#10441 NORM Not Tri: XO-1 and XO-1.5 response to invalid USB cabling configuration differs

[Zarro Boogs per Child]

#10441: XO-1 and XO-1.5 response to invalid USB cabling configuration differs
-----------------------------------+----------------------------------------
           Reporter:  Quozl        |       Owner:  wad          
               Type:  enhancement  |      Status:  closed       
           Priority:  normal       |   Milestone:  Not Triaged  
          Component:  hardware     |     Version:  not specified
         Resolution:  wontfix      |    Keywords:  XO-1, XO-1.5 
        Next_action:  never set    |    Verified:  0            
Deployment_affected:               |   Blockedby:               
           Blocking:               |  
-----------------------------------+----------------------------------------
Changes (by wad):

  * keywords:  => XO-1, XO-1.5
  * status:  new => closed
  * resolution:  => wontfix
  * type:  defect => enhancement


Comment:

 Let's be clear here.  A USB type A port is supposed to supply power.  All
 bets are off if a power source is connected to one, as this clearly
 violates the USB spec.

 This ticket is being closed as a "DON"T DO THAT --- IT WILL HARM YOUR
 LAPTOP" issue.
 The fact that the XO-1's didn't crash doesn't mean that continued use with
 this incorrect won't harm them.

 The underlying issue isn't the USB pwoer switch.  The same device is used
 in both XO-1.5 and XO-1.   In neither case does the USB power switch
 provide protection against power being applied to it's output.  In both
 cases, the power supplied through the USB port is being used to drive the
 internal +5V rail directly.   In the case of XO-1.5, this causes a crash
 because a gate drive voltage (+10V) is generated as a side effect of
 generating +5V.   When this switcher stops operating (because +5V is
 supplied by the external USB device), the gate drive voltage goes away and
 several critical power switches around the motherboard (such as +3.3V, and
 +2.5V to the DCON) are turned off.  On XO-1, this gate drive voltage was
 generated as a side effect of generating VCORE --- turning off the +5V
 switcher doesn't affect normal operation (it is only used for driving the
 speakers and USB).

-- 
Ticket URL: <http://dev.laptop.org/ticket/10441#comment:3>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system