On 03:57 PM 5/09/2001 -0700, Brad Velander said:
>Abd ul-Rahman,
> My point was only that the EBC, AK, GDS is redundant and stupid on
>the schematic because the symbols convey the function of each pin already.
>Holding only the printed schematic in your hand, the EBC, GDS and AK conveys
>what information that you didn't already know from the symbol? What does it
>tell you about where to find the Base pin on a physical device on a PCBA?
>The answer to both questions is "Nothing", right? Thus it is redundant,
>useless and clutters up your schematic with useless information.
I was going to let this one through to the keeper but nah...
1) There is no Sch clutter, the pins numbers are hidden
2) The information is not redundant, I can set a design rule that gives
different track thicknesses to all gate connections - easily. For instance
- making a Pad class with all the gates is simple as in
the Design|Classes|Pad dialog tab I can Add a new class and add
*-G
to the non-members filter and up will come all the 'G' pins (along with the
-GND pins). Makes making a pad class easier. Try that with 123-style
numbering. (I know this is a little contrived but the point is that with
EBC or GDS pin numbering you can do extra filtering that is not possible
with 123-style.)
3) Someone creating a new part knows they have to allocate EBC or GDS to
the new footprint they are creating. They are more likely to remember to
the check data sheet with the alpha numbering - this has been my long term
experience. Since the numbering starts by default as 1,2,3 etc as you
plonk down pads it is easy to forget to check that you have made the
footprint correctly. In the EBC-style the footprint is unusable until the
alpha pin numbers are provided. This is more likely to trigger a check of
the data sheet. Even if those creating footprints would normally check pin
numbering before using a footprint, mistakes happen at 2:00 am as
Production are banging on the door. If the netlist load/synch chokes you
get a healthy reminder.
> I agree that there is no one ideal solution. With the system I use,
>I have one SOT23 footprint for a particular package size. The symbols are
>designed to match that footprint.
>
> If you receive information as a netlist only, then the customer is
>responsible anyways right? They have to have used the right symbol with the
>right part. If they didn't how can you possibly tell?
Well, if a SOT-23 pin is labelled 'G' it is a fair bet that it is a
gate. A skilled designer has more information at their disposal to aid
on-line grey matter DRCs as well as potentially improving the layout. If
it is labelled '1' then you have no method of even starting to bring high
level interpretive skills to bear. Even if the schematic is on your desk
as you work, it is easy to get buried in a layout and not refer. The pin
number provides an extra memory jog.
Surely, Brad, you are not saying that PCB work is merely a join-the-dots
draughting exercise?
> Also, I don't display the pin numbers on transistors or diodes. As
>you said it would convey no more clarity because of the mixture of
>manufacturer numbering/labeling schemes. Likewise the EBC and GDS conveys no
>more information either (without the datasheet). Right? Stalemate!
Wrong, while laying out the PCB knowing that a pin is a gate rather than a
'1', whatever that is, is certainly useful. I have a TO-220 part. Which
track is the control trace and so can be thin, which are the power
tracks. I know easily without having to remember to refer to the Sch, do
you? So it does convey useful info. The pads on the PCB are a constant
reminder of function and will jog your memory as you work.
In many years of PCB work, EBC has never failed me. Lloyd's is the first
time I have seen an issue with alpha characters as pin numbers that I
recall. But my memory is not so good anymore and I have not used any other
CAE package seriously for many years.
Years ago I saw many mistakes made with 123-style numbering, many of these
were repeat mistakes. I forced a change to the system where sch parts had
EBC etc and the PCB parts remained at 123. This *forced* the PCB designers
and engineers (often one and the same) to go to databooks to check pinning
and then manually (with careful global edits) change the PCB pads to match
the sch and allow the netlist import to succeed. This was a fairly drastic
measure but it broke the cycle and made the engineers and
techs sufficiently aware of the problem that it hasn't been such an issue
recently. More recently we have been using PCB footprints that are fairly
specific. So, for instance, we have a BC8XXREFLOW footprint, which has E,
B, C pad numbers. A designer is welcome to use this footprint for any part
as long as they have checked that it does indeed conform to the pin
functions of the BC846, BC856 etc series parts. But if they use it
incorrectly then they are hung over the balcony while the vultures peck
their eyes. Coz they knew better and were just being lazy.
We use EBC-style numbering mainly as some protection against human failings.
Bye for now,
Ian Wilson
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
*
* To leave this list visit:
* http://www.techservinc.com/protelusers/leave.html
*
* Contact the list manager:
* mailto:[EMAIL PROTECTED]
*
* Forum Guidelines Rules:
* http://www.techservinc.com/protelusers/forumrules.html
*
* Browse or Search previous postings:
* http://www.mail-archive.com/[email protected]
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *