Re: why do Panamaxes die?


"Dave Platt" <dplatt@xxxxxxxxxxxx> wrote in message
news:vgiac6-sdb.ln1@xxxxxxxxxxxxxxx
In article <QemdnScgwdqwGW_UnZ2dnUVZ_vOdnZ2d@xxxxxxxxxxxx>,
Soundhaspriority <nowhere@xxxxxxxxxxx> wrote:

A pretty universal aspect of the Panamax experience is that these gadgets
die with no provocation. The one I have open before me was completely
protected from downstream surges. Without a schematic, I guessed that the
failed component would be one of the three small value electrolytics
exposed
to the AC ripple of the signal wire, but with a Lissajous box, they all
measure identically, apparently good. Before it failed completely, the
relays would cycle on and off.

Can a Lissajous box miss important leakage in a cap? The three identical
caps have approximately the same area in their circles.

I believe that in small 'lytics of recent vintage, a more common
failure mode these days is an increase in the cap's series resistance.
They don't necessarily go leaky, but if their ESR rises, they won't
filter as well.

Over the past decade or so, a lot of electronic devices have failed
due to "capacitor rot" (internal electrolysis, gassing, leakage of
electrolyte, and increase in ESR). A case of industrial espionage
gone seriously awry, reportedly... millions of capacitors were made
with a bad electrolyte formula which breaks down over time. I've had
a modem, an Apple 802.11 base station, and a DVD player fail from this
problem.

The best device to diagnose these (short of seeing the caps bulge and
leak) is an ESR meter... or one of several simple haywire arrangements
(e.g. signal generator, a couple of resistors, and an oscilloscope).

See

http://ludens.cl/Electron/esr/esr.html
http://octopus.freeyellow.com/esr.html

for some ideas. I just (this week) finished building a meter based
on the Ludens design (all junk-box components) and it works like a
charm. Good electrolytics read under .1 ohm ESR. A pair from a DVD
player power supply which had started to bulge and leak (and which
were causing excessive ripple on the supply which was causing a
deterioration of the video output quality) read around .7 and 1.2
ohms.

The AnaTek Blue is an inexpensive commercial ESR meter, of which I've
heard good things. Same idea, different implementation.

I haven't tried to look at one of these caps on a Lissajous
analyzer... how well such an analyzer can detect the rise in ESR would
depend on the design, I suppose.

You could just try replacing the suspect caps with new ones
(preferably, 105 degree C types intended for high ripple current
switching-supply applications - they'll last longer than generics) and
see if that fixes the problem.

Thanks, Dave. I actually have an ESR meter sitting on the shelf, but had
forgotten what a good ESR reading is. I was hoping that one of the caps
would "stick out" in the Lissajous display. There are only three of these
caps per unit, so I'll try replacing them.

Bob Morein
(310) 237-6511



.

Quantcast