6.1.5, DRC err on tStop layer...sometimes

On 4/13/2012 11:31 AM, Robert Scott wrote:

On Thu, 12 Apr 2012 20:51:25 -0700, Bill Martin<wwm@wwmartin.net>

wrote:

 

>> Hi,

>>

>> Most likely my error someplace, brand new to Eagle CAD. Anyway, my

>> question is: Is tStop layer invisible by default? Reason for this

>> question: I was down to zero DRC errors on my first small board layout

>> with Eagle, then I notice part of the perimeter (layer 20) had vanished

>> on my layout. So, I turned off view of all layers, then turned on just

>> layer 20 so I could redraw the border outline. Fine, done with that,

>> turn on ALL layers and ran DRC. Now have about 30 errors re tStop layer.

>> How is it possible to not see this before? Just for fun, I turned off

>> tStop layer and ran DRC again. No errors! What the heck... can anyone

>> shed some light on this, is there something I need to fix? By the way, I

>> did 99% of this layout with original default design rules, then found

>> that SunStone had a .DRU file for Eagle, so I loaded that. Still did not

>> see errors before turning on ALL layers. Should I be concerned about

>> these errors I almost did not notice before ordering boards? I made

>> Gerber files&  looked at them with GerbV, they look fine to my

>> eyeball...but i don't perceive 3mil objects all that well! That seems to

>> be the top solder mask min/max in the SunStone rules.

 

I just posted a similar issue in the support forum and got the same

answer as you got from Joern.  DRC only checks visible layers.  I also

got lots of tStop errors when I temporarily turned on tStop (29)

visibility.  It turned out that all those errors were for the

silkscreen layer overlapping the stop mask.  However even those errors

seemed to be calculated on unrealistic rules.  It seems that a

conflict is detected when any part of the silkscreen text rectangle

intersects with any part of the stop mask, even if no part of the

actual silkscreen text is anywhere near the stop mask.  I suppose it

is just very difficult for Eagle to check the rendering of each letter

in the text to see if it overlaps with the stop mask.  It is much

easier to just check the rectangle that contains the text.  In most

cases you can just approve such errors and go on.  But imagine what

would happen if some silkscreen were placed squarely on top of a

surface-mount pad.  It would make that pad unsolderable.  So we

certainly want an error to be declared in that case.

 

Robert Scott

Ypsilanti, Michigan

On 4/13/2012 11:31 AM, Robert Scott wrote:

On Thu, 12 Apr 2012 20:51:25 -0700, Bill Martin<wwm@wwmartin.net>

wrote:

 

>> Hi,

>>

>> Most likely my error someplace, brand new to Eagle CAD. Anyway, my

>> question is: Is tStop layer invisible by default? Reason for this

>> question: I was down to zero DRC errors on my first small board layout

>> with Eagle, then I notice part of the perimeter (layer 20) had vanished

>> on my layout. So, I turned off view of all layers, then turned on just

>> layer 20 so I could redraw the border outline. Fine, done with that,

>> turn on ALL layers and ran DRC. Now have about 30 errors re tStop layer.

>> How is it possible to not see this before? Just for fun, I turned off

>> tStop layer and ran DRC again. No errors! What the heck... can anyone

>> shed some light on this, is there something I need to fix? By the way, I

>> did 99% of this layout with original default design rules, then found

>> that SunStone had a .DRU file for Eagle, so I loaded that. Still did not

>> see errors before turning on ALL layers. Should I be concerned about

>> these errors I almost did not notice before ordering boards? I made

>> Gerber files&  looked at them with GerbV, they look fine to my

>> eyeball...but i don't perceive 3mil objects all that well! That seems to

>> be the top solder mask min/max in the SunStone rules.

 

I just posted a similar issue in the support forum and got the same

answer as you got from Joern.  DRC only checks visible layers.  I also

got lots of tStop errors when I temporarily turned on tStop (29)

visibility.  It turned out that all those errors were for the

silkscreen layer overlapping the stop mask.  However even those errors

seemed to be calculated on unrealistic rules.  It seems that a

conflict is detected when any part of the silkscreen text rectangle

intersects with any part of the stop mask, even if no part of the

actual silkscreen text is anywhere near the stop mask.  I suppose it

is just very difficult for Eagle to check the rendering of each letter

in the text to see if it overlaps with the stop mask.  It is much

easier to just check the rectangle that contains the text.  In most

cases you can just approve such errors and go on.  But imagine what

would happen if some silkscreen were placed squarely on top of a

surface-mount pad.  It would make that pad unsolderable.  So we

certainly want an error to be declared in that case.

 

Robert Scott

Ypsilanti, Michigan