back to article Google engineer names and shames dodgy USB Type-C cable makers

A Google engineer says some discount USB Type-C converter cables are substandard and could cause damage by drawing too much juice. One of the big advantages of the USB Type-C design is never again having to guess which way up the plug has to be to fit in its hole; the other advantage is power transmission. A Type-C 1.1 laptop …

  1. Anonymous Coward
    Anonymous Coward

    Nice one. Now do it for Apple Lightning cords/cables/leads/adapters. I don't know how many cords longer than 6 feet I've bought that don't charge the device (they lack the little authentication chip, apparently).

    1. Anonymous Coward
      Anonymous Coward

      Now do it for Apple Lightning cords/cables/leads/adapters. I don't know how many cords longer than 6 feet I've bought that don't charge the device (they lack the little authentication chip, apparently).

      This guy just validated why that may be, after all, not such a bad idea. If you pull 2.1A through a wire that cannot support it you end up with a fire hazard, so at 3A even more so.

      I've read some reviews of cheapo power supplies a while back, and the shortcuts they take in those devices are plain scary. Of course, they have fake approval logos on them, but you're looking at something that could kill. I stands to reason that those people would not worry about the right cables either..

      1. Triggerfish

        There is somewhere on the web a rather interesting comparison from and electrical engineer between a apple power adaptor and a generic chinese ebay one, he pulls it apart and goes over the design choices made, component choices etc and what makes it different, there is a lot of difference in quality.

        Edit found it.

        http://www.righto.com/2014/05/a-look-inside-ipad-chargers-pricey.html

        1. maffski

          That's a counterfeit charger, not a 'generic' one.

          1. Triggerfish

            Pardon me, it was from memory of looking at it yonks ago and forgot to re-edit that sentence, however what one do you think a generic one is more likely to resemble?

        2. John Brown (no body) Silver badge
          Thumb Up

          From the quoted link

          "Since the chip only has four pins, I expected it to be a trivial Ringing Choke Converter (RCC) circuit with just a couple transistors inside the chip - but I cracked it open with Vise-Grips and it turns out to be a fairly complex chip. I took a picture through a microscope of the IC die, which is about 1 mm across."

          Now THAT'S a tear down!

      2. ChrisC Silver badge

        Even worse than the ones with fake approvals are the ones where the manufacturer has presented a fully populated "gold-standard" unit for testing in order to gain a genuine approval mark/number for the product, and then proceeds to strip out all the protection components for the units actually manufactured/shipped. Doing a search on the approvals body website will then appear to show the unit meets the required standards, giving even more of a false sense of security than if they'd just skipped the testing entirely and slapped a completely made up approval number (or copied one off a different product - 'ere Mike, the UL site seems to think this charger is actually a microwave oven, what's up with that?) on the unit.

    2. Daimon

      Check for the MFi compliance logo with Apple accessories - that covers USB spec compliance for electrical behaviour on top of the MFi requirements. The logo should also be on the packaging when you receive the cable/speaker etc.

      1. Tom 13

        Re: Check for the MFi compliance logo

        What? You think somebody making a dodgy cable won't forge a compliance logo?

        There's an actual value to knowing your sales guy isn't dodgy and isn't buy kit from dodgy oem suppliers. The problem is first finding one, and after that making sure your own PHB doesn't ditch him.

    3. Anonymous Coward
      Anonymous Coward

      Apple Lightning cords/cables

      Like Apple are ever going to allow anyone to publish a report on dodgy Lightning connectors. Even the Apple store rating say they are total pants

      And fanbois, before you childishly down vote me for raining on your parade go and take a look at

      http://www.apple.com/shop/product/MD818AM/A/lightning-to-usb-cable?fnode=97

      Of course the reviews they show on the page are the hand full of people who think they're great rather than the massive majority that complain how total shit they are.

      Still the cheap Chinese knock off you can buy easily now seem to be much better, last much longer and don't cost $19 each.

      1. Anonymous Coward
        Anonymous Coward

        Re: Apple Lightning cords/cables

        Leave it to Apple haters to take an article about USB-C and use it to talk smack about Apple, and totally ignore the fact that Apple having the MFi logo for Lightning accessories (which they see as Apple trying to "force" you to pay inflated prices for cables) weeds out these sort of things.

