Forgot your password?
typodupeerror

How a Leather Cover Crashes the Kindle 280

Posted by CmdrTaco
from the still-no-antennagate dept.
An anonymous reader writes "Amazon has started offering refunds to Kindle owners who own the unlit leather case who claim that it causes their Kindles to reboot, but are playing dumb on the cause: "our engineering team is looking into this." People have been wondering how a leather cover could possibly crash an electronic device, and why is Amazon offering money back if they don't think there's a problem? It seems that some of the folks over at Connectify have figured it out, and it's a doozy!"
This discussion has been archived. No new comments can be posted.

How a Leather Cover Crashes the Kindle

Comments Filter:
  • Not unprecedented (Score:5, Interesting)

    by betterunixthanunix (980855) on Tuesday December 21, 2010 @03:54PM (#34632890)
    • by spun (1352)

      That really was "magic" compared to this, though. This is just a plain old short circuit.

    • by nomadic (141991)
      WHOA the ftp.sunet.se server is still around? I haven't seen that in years...
    • Re:Not unprecedented (Score:5, Interesting)

      by mswhippingboy (754599) on Tuesday December 21, 2010 @04:22PM (#34633318)
      Reminds me of an incident I encountered back in the late '70s in Pensacola, Fl. We had an IBM 4341 mainframe in our data-center that would just shut down regularly every Friday night, around the same time. We had IBM SEs come in and pour over the logs, week after week, but they could find nothing wrong and no indication of why it was shutting down. They installed monitors to check for power surges - nothing. They replaced parts - still nothing. We were in discussions with IBM to have the entire machine removed and replaced with a new machine - something IBM said they had never had to do before. After months of pulling our hair out, we discovered (not sure who made the connection - but it seemed to be a long shot at the time) the shutdown coincided with the approach of the USS Lexington (aircraft carrier) coming into port (some 10 miles or so away) from it's regular training missions. Apparently the radar from the ship was strong enough to play havoc with the circuitry causing it to trigger a shutdown. The SE installed RF shields within the box and the problem occurred no more.

      So much for magic.
      • by mangu (126918) on Tuesday December 21, 2010 @04:31PM (#34633470)

        I have a similar experience when we were installing some computers in a hydro power station control center. The old control system used electromechanical relays, so it was quite robust, but the digital computers kept crashing. There were some 500 kV lines right going over the control center, so it was assumed they were causing enough interference to crash the computers.

        After months of studies, it was decided that shielding the control center was the only solution. However there was a problem, the large glass window to the observation hall. Someone mentioned that there existed a transparent conductive paint, so they called a paint supplier:

        -"Hello, I'm looking for some invisible paint, to paint glass"

        They hung up without an answer at the other side...

      • We had IBM SEs come in and pour over the logs

        Well there's your problem. They must have spilled some of the liquid they were pouring over the mainframe.

        The RF shielding obviously had the side effect of waterproofing the circuitry case.

      • by HockeyPuck (141947) on Tuesday December 21, 2010 @05:16PM (#34634180)

        I used to work on an old IBM AS400 which provided about 150 terminals (5250) to a bank. At random times, all of the terminals would lose connection to the AS400 which was located in the datacenter which was located in the floor below where everyone sat. The connections would only drop during the daytime, we could hook up all sorts of diagnostic equipment at night and almost never saw a drop.

        After about 2 weeks of troubleshooting we determined that every time the elevator passed the cable infrastructure which was run down the elevator shaft, it would cause the terminal sessions to drop...

        Imagine everytime you left the building at 2am after not being able to find a problem; to have someone call you and say "just as you were leaving the terminals reset..."

      • by Ungrounded Lightning (62228) on Tuesday December 21, 2010 @08:18PM (#34636196) Journal

        IBM 4341 mainframe in our data-center that would just shut down regularly every Friday night, around the same time ... shutdown coincided with the approach of the USS Lexington ... Apparently the radar from the ship was strong enough to ... trigger a shutdown.

        Another IBM radar story (Third hand: CE involved -> my brother -> me.)

        Shortly after the "Foreign Attachments" suit required IBM to allow other companies' equipment to be directly connected, there were a number of multivendor projects, of which this was one.

