a throwback to a long lost day that does nothing other than add bulk to units at both side.

          Even with USB you're more than welcome to short your D+/D- pins to VCC or GND and nothing at all will happen. Also USB signals were never designed for long distance transmission where ground rise or voltage d

          • by dfsmith ( 960400 )

            You probably won't break anything shorting D+/D- to Vbus or Gnd, but <0.3V signals SE0 reset on USB, and the voltage thresholds are set at 0.3V and 2.8V relative to Gnd. Ethernet permits common mode voltages of +/-20V and 40kHz (802.3 section 12.5.3.2.5).

            Another interesting cable technology is MIDI, which was designed to connect instruments are are intentionally ungrounded. They use current signalling designed to directly drive opto-isolators!

            • You probably won't break anything shorting D+/D- to Vbus or Gnd, but <0.3V signals SE0 reset on USB, and the voltage thresholds are set at 0.3V and 2.8V relative to Gnd. Ethernet permits common mode voltages of +/-20V and 40kHz (802.3 section 12.5.3.2.5).

              Another interesting cable technology is MIDI, which was designed to connect instruments are are intentionally ungrounded. They use current signalling designed to directly drive opto-isolators!

              You definitely won't. There's specific states of the system defined for shorted D+/- lines to GND and Vcc.
              Ethernet needs to contest with common mode voltages too. When your cable run is 100m in a tray next to power cables you can inductively pick up quite a bit of voltage.

              But that isn't in itself a requirement or a result of using pulse transformers.

      • What is ground? Ground is just a reference. You may just as well say they powered the target device with -5v instead of +5v.
        This wasn't the result of connecting something to the ground, it was the result of connecting an incorrect voltage to the Vcc line.

        Also you got your thinking 100% backwards. Only on an UNGROUNDED device can you not damage something by attaching a ground connection, because only then can you ensure that no current flows. Got a wiring problem that put 240V on the earth of your USB device

  • by John Allsup ( 987 ) <slashdot@chal i s q u e.net> on Friday February 05, 2016 @06:23PM (#51450413) Homepage Journal

    We need inexpensive reliable testers for usb cables. Basically a box where you plug the cables in and it does the various electrical tests.

    • by SeaFox ( 739806 ) on Friday February 05, 2016 @06:30PM (#51450445)

      We need inexpensive reliable testers for usb cables. Basically a box where you plug the cables in and it does the various electrical tests.

      Isn't the whole point of standard that the consumer should know that two devices are equitable in abilities?
      The problem here is the owner of the USB 3.0 spec is not releasing their legal hounds on companies manufacturing "USB 3.0" cables that don't truly support the standard.

      • Makes we wonder if he could sure the USB specification people for false advertising and not enforcing their standard?

    • by Anonymous Coward

      Android tablet, inexpensive enough?

    • by Richy_T ( 111409 )

      I've thought of something similar. I tend to buy cheap cables off ebay (and elsewhere) and I've had some bad ones (and a lot of good ones). Generally, it's been a risk I'm willing to take but it would be nice to know ahead of time the bad ones (and some work fine in one situation but not in others)

  • If you plug in a non-compliant usb-c cable into a device's usb port, a compliant device should be able to recognize it as such and simply refuse to operate. It should categorically *NOT* cause the device to cease to operate.

    The fact that this guy apparently shorted a $1000 computer because of a badly made $10 cable IMO shows just as much of a flaw in the computer as it does in the cable.

    All that the computer needed to have on the port was a breaker that would trip if or when the expected limits were

    • by Anonymous Coward
      USB-C are assigned to passively provide charging power at the same position regardless of the plug's orientation. It is NOT designed to handle the case of mis-wired cables that provide external power of reversed polarity. Instead, it is simply designed so that users can not cause problems with correctly-wired cables, under the reasonable premise that users won't be manufacturing their own non-compliant USB cables. The same result could happen if any other type of USB cable had its power pins reversed and pr
      • by mark-t ( 151149 )

        Imagine what would happen if there were no protections from people attaching equipment to their phone lines.... one person could sabatoge every landline telephone on his entire block.

        If it's connecting to something made by a third party, it shouldn't matter if it is using a "standard" jack or not.... protection mechanisms should exist to ensure that noncompliant devices don't damage it.

        Doing otherwise is the hardware equivalent of allowing a stack overflow bug based on unexpected user input.

    • It was a Google employee whose company supplied computer was designed by the company themselves. He probably copped some flack from his manager for destroying a thousand dollar piece of hardware. But equally the QA team may cop bigger flack for not testing this scenario.

      Wait for the 2016 model correcting this problem. :)

    • One crappy cord, and his $1500 computer would be fried.

      From the article. It almost sounds like the guy was determined to continue his quest for a crappy cable until he destroyed an expensive laptop.

      Any sane person - or one using equipment they have paid for, themselves - would have tested on something less expensive if not actually sacrificial. But no! This guy decides that a high-end computer should be his victim.

