At 12:37 PM 6/23/01 -0500, Tim Hutcheson wrote:
>My board design is in English units, primarily, but I have several small
>outline parts for which the footprint is in Metric units and the pad pitch
>is variable and results in off grid pad centers with respect to the English
>grid. My rules are 5/5 mil and the part is a 402 class 4 part resistor
>pack, small enough that the differences in English and Metric pitch worries
>me.
One point. I and quite a few other designers prefer to put the reference
point for most SMT parts on the part center rather than on the pad 1
center. Protel does provide this as a Set Reference option in the footprint
editor. One clear advantage of doing this with 2-pin parts is that when you
have a row of parts and you rotate one of them 180 degrees, the parts
remain aligned without any adjustment.
>My question is: what is the proper convention for interfacing this part
>into an English unit layout? Does it matter?
Not a great deal. My head is programmed in Imperial units (we call them
English in the United States), so I work on an Imperial grid.
> For example, if I define the
>part layout in metric the system accommodates by allowing me to connect
>tracks to the metric center of the pads, which is fine.
You can connect tracks to pad centers anyway using snap.
Another advantage of using Imperial units is that it is the native database
system. But it is a small advantage.
> But do the odd
>angles that result create any problems for the Gerber photoplotter creating
>the tracks in tight quarters?
Gerber photoplotters are generally, nowadays, raster plotters and don't
care about things like tight corners. Basically, if a track turns by 170
degrees, it is just another line segment to the plotter, it has no memory
of the previous segment (except that it will retain the coordinates of the
previous endpoint to help define the new endpoint if necessary. I don't
think the old vector plotters had a problem either, though I suppose it
might have slowed them down a tad. Don't worry about it.
I'm assuming that those "track in tight quarters" are actually inside a
pad. For most purposes, you could simply leave out the last, connectively
redundant, tracks. But I have a habit of always completing the final
segment even if it is entirely contained within a pad. Never can tell when
the board might encounter a utility that relies on coincidence of track
endpoint and pad center to determine connectivity. (Protel does not do
this, except *maybe* with the routines that are used for dead pad removal
on inner layers. I haven't checked. The autorouter, I think, also insists
on endpoint coincidence.)
> Should tracks be laid to the nearest English
>pitch inside the pad for example? Any offer of the right way to do this, if
>it matters, is appreciated. I have noticed that the "quality" of the layout
>begins to be degraded or becomes inconsistent as the layout is edited due to
>the different units.
Basically, you could play tic-tac-toe inside the pad, it makes no
difference to the end product. As I mentioned, you do not even need to
complete that last segments.
One suggestion that I have made to Protel but I have no idea if it is going
to be implemented is a track completion command. In DOS Tango, if you were
routing from a connection line, and then, mid-route, hit the slash key, it
completed the connection to its destination as a single any-angle track.
This is a real time-saver.
(I think that the secrecy around some of Protel's plans benefits no one and
hurts everyone except the competition; it is only that everyone is that way
that makes secrecy seem to be desirable. When there is a bug, Protel should
acknowledge a commitment to fix it, if the commitment has been internally
made. When a simple suggestion has been made, Protel should let us know
that it has been accepted and is scheduled for inclusion, either in a
service pack or in the next release. Some major release plans and dates
might deserve to remain secret, but being open about improvement details
would really be a service to us. And if, for some reason, Protel is unable
to meet the commitment, we aren't going to sue! As long as the
announcements are covered by qualifying language that they represent
intention and not a binding promise, there would be no danger of that even
from the most litigious of licensees. Wouldn't it be nice if our bug
database, in the [EMAIL PROTECTED] filespace, had a field that
indicated Protel has acknowledged and is fixing the bug -- ah, "issue," or
is accepting the suggestion and intends to implement it?)
[EMAIL PROTECTED]
Abdulrahman Lomax
P.O. Box 690
El Verano, CA 95433
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
*
* To leave this list visit:
* http://www.techservinc.com/protelusers/subscrib.html
* - or email -
* mailto:[EMAIL PROTECTED]?body=leave%20proteledaforum
*
* Contact the list manager:
* mailto:[EMAIL PROTECTED]
*
* Browse or Search previous postings:
* http://www.mail-archive.com/[email protected]
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