        Each component worked fine in the respective labs. But the first integration of the whole system took place at the final site. (Why rent some space, hook it all up, get it running, tear it down, move it, and hook it up again, when you can do it once at the final site?) So they hooked it up and nothing worked right.

        Several weeks of hair-tearing and finger-pointing by exasperated CEs from several companies ensued. At one point my brother's buddy had time on his hands and decided to fix the really annoying flickering fluorescent tube. He turned off the lights - and the tube kept flickering. WTF?

        He called the other CEs over and demonstrated this. Then they all took a quick look around the environment to see what might be causing it. It was a short look: The wooden building was right next to the antenna for the airport's search radar.

        Lined the room with conductive material. Everything started working just fine. Handshakes all around, exit stage left.

  • by QuantumBeep (748940) on Tuesday December 21, 2010 @03:59PM (#34632952)

    We constantly hear about needing to "program defensively" and test for "can't happen" conditions.

    Here's one for defensive engineering.

    • I would hope, for any remotely adequate mechanical engineer or industrial designer, (or heck, even an interior designer) "potential wear, degradation, and derating of the surface coatings on parts in mechanical contact" wouldn't even fall into the category of "reacting to a 'can't happen' condition; but simply count as standard diligence.

      Now, I can imagine the organizational dysfunction where the guy speccing finishes might be told "low cost, attractive, applies to metal and aesthetically compatible with
      • by MBCook (132727)

        Shouldn't there be something (like a current limiting resister) so that even if a mistake like this DID happen, it couldn't draw too much current and cause these kind of problems? Surely you could detect how much current is being drawn and shut it off if it's too much. Other devices do that.

        Also, if the brownouts are the problem, why doesn't the Kindle notice the voltage getting low and complain? At this point, why would any electronic device be able to get to such a low voltage point where it can't contin

  • It seems to me the design flaw is with the cover rather than the Kindle. Who in their right mind would put a metal hook into an electrically "live" slot unless they intended to draw electricity? Polycarbonate would do the job, or even hard rubber.

    But I have to admit I had never even noticed those side slots on my Kindle 3 - until I read this story.

    • Re:Metal hooks? (Score:4, Interesting)

      by gander666 (723553) * on Tuesday December 21, 2010 @04:25PM (#34633372) Homepage

      Being a product manager, I would guess that whomever wrote the spec for the cover latch, specified dimensions, and what is required for it to be a sturdy fit. But that they forgot to specify that there was to be no electrical connection or conductivity between the tabs.

      The Winning bidder probably chose to make the bracket out of brass (guess here) to ensure dimensional integrity, and because a plastic mold for a thermoplastic injected part would be a couple tens of thousands of dollars.

      But, I would bet my last dollar that someone at QA at Amazon figured this out, and specified that the bracket had to be painted with a non-conductive paint as a band aid.

      This is how trivial, serial bad decisions come back to bite you in the arse

      • by MBCook (132727)

        Or, you could design the product not to have connections to it's battery exposed like that. How about a simple physical switch (not unlike those in headphone jacks) that prevents power from going through the latch slots unless a little switch is pressed in.

        That way, if the latches are full you could have the current on, and if they had a little cutout slot in them or a notch, the power could be off. Then you could use a piece of metal, even without paint, and it wouldn't cause this problem so long as it ha

  • Wow (Score:5, Funny)

    by RightwingNutjob (1302813) on Tuesday December 21, 2010 @04:03PM (#34633006)
    a malfunction in a high tech device that actually can be fixed with duct tape
  • by specialperson (1963158) on Tuesday December 21, 2010 @04:04PM (#34633034)
    First, his meter's reading 2 Megaohms, not 2 Ohms. I guess he's not much of an "Electronics Person". Second, it would appear that he's measuring conductivity though his body to achieve that number. Both of his fingers are touching the probe tips.
    • With a multimeter with only one ohm setting, no frequency, HFE or cap tester, he's just a glorified plug-tester.

      Now, can someone with a Kindle cover and a proper meter please test for us to settle it.

