|
Is It Repairable?
I accept all controllers and give each a diagnostic
examination to determine if it is repairable. The question is
then - what is repairable, or more importantly, what is not
repairable? What I want to discuss here is why I consider a
board not
repairable.
The quick answer is that if the operational
amplifier fails on
a controller made with surface mount technology, it is not
repairable.
SIEG coats the entire bottom and the surface mount
component section of the top of a controller board with a clear
coating of an epoxy like material.
I'm sure it is for a good purpose, basically to keep chips
that enter the control box from shorting any component.
However it makes my life miserable. To replace
a component, I have to get through the coating, melt the solder and
lift the bad unit off without damaging the solder pad or affecting
neighboring components.
Here is a recent example. The photo is of a
100 ohm SM resistor on top of a layout ruler. The gate pulse
which turns on the mosfets is delivered to the mosfets through two
such resistors. If the mosfet fails with a high current short,
this resistor is often taken out. Replacing it is a little
like brain surgery. Cost is not a problem. I pay $0.06
each for such a resistor. I buy 50 of any size that I need in
order to keep an inventory.
While I don't like to, I can replace these
successfully.
This one is different. This is a quad
operational amplifier IC with 14 pins. Now I have to get
through the epoxy, lift the bad one off with no damage to pads or
nearby devices, replace, solder all 14 pins AND not let any
pin short to another.
Replacing a SM resistor or diode is a snap compared
to this. My success rate has been dismal and I now will now
not accept any board with a bad opamp for repair.
There are other reasons not to repair a board such
as major arcing with PC trace damage, but this is the major one. |
|
