[Tfug] Ezgo "book-PC"

[Bexley Hall]

Hi, Ronald,

--- Ronald Sutherland <ronald.sutherland at gmail.com>
wrote:

> On Mon, May 12, 2008 at 4:54 PM, John Karns
> <johnkarns at gmail.com> wrote:
> 
> > On Thu, May 8, 2008 at 12:18 PM, Bexley Hall
> <bexley401 at yahoo.com> wrote:
> > >  I suspect it was sh*t-canned due to bad caps. 
> > >  Which brings up another question:  I've seen a
> > >  *lot* of (different) machines plagued by this
> problem.
> > >  It's not confined to the off-beat brands,
> > >  either (i.e., I have seen Big Blue machines
> > >  with the same problem).
> > >
> > >  What sort of experiences have others had with
> > >  this "failure mode"? (I want to be damn sure to
> > >  avoid any such component manufacturer!)
> 
> Don't be to quick to blame the capacitor(s), the
> system power supply can cook them. The power supply
> provides DC to the electronics, true enough, but
> its conversion processes includes a lot of noise.

Sure.  A characteristic of all switchers (hence the
L in series).

> If the power supply was
> not built correctly and was not tested well enough
> (none are) then some get
> out that include a lot of noise in the DC. Those
> capacitors on the
> electronics boards are partly for reducing noise
> from the power supply but
> mainly for noise from the electronics them self.
> They self heat due to the

They "self-heat" due to high ESR.  Good caps don't
have high ESR's for a given ripple current (that's
what the engineer's job is)

> effects of the noise and can be cooked to death if
> the power supply is
> adding to much of its own noise. Anyway one bad
> capacitor is probably the
> capacitor itself, but a board covered in puffed up,
> or leaking electrolytic
> capacitors is probably a bad power supply.

For this particular machine, I can't say -- since I
don't have the original brick that powered it.

But, I would be slow to accept this explanation.
E.g., the first machine I found this in was an IBM
Aptiva.  When told of the symptom (leaking caps),
IBM didn't argue but replaced the entire board.
If they had suspected the problem was the power
supply, they would have replaced that as well.

I think this is just a "bad cap" problem -- but one
that I haven't (yet) been able to qualify...

--don