        Apple can and will sue over counterfeit Lightning products that don't meet their standards, because the 5 cent chip that authenticates it to the iPhone has to be counterfeit in a non-approved device. In the USB-C world there's nothing anyone can do beyond what Google is doing here by simply naming and shaming. It's nice that they're trying, but neither they nor the USB Forum can do anything to stop the sale of these dangerously made cables.

        1. Martin-73 Silver badge

          Re: Apple Lightning cords/cables

          What does a defunct furniture chain have to do with compliance?

          Oh and requiring a chip from apple. Yeah... that's nice.

          People hate apple for a reason

        2. Anonymous Coward
          Anonymous Coward

          Re: Apple Lightning cords/cables

          Actually I never used to be an Apple hater till all this shit with Lightning cables. What sort of marketing genius thought that producing cabled at nearly $20 each would endear you to your customers. But I could cope with that if they actually worked. The thing about Apple used to be "it just works" well you can forgive a lot when thing do. The problem I have with Lightning cables is they'd be more reliable if they were made of cheese. In the hands of my son and many of his friends they last less than 3 months on average whereas USB leads never seem to need replacing.

          Apple can and will sue over counterfeit Lightning products that don't meet their standards

          Don't meet their standards! You mean don't die on average every 3 months. The cheap knock off ones you buy on line are now so much better than the official ones and they are now more reasonably priced. They still haven't quite reached the point with traditional micro USB leads where you look to see how many you can buy for less than $5 (I think my best buy was 8 for $5 all of which still work fine), but the $5 ones work better than the official ones.

        3. Tom 13

          Re: weeds out these sort of things.

          Actually it doesn't. Bought one a couple months ago from Best Worst Buy. It didn't work, they wouldn't give me my money back because I didn't get it out of the box soon enough. Yes, it was a genuine product with the label and cost appropriately. Ironically the "dodgy" knock off I got from the $5 place works just fine.

  2. HCV

    Flaunt: to show off, to provoke envy

    Flout: to blatantly ignore or disregard

    1. Pookietoo
      Headmaster

      "I want to hold them to task" - wrong

      "I want to take them to task" - correct

      1. DropBear
        Joke

        ...but but but I want to take and then hold them there!

        1. itzman
          Happy

          No, you want to take them to task and then hold them responsible...

      2. Tom 13

        Yes, but if he spoke Engrish very goodly nobody wood think he wuz a goud engineere.

  3. Tromos
    Joke

    No more 99p shop cables for me

    From now on it's pound shop and hang the expense!

    1. Stuart 22

      Re: No more 99p shop cables for me

      This whole thing is incredibly short sighted of Google not shipping a USB-A cable in the 5X box (unlike the 6P). At this early stage of USB-C deployment most of us will have to buy that cable to connect to a computer or if we need to share chargers and powerpacks.

      The Google one costs £10.99. The Chinese do adaptors for 96p delivered on eBay. Hard choice. I bottled and got a CPC-Farnell cable for £2.68. The cable is re-assuringly thick so I hope they haven't skimped on the resistor. I'm too mean/poor to buy a Pixel so I can't be sure. But really Google adding the cable to the BoM would have been peanuts especially on phones not competing in the budget market. Even if they had raised the price a little both Google and the consumer would have been mostly better off.

      1. DropBear

        Re: No more 99p shop cables for me

        "The cable is re-assuringly thick"

        If it's really cheap and you really give a #### buy two and section one somewhere. Then potentially marvel at the 0.01mm thick copper ($$$$$) wires inside the 5mm thick plastic ($0.001) - at least, if car jumper cables are anything to go by considering it's rare to gaze upon the thickness of the actual wire underneath the lavishly thick rubber without getting the heebeejeebees with most of them...

        1. Martin-73 Silver badge

          Re: No more 99p shop cables for me

          Copper? Luxury.. most of the cheap chinese cables these days are using copper coloured aluminium.

    2. TRT Silver badge

      Re: No more 99p shop cables for me

      The good thing about the 99p store is that you can pick up a fire extinguisher at the same time and even after spending £1.98 you're STILL almost £9 up on the deal.

  4. Kevin McMurtrie Silver badge
    Boffin

    Er....

