Prev: Panasonic 27XF25CW 27" TV Horizontal Line until warms up
Next: Maplin/Trust 500 UPS
From: b on 15 Mar 2010 10:19
On Mar 15, 3:14 am, "Arfa Daily" <arfa.da...(a)ntlworld.com> wrote:
> "William R. Walsh" <newsgrou...(a)idontwantjunqueemail.walshcomptech.com>
> wrote in messagenews:j-WdnVY6usNsGQDWnZ2dnUVZ_uOdnZ2d(a)mchsi.com...
>
>
>
> > Hi!
>
> >> I bought a Philips DVP 642 DVD player in December 2005, and 2 months
> >> shy of 2 years, it failed
>
> > This sort of thing is completely inexcusable. I suppose it's brought on by
> > cost, or at least I hope so, given how many years have gone by since the
> > electrolytic plague took place. This computer (see the sig) is running on
> > all of its original electrolytics, as is the Deskpro EN a few places down
> > and many other devices. The EN runs some of its 'lytics a little bit warm
> > to
> > the touch, but it's still going. It runs 24/7.
>
> > This machine also spends a large amount of time powered on, and was used
> > as
> > a server prior to falling into my possession sometime in 2002.
>
> > I tried to save a 16-port SMC ethernet switch after it started acting up.
> > One of the electrolytics was blown up, but a replacement did not restore
> > normal operation. My guess is that damage to the other circuitry had taken
> > place.
>
> > So it *can* be done. I suppose the only reason it doesn't always work out
> > is
> > due to cost and the odd defective unit.
>
> > William
>
> Further circuitry damage can often be a consequence of failing smps
> secondary-side electrolytics. If the supply monitors say the 12v rail for
> regulation feedback, and the filter cap on that rail goes bad, the resulting
> hash and ripple can appear to the sensing circuit as a low output. This
> causes the m/s ratio of the chopper drive to open up in an effort to restore
> the rail to the correct value. As all the other rails are tightly
> magnetically coupled to the bad rail as a consequence of them all sharing
> the same transformer core, the end result is that the 3.3v and 5v rails can
> go sky-high, causing a trail of catastrophic damage to various LSIs in the
> equipment.
>
> Arfa

I've seen this with quite dramatic consequences. 2 years ago a friend
brought me a budget DTT-DVD player which was dead. Opened up,and
several caps had literally exploded, but only one was in the psu. the
others were downstream, on the main/processor pcb. Clearly there had
been some catastrophic voltage rise.

The other week I picked up a DTT set top box. Same thing - lots of
exploded caps, with only the legs still on the pcb. The few remaining
electolytics on the main pcb were bulging.
Needless to say, both these units were promptly scrapped as BER.
The lesson is: when something is acting erratically, don't wait before
changing the caps!
-B
From: William R. Walsh on 16 Mar 2010 23:12
Hi!

> Further circuitry damage can often be a consequence of failing smps
> secondary-side electrolytics.

Every cap in the switch's SMPS was fine, and the outputs were right on. A
large (filter?) cap on the main board blew its top. It looked like it had
been working up to this for a while. It had been working great up until the
moment it fell off the wagon.

A replacement cap didn't help much, and I sure couldn't complain about six
years of service from a used switch that I paid all of $10 for. I saved the
SMPS for later use and recycled the rest. A new D-Link switch replaced it.
What struck me was how the components have shrunk. The SMC switch had
several large ICs, two of them responsible for switching and one that
appeared to be a sort of processor (for what I have no idea, it was not a
managed switch). There were some artifacts near the processor for RAM and
ROM, probably used to add management functionality?

The D-Link switch is based on a Realtek platform. It runs cooler, performs
just as well and is a simpler design.

William


First  |  Prev  | 
Pages: 1 2 3 4 5
Prev: Panasonic 27XF25CW 27" TV Horizontal Line until warms up
Next: Maplin/Trust 500 UPS