    • First, his meter's reading 2 Megaohms, not 2 Ohms. I guess he's not much of an "Electronics Person".

      Second, it would appear that he's measuring conductivity though his body to achieve that number. Both of his fingers are touching the probe tips.

      That was the first thing I thought of when seeing the picture as well... Thank goodness he posted the full res version of that so we can very clearly see the M on the meter. What a maroon.

    • by vlm (69642)

      First, his meter's reading 2 Megaohms, not 2 Ohms. I guess he's not much of an "Electronics Person".

      Second, it would appear that he's measuring conductivity though his body to achieve that number. Both of his fingers are touching the probe tips.

      He would get about the same resistance if he skipped the whole leather cover thing and just held the meter probes. You'd think they'd notice something like that.

      And Connectify earns their spot on "57 Lamest Tech Moments of 2010"

      seriously though pretty much any tightly fitting leather case will probably put the device under some continuous strain, probably leading to something internal flexing, then a reboot. With those nice strong metal clamps gripping the case tightly, I could imagine it.

      • by makomk (752139)

        He would get about the same resistance if he skipped the whole leather cover thing and just held the meter probes.

        He may well effectively be doing just that. Given how difficult it is to make reliable electrical contact with a small area of metal using standard multimeter probes, together with the general level of competence displayed, I wouldn't be surprised to find the probes aren't in contact with the metal at all.

      • by blair1q (305137) on Tuesday December 21, 2010 @04:38PM (#34633586) Journal

        Reminds me of the Garmin Edge reboot and USB connectivity problems. They didn't use a flex to connect the rear case electronics to the front case electronics, they used a riser with flexible fingers. Fair enough, but they integrated this riser with the mini-USB port jack, and because of that surrounded the case opening with a thick gasket of stiff rubber. See what's coming? When the case is closed, the gasket puts a high spring force between the two circuit boards right where the fingers are mounted, reducing the spring force the fingers can apply to their mating contacts. When using the unit in a vibrating situation (you know, like on a bike, especially an MTB in typical MTB terrain), intermittent loss of contact results in power-bus glitches, which results in inadvertent power-cycling. And these things boot slower than a netbook running Windows Vista, so not only is it wearing on your data-gathering sensibilities, it's fracking boring waiting for the thing to come back to usable state so you can sweat while you wonder if it'll blow itself out again.

        Also, repeated insertion and removal of the USB connector leads to loose USB connectivity, and reboots while plugged into the computer.

        It took Garmin nearly a year to "figure it out", while everyone online who knew what the insides looked like knew within seconds what was going on. And Garmin's solution was to introduce the next model (at 3X the price). People owning the buggy model were offered a chance to mail in the device for a fix, but most were out of warranty, and the fix was not reputed to be a sure one.

        Moral: Never -- ever -- trust a corporation when the potential for money flow is negative to them.

    • by pz (113803)

      First, his meter's reading 2 Megaohms, not 2 Ohms. I guess he's not much of an "Electronics Person".

      Second, it would appear that he's measuring conductivity though his body to achieve that number. Both of his fingers are touching the probe tips.

      And unless the metal we assume exists between the hooks were exceedingly thin, or the spots where the paint was rubbed off exceedingly small, 2 ohms would not be a reasonable resistance for a wire of that length. If the spots where the paint was rubbed off were small enough that the contact resistance was 2 ohms, then they would be difficult to find with the probes. Metal at macroscopic sizes conducts quite well. A quick check with my handy Fluke 179 shows that a clip lead, roughly the same length as the

      • by vlm (69642)

        If the spots where the paint was rubbed off were small enough that the contact resistance was 2 ohms, then they would be difficult to find with the probes. Metal at macroscopic sizes conducts quite well. A quick check with my handy Fluke 179 shows that a clip lead, roughly the same length as the distance between the two hooks in the Kindle cover, has a resistance that is below the threshold of measurability on the meter (0.1 Ohm).

        http://en.wikipedia.org/wiki/American_wire_gauge [wikipedia.org]

        Lets assume the metal wire in the kindle case is copper and a foot long (makes it simpler), you'd need a wire gauge quite a bit below 40 AWG to get 2 ohms in a foot. seeing as thats about as small as commercially available, yet is till way too big, and yet is only 3 thousandths of an inch across... doing some reasonable extrapolation thats a piece of copper about a thousandth of an inch in diameter. That will snap before it falls off the assembly line.