  • The Cost of Quality (Score:2, Informative)

    by Anonymous Coward

    A quick analysis of the workbook by TheRealHocusLocus finds that "Approved" cables cost ~60% more than unapproved cables on the average.

    Apprvd Average of Cost
    NO $10.61
    Yes $16.71

    -- Jared

  • The damage is the fault of the port design, not the cable. The port should be able to handle short circuits, switched wires, and incorrect power control signals without permanent damage. What if the perfectly approved Apple cable has been chafed and is now shorted to ground? Fail gracefully.
    • What if the perfectly approved Apple cable has been chafed and is now shorted to ground? Fail gracefully.

      That's what you might think, but I'm gonna go out on a limb and say a cable that would fail gracefully when shorted to ground would not pass mains voltage [dailymail.co.uk] mains [dailymail.co.uk] to ground (e.g. the phone case). Read the articles before you blame shoddy aftermarket chargers, as the 2nd one involved a genuine Apple charger; but what you're saying is that the cable, connector, and/or device should somehow prevent this.

      • by Anonymous Coward

        The second one apparently involved a licensed charger. The question is though, why has this not happened outside of Russia and China? Given that the number of iPhone/iPad users outside of China and Russia would outnumber the Russian/Chinese users by quite a lot, why are these incidents isolated to those particular countries? (I did find one in Thailand with a guy dying who was using a thirdparty unlicensed charger).
        Is it possible that all of these deaths can be attributed to third party chargers which fa

      • The device should prevent this. USB has to negotiate higher currents so high currents shouldn't be present in the first place. Second, short circuits and overcurrents should be handled by the host as well as reverse wiring by the device. In the device it's easy: 1 diode short circuits a reverse polarity situation causing an overcurrent >500mA on the host which switches it's polyfuse.

        • Wow, where'd you get your EE degree? That's impressive!

          Your complete lack of understanding of how electricity works, that is. And USB, for that matter.

          For starters, the way USB handles overcurrent conditions is by limiting current to slightly more than requested (or slightly more than 500mA, 900mA for USB 3.1, if no negotiation has occurred) and monitoring current draw. In that way, even a dead short can not damage the device and the moment an overcurrent condition is detected, the port is shut down. Th
          • by guruevi ( 827432 )

            Sorry, I got my EE degree in a non-English language and haven't used it much in my US career.

            How does a wire cause a reverse voltage (-5V on Pin 1) when all it gets is +5V?
            Pin 1 VCC (+5 V, red wire)
            Pin 4 Ground (black wire)

            We're talking about the wires (pin 1 and 4) carrying the voltages being reversed right? The data wires should be on optocouplers or some form of transformer.

            The simplest protection against miswiring a voltage source or preventing reverse current from flowing (eg. in PV panel

            • I lack an EE degree as I'm a software developer and not a hardware guy (some day, perhaps...) but I do generally get the basics. It still comes down to planning for the least likely scenario when you can't do anything about the more likely ones, but that's product design, not port design; even if that level of protection was in the spec (and I haven't looked at 3.1 in detail) there's nothing preventing manufacturers from leaving it out and just not using the logo. Beyond that, while it would have been prefe
          • I believe the suggestion was a parallel diode that would be reverse-biased in normal operation. If a cable was connected with the polarity swapped, it would clamp the negative voltage and trigger the overcurrent protection on the driving end.
            • Now that explanation makes sense. We're still relying on external equipment to function per the spec, though, and actually have overcurrent protection; we're also hoping that said protection activates before our equipment is destroyed. And it's still the much less likely scenario, compared to overvoltage, which is more difficult to protect against; if that's not being handled, it still doesn't make sense to handle reverse voltage.
  • by Anonymous Coward on Friday February 05, 2016 @07:37PM (#51450703)
    One connects to all of your devices and accesses the data, the other is a hardware standard.
  • Just looking over Benson Leung's reviews list [amazon.com], some of these items are transparently fishy. For instance, this Benson review [amazon.com] currently has nearly 50% "unhelpful", while the product page is fully of sketchy positive reviews. It's a USB cable, who takes photos and writes essays about them (people with multimeters and cable testers aside)? Tried it in FakeSpot, and it figures the product has "68% low quality reviews" [fakespot.com].

    Fake positive reviews and "unhelpful" spamming on negative reviews. Does Amazon have any mecha

  • Just looking over Benson Leung's reviews list [amazon.com], some of these items are transparently fishy. For instance, this Benson review [amazon.com] has nearly 50% "unhelpful", while the product page is fully of sketchy positive reviews. It's a USB cable, who takes photos and writes essays about them? Tried it in FakeSpot, and it figures the product has "68% low quality reviews" [fakespot.com].

    Fake positive reviews and "unhelpful" spamming on negative reviews. Does Amazon have any mechanism to deal with this type of gaming?

  • Comment removed based on user account deletion

Every nonzero finite dimensional inner product space has an orthonormal basis. It makes sense, when you don't think about it.

Working...