Thanks for all the responses on footprints. This whole issue is pretty sickening actually. Since we produce low quantities of diverse products we have no dedicated PCB layout people. All engineers do their own circuit designs and parts specification and ultimately are expected to do tiny PCB layouts of everything and get them to work. The thing that gets me is that it seems like it would be extremely simple for parts vendors to provide land patterns for their parts along with the mechanical drawings of the parts themselves. Some do but most don't. I just talked to Maxim about this and they said they simply don't provide this information. They recommended IPCSM782. Of course a good percentage of the parts you need are not listed in this document and a lot of them that are there do not match the recommendations of the vendors of the parts. I asked Maxim how they layout their own eval boards since they provide no guidelines and no guidelines exist in IPCSM782. They didn't have an answer but I suspect they rely on rules of thumb and intuition, which is what we end up doing with our designs here most of the time. After enough bad yields and scolding from our PCB fabricators we manage to stumble into something that seems to work. I did find what I thought was a good layout for 0402, 0603, etc. from AVX capacitors. Upon closer inspection, however, I found that their recommended footprints violated their own guidelines given on a different page of the same document. Go figure!

Ray Mitchell

At 04:59 PM 3/11/2004 +0000, you wrote:
> -----Original Message-----
> From: Ray Mitchell [mailto:[EMAIL PROTECTED]
> Sent: Wednesday, March 10, 2004 5:36 PM
> Subject: [PEDA] Common PCB footprint specifications
> Hello,
> I'm sure this is a repeat, but is there a simple
> specification readily available that gives the "commonly
> accepted" (if there is such a thing) dimensions for 0402,
> 0603, ..., SIOC-14, etc., and all the other "standard"
> footprints?  I don't really want to wade through a bunch of
> technical stuff to derive all of this myself and I certainly
> don't want to trust a priori the patterns that come with
> Protel or any other product.  It's really annoying when part
> manufacturers don't provide these footprints, assuming they
> are common knowledge.


I have accumulated quite a library of such footprints but most of them
will have been optimised to suit our in house processes more than
following the IPC standards.

The supplied Protel IPC land patterns are not too bad, they are
certainly a good basis to build your own on. But most libraries stop at
the land pattern stage, which is what the IPC are looking to change.

A lot of the way the IPC are trying to structure library conventions are
along the lines of what I was already doing for years anyway, not
because it is good, but because it make life easier for us internally if
the naming conventions for footprints already match vision library
footprints on placement machines (which then relates to mechanical
dimensions as well, as a Murata 16V X7R 0603 will have different
dimensions to a Kemet 16V X7R 0603 in same voltage) and other EDA
packages we use etc.

I especially like the way the new IPC recommendations take account of
things like, 0 deg positions in tape or tray, if Protel could also make
allowances for rotation on non-polarised chip parts (only use 0,90) to
reduce un-necessary head rotations on turret head placers that would be
even better as I currently use an in house utility to parse the P&P
files and check for string matches on footprint & part number to
identify non-polarised parts and it will replace 360 or 180 values with
0 and 270 values with 90.

In DXP I planned to use a parameter for that at SCH level, so I only
need to check for one match, but that's another story, no time for
documenting or agreeing how this should be done internally yet.

Same with pad sizes, I slightly oversize SMT pads in some cases against
IPC recommendations (not much) to allow for place tolerances when
reducing Z height & down pressure, Vac release and place speed,
especially on Chip r/c's as well as wave flow direction and so on.
Same for connector placement, especially for IDC and connector rows
<2.54mm, I sometimes enlarge the pads beyond IPC recommendations in one
direction to get the best out of the features on our wave soldering
equipment (Vitronics-Soltec with Select-X debridging).

If Protel could assign a different footprint for rotation, or side,
based on some sort of logical system, then it would make life so much
easier to define DFM rules even at SCH level.
Perhaps that's worth a new feature request on the DXP forum :)

To me a library has to be more than just a symbols collection, or the
manual pre-processing required diminishes its value, very little third
party libraries do this, so IMO are not worth it.

I like the IPC new offerings for library recommendations very much, and
would like to see it adopted, even although some of the naming
recommendations may choke some placement machines offline programming
software or optimisers software a bit like white space, characters, case
sensitivity and a lot of other things that should be non-issues in this
day/age. I prefer a direct import approach to programming these
machines, Gerber import, pattern search and processing is alright, but
takes to long and can be error prone.

If anyone wants me to split & upload the library contents I have here,
ill do it as a part time job, but I guess most people will have these
things already, or prefer to use their own in-house libraries.


Ray Mitchell
Engineer, Code 2732
SPAWAR Systems Center
San Diego, CA. 92152

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
* To leave this list visit:
* http://www.techservinc.com/protelusers/leave.html
* Contact the list manager:
* Forum Guidelines Rules:
* http://www.techservinc.com/protelusers/forumrules.html
* Browse or Search previous postings:
* http://www.mail-archive.com/[EMAIL PROTECTED]
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Reply via email to