    Old USB allows up to 5A for chargers. With such a low voltage used, it's difficult for there to be more than 2.5 W of losses before the power consumer no longer has enough voltage to function. 2.5 W lost across the length of a wimpy cable won't generate much heat. 0.5 W lost in a heavy cable with 2 W lost at a spot of fraying sill won't generate much heat due to good thermal conduction.

    I could be reading the spec wrong, but it doesn't look like the resistor needs to be there except to possibly improve efficiency. USB-C high voltage mode needs a data handshake.

    1. Phil Endecott

      Re: Er....

      > Old USB allows up to 5A

      That's not how I remember it. Maybe you meant 5W ?

      1. phil dude
        Boffin

        Re: Er....

        decimal point? I thought 2A max is what you get from normal USB...

        P.

        1. david 12 Silver badge

          Re: Er....

          USB is 100mA, (0.5W) negotiable up to 500mA (2.5W). Apple negotiated that limit up (by ignoring it), to provide 1.8A (9W): this was why you could only do rapid charging using an Authentic Apple Charger. The new standard was called a "USB Charging Port", and it was added to USB 2. That is, it was not an original part of USB 2.0

          The USB 3 specifiction is negotiable up to 900mA (4.5W), and the charging port specification is 1.5A up to a total of 5A for multiple devices.

          I don't know why a USB 2 charging port is 1.8A and a USB 3 charging port is 1.5A, but I'm guessing it's because I've misread/missunderstood something.

          1. Charles 9

            Re: Er....

            I don't know why a USB 2 charging port is 1.8A and a USB 3 charging port is 1.5A, but I'm guessing it's because I've misread/missunderstood something.

            Because USB Charging over the traditional cable was always limited by the spec to 1.5A. It's just that many devices push the envelope towards 2A. Type-C cables BTW are allowed to go up to 3A by spec.

          2. Anonymous Coward
            Anonymous Coward

            Re: Er....

            "I don't know why a USB 2 charging port is 1.8A and a USB 3 charging port is 1.5A, but I'm guessing it's because I've misread/missunderstood something."

            IIRC, the USB 3.1 spec was more complicated than that. (off to the Google machine...) I can't find my copy of the USB spec right now, but according to Wikipedia the spec is 2.0 A at 5V, and up to 5A at 20V.

      2. Kevin McMurtrie Silver badge

        Re: Er....

        There's a specification for dumb chargers where the data lines are shorted together and current is negotiated by monitoring voltage droop. Despite some warnings that exceeding 1 amp may not be a good idea, many chargers provide 4 to 5 amps. USB 1 and 2 data devices are limited to ~1 amp because you don't want the power lines to have less voltage than the data lines.

      3. John Tserkezis

        Re: Er....

        "That's not how I remember it. Maybe you meant 5W ?"

        Less than that. 500mA max, and that's with the devices "asking nicely" (there's some USB negotiation before it'll do that.

        1A is however, routine, if not out of spec, and I've seen them squeeze 2A out of it too. Standards are treated like suggestions, if it doesn't fit your needs - just ignore the documents and do what you want.

      4. Phil O'Sophical Silver badge

        Re: Er....

        "Battery Charging Spec V1.2" from December 2010 raised the limit to 5A.

  5. Randy Hudson

    If I had a nickel

    For every time someone skimped out on a pull-up resistor...

    1. This post has been deleted by its author

    2. DropBear
      Trollface

      Re: If I had a nickel

      Sorry, that nickel (per device) was had by them instead, and it's safe to assume they sold a lot of them...

    3. Crazy Operations Guy

      Re: If I had a nickel

      Had a $1000 logic analyzer that did that (well, missing pull-down resistors)... You get all sorts of results when the other end is left floating (the lead was the perfect length to adsorb harmonic frequencies off the AM radio tower nearby and generated enough noise that the analyzer thought it to be the 1.25V signals we were working with...)

  6. Phil O'Sophical Silver badge

    a Type-C gadget could attempt to draw 3A from a USB 2.0 host or charger

    A USB 2.0 device could attempt to draw 3A as well, it's up to the charger to limit the current drawn to whatever it can safely supply, either by a fuse or something more sophisticated. The real problem here is the device manufacturers who just connect the USB power pins to the raw 5v and assume that the conected device will 'play nice'.

    1. John Tserkezis

      "The real problem here is the device manufacturers who just connect the USB power pins to the raw 5v and assume that the conected device will 'play nice'."

