Re: gEDA-user: Newbie PCB & gschem

[Magnus Danielson <cfmd@bredband.net>]

From: Dave McGuire <mcguire@neurotica.com>
Subject: Re: gEDA-user: Newbie PCB & gschem
Date: Sun, 23 May 2004 11:40:33 -0400
Message-ID: <87153A84-ACCF-11D8-A0CC-000A95C4A1C2@neurotica.com>

> On May 23, 2004, at 8:48 AM, Bob Paddock wrote:
> >>  A document explaining the basics and
> >> the terminology (I still don't know what a "thermal" is) might help.
> >
> > A "thermal" is usually used when connecting a pad or pin to an 
> > interlayer.
> >
> > It is generally a shape of "X" or "+", where only the four end points 
> > connect
> > to the interlayer.
> >
> > If you do not use a "thermal" the entire interlayer plane becomes a 
> > heat-sink
> > that sucks away the heat, hence thermal, making it impossible to 
> > solder that
> > point, because you can never get it hot enough to melt the solder.
> 
>    ...and if you finally get just enough heat in there to get it melted, 
> you will likely wind up with your soldering iron permanently welded to 
> the PCB.  Been there, done that...doh! :-(

What one needs to do to get around this is to pre-heat the PCB with a hot-air
gun or better even, put it on a hot plate. The temperature-gradients are *much*
lower that way than on a room temperatured board so soldering is easier. When
we solder and desolder BGAs we have regulated heating from one side and hot air
on the "component" side (whichever side that is).

However, one really *should* use thermals on all layers to start with. It is
a sound engineering-principle. Never hook up a pad/hole directly to any surface
of copper around it without the thermals. You could also do with a pure pad/
hole and one or a pair of lines over to the Cu-surface, it is about equalent,
but just putting a pad in the middle of a surface is a no-no.

Cheers,
Magnus