At 02:41 PM 5/14/2002 -0700, Dennis Saputelli wrote:
>could you repost your alternative?

possible implementation: enable an unused mech layer and label it "shorts." 
Add track on this layer to a jumper footprint so that it will short the 
jumper if added to the plot for the copper layer. Use the "shorted" jumper 
on schematics to isolate a net section (i.e., to control connection path). 
Create a separate gerber definition file under the CAM manager for the 
copper layer used which adds the shorts mech layer to the plot; disable 
plot for that copper layer in the regular CAM gerber definition file. If 
both definitions are enabled for plot in the CAM manager, both will be plotted.

(Copy the main CAM def file to make the special one, then disable all 
layers except the one you want to plot and change the mech layer to plot 
simultaneously in on the mech layer tab.)

I have not actually used this procedure, only the microgap virtual short, 
but it seems quite straightforward in theory. If anyone has used it, please 
confirm that it works or tell us how it didn't, if known.

I would expect that the most likely failure would be that the designer 
forgets to set the special CAM definition, which would be relatively 
harmless at the prototype level since the shorting jumper could manually be 
shorted. (This is also the case with the microgap short).

This, by the way, is a good argument for requiring the fabricator to 
provide its customer the actual CAM files used for fabrication. Fabricators 
routinely make small changes without informing customers in order to 
improve manufacturability.... But I've never seen a fabricator provide the 
actual files.

In any case, the fabricator who unshorted the pads should very clearly have 
queried the customer as to whether or not the pads should be shorted (as 
they would if the board had been fabricated as-is) or opened to a 
fabricatable gap. So those missing shorts were due to fabricator error...

>i missed it
>i did find that origin pin 1 seemed to nail the issue and make no gap
>even in gerber 2.4
>my gap was i think .005 mil
>in the (bad) case of using center origin and using gerber 2.3 i think
>result was a 2 mil gap as the edges of the pads were pulled back toward
>the center to the nearest mil
>this notwithstanding it has proven to be a useful and clever tool
>Dennis Saputelli
>Abd ulRahman Lomax wrote:
> >
> > If gerber plots were not rounded off, there would be no problem with the
> > "virtual shorts," and, in fact, if fab houses fabbed the boards as-is
> > without modifying the gerber, there would also be no problem.
> >
> > But Protel does some rounding and it is not easy to exactly control
> > aperture assignments while using the much easier RS-274X, though it can be
> > done; properly implemented, aperture match would cause the gap to actually
> > disappear as long as the pad distance is such as to leave the pads on a 1
> > mil grid. But as Mr. Saputelli discovered, it is fairly easy to set up and
> > place a virtual short footprint in such a way as to leave a tiny gap,
> > enough to puzzle the fab house inspector.
> >
> > Complexities like this have led me to recommend the alternate method I
> > indicated in another post in this thread. It is a little easier to document
> > and no fab house will be tempted to modify the gerber.
> >
>            Integrated Controls, Inc.
>    tel: 415-647-0480                        2851 21st Street
>       fax: 415-647-3003                        San Francisco, CA 94110

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
* To leave this list visit:
* Contact the list manager:
* Forum Guidelines Rules:
* Browse or Search previous postings:
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Reply via email to