[svlug] Laptop horrors: AC Power supply breakage

Mark S Bilk mark at cosmicpenguin.com
Tue Nov 20 08:56:02 PST 2001

In-Reply-To: <3BFA29CD.9080908 at quiotix.com>; from jbs at quiotix.com on Tue, Nov 20, 2001 at 02:00:45AM -0800

On Tue, Nov 20, 2001 at 02:00:45AM -0800, Jeffrey Siegal wrote:
>Karsten M. Self wrote:
> > There's this thing called a design envelope, Raph, and laptops falling
> > off of desks should fall within this.  Not frequently, needs be, but
> > some basic shock tolerance.
>Maybe it should be, but I'm afraid it really isn't.  Most laptops are 
>very shock sensitive and can be damaged even by putting them down on a 
>table too hard.  Probably the number one complaint about laptops is damage.
> > You'll note that automobile manufacturers have these little events
> > they call "recalls".  That's not 'coz they want to do them.
>It's mostly 'cos they're worried about the liability that could occur 
>should someone get injured. Not applicable to the issue of a fragile laptop.
>I agree caveat emptor is applicable here.  Laptop reviews should include 
>durability testing (unfortunately, should is not the same as do) and 
>buyers should pay attention close attention to these tests.
>Of course, none of this excuses bad customer service, or refusal to 
>honor a warranty on a failure which has not been clearly shown to be due 
>to damage.  It could well have been a bad connection to begin with, or 
>even damaged prior to delivery.  There is no way to know that your drop 
>caused the problem, especially since it did not occur immediately 
>thereafter.  The manufacturer should give you the benefit of the doubt, 
>unless they can show otherwise.

If the motherboard and the power supply are soldered directly
together, instead of using short jumper wires to connect them,
this is bad design on its face, rendering the machine much 
more susceptible to being damaged by the inevitable minor 
impacts, like someone bumping it while it's being carried 
in a crowd.  Those two relatively heavy components are going
to move in different directions, and twist with respect to
each other.  Designing the machine so that the resulting force 
ends up stressing the solder bond between the two parts, and 
worse, stressing the bond between the copper foils and the 
epoxy boards underneath them, is a violation of basic, simple,
well-known engineering practice, and should render the 
manufacturer responsible for repairing the damage that ensues.

The machine must be designed to withstand some expected degree 
of impacts without sustaining damage.  After all, it's sold as 
a _portable_ computer!  Anything portable is going to get 
bumped if it's used in the mode that it's advertised as being 
capable of, namely being carried around.  If it can't withstand 
such _unavoidable_ impacts, then it's being sold under false 

The degree of impact can be quantified as acceleration (G force)
and rate of change of acceleration (labelled in physics by the 
term "jerk" -- http://home.earthlink.net/~tdp/jerk.html).  
There are testing labs that will subject a piece of equipment 
to precisely quantified abrupt movements so the manufacturer 
can determine if it survives without damage.  There are also 
very cheap little indicator devices that can be incorporated 
into equipment that show (perhaps by a color change) if it 
has ever been subjected to a movement beyond a certain degree 
of acceleration or jerk; I think some disk drives have these 
glued on -- they're about the size of two pennies in a stack, 
if I'm remembering correctly.  

The computer's specs should say how much G and jerk it can 
withstand without damage, and the indicator should change color 
beyond that.  This provides an objective determination of 
whether the failure was due to a manufacturing defect, or to
out-of-spec force encountered during use, and thus whether the 
manufacturer or the customer should pay for the repair.

In the absence of such good engineering practice, I guess it 
all comes down to the thickness and cushiness of Karsten's 
carpet, to decide whether the impact falls within the range 
of expected portable usage.  But some equipment is designed 
so it can be dropped from 6 feet onto a concrete floor, 
by incorporating rubber isolators between the case and the 
internal components.  These are just little rubber grommets 
that decrease the instantaneous levels of acceleration and 
jerk that the components are subjected to when the case hits 
the floor.  Disk drives have them inside, between the case 
and the head/disk assembly.  They are very cheap, and failure 
to design them in is unforgiveable.

More information about the svlug mailing list