      I've seen this more often than I'd like!

  7. Scoular

    Depend on a cable?

    Surely in a sane world the charger would be self protecting including from a dead short, it is quite possible to do.

    Similarly at the other end the device should limit the current drawn to whatever limit the designer intended and that is also possible.

    Why would anyone in the right mind depend on a cable from an unknown supplier to protect their expensive kit. The pullup resistor as part of a signaling scheme? Why not have that in the device so you know it is there.

    1. Anonymous Coward
      Anonymous Coward

      Re: Depend on a cable?

      Surely in a sane world the charger would be self protecting including from a dead short, it is quite possible to do.

      That is only ever going to happen if the components to do so would be cheaper than the parts involved in NOT doing it. That industry goes for large volume, and every cent saved counts. Your life doesn't.

      1. Anonymous Coward
        Anonymous Coward

        Re: every cent saved counts

        That argument is just as effective for cables, except the cable manufacturers often don't have a brand to protect.

        The op is right, expecting the cable to be anything other than a basic piece of wire with the right connectors is stupid. Make the £300 device do the job. You can absorb the cost at the device end more easily than with the cable.

        Seriously, engineers came up with this design? They're allowed to blindly trust their input? It'll end in tears....

        1. The First Dave

          Re: every cent saved counts

          "Make the £300 device do the job."

          How the hell is the device meant to know what charger it is attached to, and with what cable/adaptor ?

          1. Anonymous Coward
            Anonymous Coward

            Re: every cent saved counts

            "How the hell is the device meant to know what charger it is attached to, and with what cable/adaptor ?"

            That is what the 56k pull-up resistor is there for, to signal that a non type-C device is being connected. But that is what's missing from the cables.

        2. Anonymous Coward
          Anonymous Coward

          Re: every cent saved counts

          The problem is that the power delivery spec needs to be backwards compatible with existing USB cables. In order to allow for fancy stuff like charging at 100W, there needed to be some way of having the cable indicate its capability. How else do you let your iWhatever and your beefy charging station determine that they can charge at the "fast" rate? Keep in mind, you also need to make sure a port can supply OR consume power.

    2. Dave Horn

      Re: Depend on a cable?

      Why would anyone in the right mind depend on a cable from an unknown supplier to protect their expensive kit.

      I suspect that would be because the new Nexus series of phones only come with a USB type C cable. Good luck copying anything to/from your phone unless you're willing to sit and wait while it goes over WiFi.

    3. DropBear

      Re: Depend on a cable?

      "Why would anyone in the right mind depend on a cable from an unknown supplier to protect their expensive kit."

      So what should happen when your exquisitely crafted $500 3A-capable power brick and equally exquisitely crafted $500 3A-guzzling USB device get hooked together with the first USB cable you can get your hands on, possibly originally sold with a self- and / or micro-powered gadget slurping not even 100mA, with its power leads sized accordingly...? Attach device -> "I need moar power!" -> "Roger, I'm fine with that!" -> "Bleurghhargfraggh... braaaaains..."

  8. Christian Berger

    So... the news is USB doesn't work in real life...

    ... I always thought we'd knew this since the Microsoft demo which crashed their Windows 98.

    Seriously, USB has so many flaws. For example the electrical side is not properly symmetrical. There is no galvanic separation which makes ground loops a problem. The software side is highly complex and uses 16 bit Unicode as the text encoding, probably the last place left doing so.

    1. Michael H

      Re: So... the news is USB doesn't work in real life...

      Actually, the only support for Unicode in the Win32 APIs is through UTF-16, so Windows software uses it lots. Considering the Wintel environment the original USB spec was borne from, it's probably the reason USB uses it.

    2. John Tserkezis

      Re: So... the news is USB doesn't work in real life...

      "Seriously, USB has so many flaws"

      Still not as bad as bluetooth. If it works, it works great, otherwise their stance is "screw you".

  9. graeme leggett Silver badge

    summary

    Manufacturers don't follow spec if they think they can get away with it.

    Don't suppose trading standards are much use against Chinese import on amazon/ebay etc?

    1. Anonymous Coward
      Anonymous Coward

      Re: summary

      Don't suppose trading standards are much use against Chinese import on amazon/ebay etc?

