Gigaworks S750 repaired by Regenboog Heerlen

Hi

Ok seems like together we could get to the "bottom" of this !

Here is reaction from "repair specialist":

"
The printed board is double sided and there are connections from one
side of the board through to the other. These connections are getting
corroded by the glue.
"

So this is some new information and possibly even a new hypothesis of what is happening to these boards and causing short circuits.

"Corroded" he writes. Could this be causing the short circuits ?

Hmm an interesting hypothesis.

Bye,
Skybuck.

wrote:
Hi

Ok seems like together we could get to the "bottom" of this !

Here is reaction from "repair specialist":

"
The printed board is double sided and there are connections from one
side of the board through to the other. These connections are getting
corroded by the glue.
"

So this is some new information and possibly even a new hypothesis of what is happening to these boards and causing short circuits.

"Corroded" he writes. Could this be causing the short circuits ?

Hmm an interesting hypothesis.

Bye,
Skybuck.

You cannot really stop corrosion like that.

It all depends on how the PCB is coated, as to how protected it is.

For the bottom of the board, there is HASL (hot air solder leveling), which
coats the copper with tin/lead. Then, a conformal coating goes over top of
that, leaving gaps in the coating, so you can apply solder to points that
need to be soldered.

On the top side, there is solder mask, and there can also be conformal coating
applied.

The copper tracks themselves, can have a plating applied to the top
layer. Some assemblies can have gold (precious metal) applied
over the copper. Gold is quite resistant to chemical attack. It may
take a sequence of metals, plated up, before the gold is applied.

For the PCBs I made a home in the kitchen, those had no coating at
all, and it only takes about ten years, before chemicals in the air
eat through the copper tracks, ruining the circuit. I built my own
digital clock, with vacuum fluorescent display tubes, and that's what
happened to the PCB. In school, we coated PCBs with silver on the copper
tracks, which doesn't give too much extra protection, but is better
than nothing. That isn't electroplated, and is done just by dipping
the board in a toxic chemical brew with silver in solution.

*******

The eating of the tracks, is not causing the fuse to blow.

There is a conducting path there somewhere, which is a problem
and needs to be cleaned up. The design would have to be checked
on the primary side of the buck converter, where the AC from the
wall is rectified to DC and rests at a high voltage. The
secondary side (the voltage all the Class D amplifiers share),
that should be a lower voltage, less susceptable to "flash/bang"
accidents :-)

AFAIK, your amp is ten channels, sharing a common programmable
rail voltage, and the controller chip sets the common rail voltage
high enough for the "loudest channel". The sub has three channels
tied to it, to get three times the power, leaving seven channels for
the satellites. By making all the amplifiers the same, they need the
same rail voltage, and thus, can share the same power source. It
is that common power source, that is the most dangerous part
(as it has the high voltage on the primary side). And the glue ?
I don't know if there is a good chemical way to remove it.
Scraping it, is a mistake, because of the other damages that
could cause. If there was a good solvent for the glue, I'm sure the
repair people would be ecstatic.

Paul

Some evidence is found for this hypothesis, which also happens to be an amp/audio restoration/repair guy:

http://www.cdkands.com/corroded.html

Very interesting:

"Every lead in contact with the glue is corroded. The green color indicates copper oxide. I am no chemist
but I believe that the glue becomes conductive, creating an electrolyte that allows a current flow across the
board. This sets up a condition for electrolysis to occur between the copper core of the lead and the silver/lead
tinning.
"

The lead had eroded clear through and caused the failure of the channel

In this sense/context eroded means something different from corroded... I think in this sense he means "eroded" is like this stuff start crawling across the board creating unwanted connections... apperently a few millimeters is already enough to cause havoc ?

Not entirely sure why this would cause havoc... on the picture show... since that lead is going to the resistor anyway ? Hmmm...