On Feb 1, 2008 1:45 PM, Levente <[EMAIL PROTECTED]> wrote:
> Cool!!!! I haven't tried it yet, but sounds good. I finish some work, and I
> give it a try. My question was that how do you solve the "transistor
> problem"?

The example circuit (common-emitter.sch) is in fact an example of how
the "transistor problem" (if I understand it correctly) can now be solved.

You put down a generic BJT symbol and connect stuff to it.  This symbol
only knows that it is an NPN BJT - it has no pinnumbers, no footprint, no
hFE, but it DOES have pinseq attributes on the pins.  (I may have
misunderstood the requirements for a flow involving spice, but this is
easy to change.)

At some point you need to decide that you'll use a 2N3904, so you place
a "heavy" transistor symbol somewhere out of sight, fire up the Slot
Chooser, and select the one slot it has available for NPN BJT's.
Automagically libgeda "inherits" the 2N3904's refdes to the light symbol,
as well as numbering the pins.

To finish the solution to the "transistor problem", you need vast tables
of transistor data, and a perl script that generates only the *heavy*
symbols.  Your "light" symbol that's visually part of your circuit can stay
as it is; you only need to re-associate the light symbol with a different
heavy symbol if you want a different transistor.

HTH


_______________________________________________
geda-dev mailing list
[email protected]
http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev

Reply via email to