# Re: [PEDA] Best settings for polygon fill on signal layer?

```At 09:37 AM 9/30/2002 -0700, Brad Velander wrote:
>Tim,
>         first my observations. I have always found that 45 degree polygons
>are a lot slower than 90 degree polygons at repouring.```
```
Not surprising, since the math is quite a bit more complex....

>  Since yours is
>completely filled I don't see where 45 degree is doing anything for you.

Rarely would it be useful. If one were filling a more or less rectangular
space, it *might* be more efficient in terms of numbers of tracks created.
It might have been better for the programmers to have put more effort into
pour optimization and accuracy -- there are still a few bugs or
deficiencies -- than into 45 degree pours. If someone really needed a 45
degree pour, a rare case, they could rotate the whole board by 45 degrees
and then place a regular orthagonal pour.

>Second, rather than using a grid matching the track width, I normally use a
>"0" grid which spaces the tracks according to the track width. Just a couple
>of tips from my experiences.

Took the words right out of my mouth. I have never seen a reason to use any
other grid setting than 0 -- which is interpreted as "fill efficiently",
unless I want a gridded fill, in which case, of course, grid will be set
larger than track width.

Fill track size is controllable through rules. Obviously, the larger the
track size, the smaller the pour size in terms of numbers of primitives.
But too large, and one will not get, for example, track between the
narrower gaps.

>         As for your other questions, I have never observed Protel to do any
>gerber optimization for polygons.

Protel does not really support true polygon primitives. If it did, it could
do polygons much more efficiently. (It would use positive/negative merges,
easy in RS-274X, to create true polygons with blowouts.) Instead, the
"polygons" are a set of instructions, so to speak, for creating a pile of
track (and/or arc) primitives which together create a pour. Because these
primitives can be DRC'd just like any other track, these "polygons" are
easier to DRC than would a true polygon primitive. Plotting this entity is
also easy for the same reason, polygon primitives are plotted just like any
other tracks and arc. But *working* with these pseudo-polygons is
substantially more difficult, because of slowness of redraw, the sheer
number of primitives which can be involved, etc.

I don't think DXP has introduced true polygons. (They would be very easy to
display!) I'd recommend them for future development.

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* 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/proteledaforum@techservinc.com
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
```