TECHNET Archives

August 1998

TechNet@IPC.ORG

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Ahne Oosterhof <[log in to unmask]>
Reply To:
Date:
Fri, 7 Aug 1998 13:54:30 -0700
Content-Type:
text/plain
Parts/Attachments:
text/plain (127 lines)
Another penny to contribute:
Changing the layout of a circuit board can be a simple innocent endeavor or
you can be taking a walk into a minefield.
Analog boards: if the circuit is using high frequencies (and that can start
below 10 MHz) certain parts should not be close together, like inputs and
outputs or large signals and sensitive parts of the circuit, in order to
prevent crosstalk. Different voltages levels require different minimum run
spacings. Different amounts of current require different minimum run widths.
And then there are sensitive differential circuits with their own demands.
So you may have to talk to the electrical design engineer to get it right.
And of course there are the considerations mentioned in other messages, like
heat/cooling, size and weight, for which you may have to consult with the
mechanical design engineer.

Digital boards: similar problems, different considerations. In many digital
circuits crosstalk is less of a problem; if you are lucky the circuit senses
final levels, rather than transitions, but don't count on it.
Voltage levels are typically low and therefore don't present issues. Current
may become a problem, but supplies are often provided in planes rather than
runs. Run lengths may be a problem when dealing with high clock rates: runs
that carry differential signals or part of a bus may be required to be
'exactly' the same length to assure that signals arrive at the other end at
the same time.

And after all that: you won't even know that the resulting layout is
acceptable until you have done a full performance evaluation. Oops, did not
even mention issues that may show up in the assembly process, e.g. spacings
between large and small parts, where you need rooms for human or machine
fingers, or access to adjustable parts, e.g. potentiometers.

Maybe this was even three cents worth, in any case good luck.

Ahne Oosterhof, A-Laser, Inc.    (I wuz an electrical injeneer, once upon a
time)

-----Original Message-----
From:   TechNet [mailto:[log in to unmask]] On Behalf Of Larry Campbell
Sent:   Friday, August 07, 1998 05:59
To:     [log in to unmask]
Subject:        [TN] New Designer!! -Reply

Jason,

You are experiencing nothing new.  Most designers, at least the
older ones, have never recieved any formal training.  There was
none available.  It was always on the job training.

In all probability, the older boards you are going to convert, are
analog circuits and don't require the same expertise that the
boards of today require, with their higher speeds and digital
circuitry.  Yes, it was more aesthetic than it is today, but not
necessarily a bad board.  I think you'll have a harder time
regenerating them into surface mount components than
anything else, simply because the routing will have to be so
different.  You might as well not look at anything except the
schematic and revised parts list.

As you said, your company has been in the business a long
time, so the boards generated over the years must have been
good.  Don't forget that something that may appear to be an
asthetic design could have other reasons too, such as a height
requirement, heat (thermal), weight or vibration.  The same rule
applies today as it did 10-15 years ago.  Let 10 people design
the same board and you'll get 10 different layouts.

I would recommend to you to read anything you can get your
hands on.  IPC specs, trade magazines and such.  Get some
formal training if you can (It's still hard to do, even today).  Talk to
your board vendor(s) and assemblers.  They can give you advice
as to what they can, or cannot do and what they need from you.
Pick the old guys brains for their advice and help.  Most of it is
still just as useful today as it was yesterday (unless the laws of
physics has changed and nobody told me).

Good luck with your endeavor.

Larry Campbell

>>> "Markosky, Jason" <[log in to unmask]> 08/06/98
06:11pm >>>
Howdy,

I am relatively new to the designing of circuit boards so please
bear
with me and my petty questions.  The company I work for has
been in the
business of designing its own circuit boards for several years
now. I
have found out that past designers had no formal training in the
processes of designing of the boards.  What I am trying to say is
previous designers had vast knowledge in the way the designed
theoretically worked but, when it came to the layout of the board
and
where components were  placed became a  random process
throughout the
area of the board. There was some knowledge of where to put
certain
devices, but the designer  seemed more interested in aesthetics
of the
board rather than proper function.  What I am looking for is some
rules,
tips, and any information that will help me improve old and future
designs.  Also, I have been given the task of converting old
designs
with regular components on them to an all surface mount
design.  Again,
if anyone has any tips or past experience I would greatly
appreciate it.

THANKS FOR ALL THE HELP!!!

JASON MARKOSKY
[log in to unmask]
<mailto:[log in to unmask]>

################################################################
TechNet E-Mail Forum provided as a free service by IPC using LISTSERV 1.8c
################################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in the body:
To subscribe:   SUBSCRIBE TechNet <your full name>
To unsubscribe:   SIGNOFF TechNet 
################################################################
Please visit IPC's web site (http://www.ipc.org) "On-Line Services" section for additional information.
For technical support contact Hugo Scaramuzza at [log in to unmask] or 847-509-9700 ext.312
################################################################


ATOM RSS1 RSS2