        I su

    • His left finger looks like it's touching, but I'm not entirely convinced the right one is.

      Regardless of that nitpick, it is still showing 2 MOhms, which shouldn't draw nearly enough power to do anything he thinks it's doing.

      Purely speculation, but does it look like the black probe is even touching the metal...?
    • by mrjatsun (543322) on Tuesday December 21, 2010 @04:32PM (#34633492)

      My son's Kindle has this problem... I removed the cover the other day and it has not had it sense. I just broke
      out the multimeter, I was unable to get an electrical path even when scraping the paint on the hooks.

      • Re: (Score:3, Interesting)

        by Sir_Gimpy (861124)
        I just tested with my kindle case and a Fluke 73III meter. I've had a few restarts, but I figured they were due to the new software or the fact that I dropped it a while back. I have tested the hooks as well as I can. They are NOT connected as far as I can tell in my Kindle3 Case. I scratched the paint off and I was able to get a connection between two points on the same hook, but never anything across the hooks. (What I am trying to say is that I verified that my meter was making a connection to the metal
    • by mangu (126918) on Tuesday December 21, 2010 @05:02PM (#34633974)

      First, his meter's reading 2 Megaohms, not 2 Ohms. I guess he's not much of an "Electronics Person"

      Maybe he's not much of a "Photography Person". It's not easy to snap the shot while holding two probes in contact with the metal hooks. The photo is there to give an idea of how it went, it's not supposed to be an accurate document of the measured value.

      The "non-electronic" persons are those who calculated the electric resistance of the metal hooks. I got my EE degree in 1979 and have been an electronic hobbyist since 1969 and I know that 2 ohms is typical of what you may get from measuring a short circuit. There's always some dirt and oxide around, it's not usual to read the true resistance of the metal itself.

  • Phaser on overload. (Depending on the short-circuit current capacity of the Kindle's battery and the resistance of the shorting bar,that is.)

  • 2 Ohm or 2 Megaohm? (Score:5, Informative)

    by Danh (79528) on Tuesday December 21, 2010 @04:06PM (#34633062) Homepage

    The linked article at Connectify says they measured a resistance of 2 Ohm, but on the picture I read 2 MOhm!

    Check yourself with the large version of the picture.

    • Careful inspection of the picture leads me to conclude that the air gap between the negative probe and the hook probably does have a 2 MOhm resistance
  • The upside (Score:3, Funny)

    by willoughby (1367773) on Tuesday December 21, 2010 @04:06PM (#34633068)
    Well, as long as it's rebooting that gives you at least a few moments while Amazon cannot delete your files...
  • The line starts over there to bash on Amazon and/or the Kindle.

    Don't worry, you don't need a valid or even sensible reason to get your chance. Just be frothing mad and they'll let you in.

    • by jeffmeden (135043)

      Where is the line for bashing the "expert" at Connectify who in the same sentence derides anyone who doesn't know what an ohm is, and demonstrates that he doesn't know how to use a multimeter... That's the line *I* want!

  • Whomever designed the case should be fired with extreme prejudice. They're lucky they didn't fry a significant number of very expensive ebooks with something this stupid. If I were a victim of this I'd demand a replacement kindle while I was at it... no telling what long term affect this had on the device.
    • by vlm (69642)

      Whomever designed the case should be fired with extreme prejudice. They're lucky they didn't fry a significant number of very expensive ebooks with something this stupid. If I were a victim of this I'd demand a replacement kindle while I was at it... no telling what long term affect this had on the device.

      The article pic shows about 2 megs not 2 ohms of resistance. Thats not too unlikely for a persons dry skin. If you think about what leather is made out of, it makes sense that a human body and a leather case would have about the same resistance. I think it would be safe to assume the kindle engineers designed it to survive dry skin contact, so the case designers building their case out of dry animal skin is not exactly the dumbest intersection of the fashion designer and electronics world I have ever see

    • Re:KaWow (Score:5, Interesting)

      by anUnhandledException (1900222) <davis.gerald@gmail . c om> on Tuesday December 21, 2010 @04:38PM (#34633588)

      How exactly do you fry an ebook?