      The majority of the population has no appreciation for the danger these parts represent and, to be fair, it's not all of them, I've had some decent kit from some brands. The problem is that few think about the reason why some parts are a "bargain", but we see that with services too. After all, there are billions who think they are smart by using a "free" service..

      Typically, a good sign is if they have a local distributor, because an importer ends up with responsibilities and liabilities - go direct and you're taking more of a gamble.

    2. Anonymous Coward
      Anonymous Coward

      Re: summary

      trading standards play the usual game: understaffed, overworked, etc.

      1. TRT Silver badge

        Re: summary

        Yeah, they completely fail to live up to their advertised claims. Someone should report them to the... oh.

  10. king of foo

    KZZZERK!

    Why PFY, with this new android based cattle prod you are really spoiling us....

  11. Cuddles

    Alarming?

    More like entirely expected. For some things, buying cheap (mainly Chinese) knock-offs on Amazon and eBay is a reasonable way to save money. For chargers and batteries, it's madness. You'll only save a couple of quid at most on a USB charger or cable, but only a complete idiot would be surprised that the cheap one is a massive fire hazard.

  12. A K Stiles
    Childcatcher

    In an ideal world...

    Surely the solution would be for each part of the system to limit the current - so the supplying device doesn't provide more than it should, the end device won't receive more than it should and the cable won't transmit more than it is capable of handling. Why does there need to be any active communication between the devices?

    1. Anonymous Coward
      Anonymous Coward

      Re: In an ideal world...

      Becuase a dumb cable has no physical means of limiting just how much current it can safely draw at the spec'd voltage, and given the desire to make these things as cheaply as possible, you have to assume the cable (the simplest part of the path) will be the cheapest and dumbest: just wires, connectors, and plastic. Since the first sign of a dumb cable overdrawing could immediately jump to dangerous, it's gonna have to be up to the rest of the chain to prevent this.

  13. Unicornpiss
    Alert

    Here's something similar

    Here's a link to an analysis of a counterfeit Dell power adapter.

    http://www.electricstuff.co.uk/acadapter.html

    The rest of the guy's site is pretty cool too if you're into tinkering with electronics...

    1. graeme leggett Silver badge

      Re: Here's something similar

      This one is a quite comprehensive look at dodgy fake of ithingy charger

      http://www.righto.com/2014/05/a-look-inside-ipad-chargers-pricey.html?m=1

  14. I am a machine (says Turing test)

    Do nuclear power plants use USB Type-C cables?

    After reading this, I just hope USB Type-C cables are not part of the standard Chinese nuclear power plant design, particularly the "export" type.

    1. TRT Silver badge

      Re: Do nuclear power plants use USB Type-C cables?

      It gets worse when you consider some of the power components may also be French.

      1. sabroni Silver badge
        Mushroom

        Re: some of the power components may also be French.

        BACK OFF BRUSSELS!!!!

    2. Crazy Operations Guy

      Re: Do nuclear power plants use USB Type-C cables?

      Don't worry, Nuclear power plants still use old IBM scientific-spec s/360 boxen. And that's on the upgraded plants, the older ones just use relays and vacuum tubes. So they're immune to problems with USB cables. If it works, why bother replacing it when doing so would mean decades of regulatory approval.

      1. Unicornpiss
        Meh

        Re: Do nuclear power plants use USB Type-C cables?

        I'd expect that nuke plants and most of the scientific world still use IEEE-488 cables, though likely not to export data or charge their phones :)

  15. Danny 2

    Excellent use of sark

    "Understandably, Leung used his new Google Chrome Pixel laptop and Nexus smartphone for the tests."

    I've an old VMEbus cable up the loft, cost £100 thirty years ago and that was cost. Twisted pairs, shielded ribbon, resistors too, as used in CERN. Sentimental value - £5 to a museum.

  16. Zmodem

    most devices have capacitors in, so they are globally usable, more so if they are mains powered, and run on all countries power AC/DC 110-240v, then they have to comply with EU`s ErP which is more strict then USA energy star rating power usage thing

    so production lines take the best out of both ErP and engery star, and make a single globally usable production line to keep the cost down

    1. Bloakey1

      <snip>

      "so production lines take the best out of both ErP and engery star, and make a single globally usable production line to keep the cost down"

      Pssst. Why was Sol annoyed?

