Re: gEDA-user: Hello, Bills missing transistors and Attribute questions

[Peter Clifton <pcjc2@xxxxxxxxx>]

On Sun, 2008-03-30 at 04:01 -0700, Mark wrote:

> My names Mark and I am from Malaysia. Im a newbie and
> recent convert to the world of gEDA and have been
> really struggling trying to understand it all!

Hello, and welcome...

> The first thing is that when one looks thru Bill
> Wilson's really helpful gsch2pcb_Tutorial, there is a
> link to a library of transistor symbols and elements
> that Bill created. The URL is in the "Custom File
> Elements" section of the document:
> http://www.geda.seul.org/docs/current/tutorials/gsch2pcb/gsch2pcb-libs-20040110.tar.gz
> 
> Unfortunately it leads to an object not found page.
> Would anyone have a copy of this library? 

I don't unfortunately, perhaps someone else will be able to help though.

> So, if I have a BC550 NPN GENERAL PURPOSE TRANSISTOR,
> what are the preferred values to give:
> 
> Device=?
> Description=?
> 
> My guess is that:
> 
> Device=NPN TRANSISTOR
> Description= BC550 NPN GENERAL PURPOSE TRANSISTOR
> 
> Would this be correct or would you suggest some other
> values?

Those sounds good to me. As a general rule though, these are for your
benefit. This said, SPICE or gnucap backends may attach meaning to the
Device= attribute. In the existing symbols, see mostly:

device=NPN_TRANSISTOR

There is also device=SPICE-NPN

(Note its "device" not "Device", similarly "description").

Also might be of interest, is the attribute:

"documentation=...." If you set the attribute to a URL, then you can
access that URL in a web-browser from the menu
"Hierarchy->Documentation...." when you have a pacticular component
selected.

> My next problem is with the pins. I think what is most
> important is that the pin numbers of the symbol pins
> match the physical pin numbers of the particular
> transistor when placed inside the PCB. Difficult with
> TO92 as there are lots of em ;)

Yes, a common problem. Also, rather surprisingly I can't find a TO92
package in PCB. My suggestion would be (if you're targeting PCB for
board design), to find the right footprint in PCB and match your
pinnumbers in your specific symbol.

Al Davis or Start Brorson may be able to answer more about the
requirements for using gnucap or some SPICE variant to simulate your
design.

You might take a look at:

http://www.brorson.com/gEDA/
http://www.brorson.com/gEDA/SPICE/


> As you know... these small package transistors have
> specific manufacturer numbered legs. In the case of my
> BC550, I have a Leg #1 (Collector), Leg #2 (Base) and
> Leg #3(Emitter).
> 
> Does it matter say if pinseq=1 is where transistor Leg
>  #2 goes? Can we attach any Leg # to any pinseq#?

pinseq defines the output ordering for the spice backend. (And allows us
to reference specific pins which might have different pinnumbers when
using "slotted" components (e.g. one opamp symbol for all opamps in a
chip with 2 or 4 inside).

For SPICE, the order you _need_ pinseq incrementing is:

collector (pinseq=1), base (pinseq=2) , emitter (pinseq=3). If you were
modelling a transistor in an IC design, you could also hook up to the
silicon substrate of the transistor with pinseq=4.

> On the subject of pinnumber, what would a better
> approach be? Should I use values of 1, 2 and 3? Or
> would it be smarter to use values of E, B and C for my
> transistor symbol?

You can do either, so long as the pin names in PCB match the pinnumber
in gschem. pinnumber also works with letters - it just means that you'd
be making specific PCB footprints for all the TO92 pinout variants you
want to use.

> On the subject of pintype for transistors, what should
> I be using? pas? I am uncertain how open collector and
> emmiter fit into the picture here. As I might wish to
> one day run spice... I would really appreciate some
> info on what values I should be setting for the E, B &
> C legs on my transistors.

pas sounds good to me, but its probably not generally appropriate. Those
values are only used by the design rule checker, and appear to be much
more geared to digital logic - where there are clear power, input,
output pins etc.. The purpose would be to check (for example) there
aren't two outputs connected together, or inputs which aren't driven.

> For pinlabel, would it be correct to say that I do not
> need to include this attribute unless I am connecting
> a particular transistor leg to another schematic
> somewhere else? Or do I need to include this
> attribute. If so.. what would I label them for the E,
> B & C legs on my transistors?

I'd add it if it improves clarity. I generally use them to show the
signal names of signals on more complex chips. Eg. "IRQ" "D0" "A12",
etc..

> Well guys... thank so much for your time going thru
> this. Really appreciate it!

No problem, we hope you enjoy using gEDA - any further queries, just ask
here.

Best wishes,

-- 
Peter Clifton

Electrical Engineering Division,
Engineering Department,
University of Cambridge,
9, JJ Thomson Avenue,
Cambridge
CB3 0FA

Tel: +44 (0)7729 980173 - (No signal in the lab!)



_______________________________________________
geda-user mailing list
geda-user@xxxxxxxxxxxxxx
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user