      A demonstration for you:
      1) Purchase Kindle
      2) Purchase and download 1000 ebooks to Kindle
      3) Throw kindle into incinerator
      4) Purchase new Kindle and click "Sync"
      5) 1000 ebooks "magically" appear on new kindle and more remarkable show no signs of fire damage.

      • Re:KaWow (Score:5, Funny)

        by Jtheletter (686279) on Tuesday December 21, 2010 @04:44PM (#34633688)
        Fahrenheit 404?
      • by DragonWriter (970822) on Tuesday December 21, 2010 @06:53PM (#34635408)

        How exactly do you fry an ebook?

        A demonstration for you:
        1) Purchase Kindle
        2) Purchase and download 1000 ebooks to Kindle
        3) Throw kindle into incinerator
        4) Purchase new Kindle and click "Sync"
        5) 1000 ebooks "magically" appear on new kindle and more remarkable show no signs of fire damage.

        1000 minus, of course, the number of those ebooks that Amazon has decided can no longer be downloaded since the time they were downloaded into the old Kindle. Now, depending on how your tastes in ebooks line up with Amazon's whims in maintaining their public interest, that difference might be zero, or 1000, or anywhere in between.

  • This arrangement effectively shorts the power supply - the users are lucky that it doesn't destroy their devices. They're even luckier that the problem seems to go away once the shorted power supply has crashed the Kindle, and presumably turned the power supply off. A shorted battery with all its power flowing through a nice, flammable animal product would be even worse - either for the leather or the battery.

    Disclaimer: not a Kindle owner, just sowing a bit of FUD.

    • It's probably not the power bus directly. Most battery packs and I would assume these devices with exposed leads have built-in current limiters that shut down on an overcurrent condition. I recall seeing a news story a while back about knock-off cell phone batteries that didn't have this circuitry being a fire/safety hazard.
  • by wfolta (603698) on Tuesday December 21, 2010 @04:29PM (#34633422)

    I got a Marware cover for my iPad and love it. One issue it had though, was that the iPad's compass simply never worked. It always gave me the Figure-8 Shake warning, and I eventually thought that perhaps my iPad was defective... Then one day I noticed that the flip out "foot" in the cover is held in place by two magnets. Whoops. Really only an issue if you use a compass app or if you want to figure out directions while not moving, but an interesting design issue none-the-less.

  • by Gothmolly (148874) on Tuesday December 21, 2010 @05:22PM (#34634266)

    So all you have to do is scrape the paint, and wire in a little LED and resistor, and you have a free lit case? Thanks Amazon !

  • by Caerdwyn (829058) on Tuesday December 21, 2010 @05:35PM (#34634464) Journal

    Jeff Bezos: "You're reading it wrong."

    :)

  • by wjousts (1529427) on Tuesday December 21, 2010 @05:48PM (#34634660)

    The paint on the hooks wears off and shorts out the device!

    There now was it so fucking hard to put that in the fucking summary?

  • by scdeimos (632778) on Tuesday December 21, 2010 @08:20PM (#34636206)

    As already mentioned above, the multimeter in the picture is reading 2.164 megaohms which is quite a high resistance and would make no difference at all to the operation of the Kindle.

    It seems that the blog owner has realised their mistake and replaced their blog entry with the content of another, but not before it made it's way into Google Cache [googleusercontent.com]

    For those interesting in seeing the high-resolution "Oopsie" image, it is here [blogspot.com].

  • Am I going blind? (Score:3, Informative)

    by McPierce (259936) on Tuesday December 21, 2010 @09:21PM (#34636704) Homepage

    I'm on the Connectify blog and I don't see anything about leather cases or flaws. When I search the page for "leather" I only see the tag and nothing else.

You can measure a programmer's perspective by noting his attitude on the continuing viability of FORTRAN. -- Alan Perlis

Working...