      Because he could not take Mega Ohm.

      A joke told to me 40 years ago when I was building things and soldering stuff.

      1. Zmodem

        3A on a usb 2.0 device would just be like a constant spike, and not a massive concern anyway, only apple would come with only usb 3.0 type c and no other usb ports

  17. Gene Cash Silver badge
    Unhappy

    Not "Linux commands"

    More like Chromebook Pixel commands, as "ectool" is a Chromebook/Coreboot specific tool.

    1. Crazy Operations Guy

      Re: Not "Linux commands"

      Why bother with system command in the first place? A $10 multi-meter from Radio Shack would be much easier...

      1. Charles 9

        Re: Not "Linux commands"

        Uh...what Radio Shack?

  18. elwe

    The USB 1.1 and 2.0 specifications mandated that host ports monitor the power drawn against that negotiated via protocol and cut power to devices which draw more than they negotiated.

    Unfortunately virtually nobody implemented this, so devices were made which drew more than 500mA, didn't negotiate more than the initial 100mA etc. and we ended up in the situation we are in today.

    Interestingly if you buy a device claiming USB compatibility and using the logos and you connect something which draws too much power and damages it, under the sale of goods act you would be well within your rights to demand the seller fix the device at any point in its life as there was a fault present at the time of sale stopping it from meeting the specification.

  19. Hoddy

    For Chrissake - it's NOT 'Flaunt'

    "I have started reviewing USB cables on Amazon because I have gotten fed up with the early cables from third-party vendors that so blatantly flaunt the specification, and I want to hold them to task," he wrote on Google+.

    I know it's the Google Twonk speaking (clearly, their much vaunted recruitment test doesn't check on literacy) but El Reg could at least have put ‘(sic)’ after the misused word.

    Flaunt != Flout

    While I'm ranting - 'Enormity' means very, very bad, not very big.

    So he's committed the enormity of using flaunt when he means flout.

    I feel better now. Thank you for reading.

  20. techmind

    Watch out for iron wires too

    I discovered some months back that a surprising number of mouse and HDMI cables stick to a magnet (and cut them open and found it applies to the inner conductors by themselves too). This is truly strange, as you'd expect them to be copper (which isn't magnetic).

    Are the Chinese using iron wire or nickel plating or something weird, to save money on copper? That'll increase the resistance too.

    1. PNGuinn
      Mushroom

      Re: Watch out for iron wires too

      YUp. standard practice for many cables.

      Cheap internal telephone cable is often steel / copper plated rather than copper, catx and alarm cables are often aluminium / copper plated. You can get them from cheapjack suppliers with no indication (apart from price) of what you are getting, so caviat emptor.

      Places like Cpc Farnell sell em also, but do indicate what they are in the description. (Look for the abbreviations CPS and CPA as well as a full description.

      I've never tried steel phone cable - have my doubts as to what the effects would be on the idc blades of the connectors.

      Ally alarm cable works ok but beware - its significantly higher resistance, which matters for some circuits, and very brittle in comparison with copper.

      Note that ally catx IS NOT catx - as I understand it the spec requires copper.

      Note that it is NOT safe to weigh the reels to check composition. Some jammy ba****ds weight the reels to fool you. You have to strip and check the actual cable itself.

      I insist these days that the far end has to protrude through the end of the reel so that I can do a continuity AND resistance check. Overdrawing the copper, giving thinner conductors / higher resistance is common.

      Applies to power cables as well. I recently rejected a reel of "BS" 10 mm2 earth bonding cable that was about 8 mm2. Approximately correct outer sheath, thin strands. Use a micometer if you get faced with a part used reel.

      A BASEC approved Turkish manufacturer got their approval bounced by BASEC a while ago for doing just that.

      Not sure of the appropriate icon - decided to go fot the biggie.

      1. Martin-73 Silver badge

        Re: Watch out for iron wires too

        The copper clad steel phone wire isn't compliant to CW1308 which requires tinned copper (note, not bare copper, for historical reasons... soldered joints..)

        I'd also be worried about the self inductance causing problems with DSL

  21. Kelenling

    ESEEKGO USB Type C Cable You Can Try

    I am using eseekgo usb type c cable. It work well and charge so fast. High quality design can use a long time. Hope this helps!

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like