Augs inverted image woes

[Richard Cullin]

i'm still playing around with making pcbs  , now trying out auggie's image rastering,  the TRACKS when lasered with a positive image are
correct width {although quite useless for pcb etching }  When the image is inverted the TRACKS and PADS all seem to shrink ,
it all looks good on the screen but the result is disappointing when lasered.
looks to me that the laser dot with is being subtracted from the region .
attached are the org pattern and the pos/neg results  , raster was 30mm wide .5mm overscan, @.15mm raster

[Richard Cullin]

I inverted the image elsewhere and retried augs  , I still feel the tracks are cropped .
is there any scope for laser dot size adjustment ?

[ArtF]

Hi Richard:

  Id have thought it would get slightly larger.. due to the laser being considered in auggie to be infinitely small
in diameter. ( I havent tried any engraving such as your doing.. ). I don't think I added any option to make it turn on sooner or shut off later. It IS odd though, I can picture th eon taking a bit too long to occur, thus shrinking the leading edge, but Id have thought the same effect would lengthen the lagging edge, for an equal or wider line width,
not smaller.

  To make a line rastering across thinner would take a delay in turning on perhaps with no off delay. That I could
see thinning such a line out. You may want to simply try going slower to see if it alleviates it any? That would
make logical sense if it is an on delay issue...

Art

[Richard Cullin]

hi art
I don't think its a timing thing , traces are thinned n/s as well as e/w . I did try though ,slowing from 500 to 200mm/m makes little difference.
it looks more like milling a shape with a 4mm bit when all the offsets were calculated for a 3mm bit. 
traces and pads I can be oversized to compensate but it would be nice for etching logo's and such if negative images were true to form

ps trying  dmap2gcode , results so far no better

[Richard Cullin]

ps
I increased the dpi of the source image x 4  thinking it may be a rounding issue .  makes negligible difference

[ArtF]

Ahh, I can see it then, makes a kind of sense. Since it thinks a beam is infinitely small, when it turns on it cuts a
radial distance into anything its cutting. My laser has a very small kerf, so I may not notice it as much, though
I probably would if I tried something like a circuit board trace.

  Other than somehow offsetting the image, Im not sure I can do anything about it in the raster engraving,
it would entail turning on the beam early and off late to make up for the diameter, and that'd be true in
both e/w and n/s directions..thats hard to do on the fly with images, not impossible, just hard.

  Be better to have something that grows the image a bit in all dimensions or something that focuses the
beam a bit tighter..  This does seem to be something Ill face eventually though, as I can see where trying to
do inlays will be affected as the positive and negative images wont match... Ill give this some thought..

Art

[tweakie]

Hi Richard,

Did you manage to solve your track width etching problem ?

Looking at your pictures my thoughts are ( as you have said ?milling a shape with a 4mm bit when all the offsets were calculate d for a 3mm bit? ) that your laser focussed ?dot? size is too large.

Just to see if I could replicate the problem I copied your pic. and generated the code for Auggie and this was the result.

Hope this helps.

Tweakie.

[Richard Cullin]

Did you manage to solve your track width etching problem ?

not really ,i have gone back to gcode generated by the isolation milling pgm for the laser. its a bit slow but i get good results.

my diode laser is fixed focus @18mm focal length and a dot size approx .12mm ,the dot is not perfectly circular either
your laser has done a better job , that would etch quite well.
i'm looking to upgrade to a chinese 50w job  and  will definitely convert it to auggie running gear.