GRBL Laser PP+PPT

[dave benson]

Pics to follow.

[dave benson]

more...

[dave benson]

more...

[dave benson]

last one.
Couldn't get it to load to the forum, heres a med res one instead.
Dave

[dave benson]

Here's an update to the GRBL\Mach3 PPT and PP for Lasers.

An explanation of the new command line:
 -.255 0 0 255 XA

The first two values are the max and min depth values that you used in the Height Map generator
to generate the file.
The next two  values are the scaling values for the laser power.
and lastly the next two letters are laser head designations.

The default for this line is  -.255 0 0 255 X
where X designates laser head one (most people will only have one).

The next option for this line is  -.255 0 0 255 A
where A designates laser head Two.

The last option for this line is  -.255 0 0 255 XA
where XA designates use Both laser heads.

To be clear  this won't work  -.255 0 0 255 X  A   don't leave a space between the letters.

The PPT also does vectors now.
say for example you want to raster an image and the put some text around the image.

Then you would generate your raster line with the height map generator  and select a engrave mop
and use your laser style if you have one set up. this must be the first part and mop.

Then to do vectors make a new part and engrave mop (some where in the mop name include the word vector set your depth -0.127  for example is half power.

Always put the vectors last in the file after all Rastering is done.

For Mach3 users you'll have to then open up the file and change the one and only M4 in to M3 in the vector mop.

Dave

[dave benson]

Theoretically you could have 4 laser heads running simultaneously with the hardware I have.
As with 3D printing (some prints can take days) Lasing is a slow process
for Rastering images, vectors can be pretty fast though.
So to speed things up a bit I added the ability to use a second laser head
which can be used by itself or with the other laser head.

At this point I have two lasers a small 2 watt one and a 15 watt.
and the intention is to use the smaller one for fun, and the larger one
for cutting out gaskets.

I have run the X and the A Axis slaved together, which works well to somewhere
around 3-4  m\m  then the combined mass of the laser heads (both going the same direction)
starts to cause the end plates to flex (6mm plexiglass)  so I ran the heads in opposition to each
other, so as to balance the forces on the gantry end plates, this worked better than expected
as the laser heads have different masses, which is mostly down to the motor for the 15 watt head
as it's spare from the grinder project, and is the most power full in that frame size however it is
three times the weight or the other motor. the only thing with this technique is the second image is
a mirror Image.   I could rewrite the PPT to output two identical images but this is not practical
as I'm running out of memory (out of memory exceptions in CB) and have had to  keep
the images to a reasonable size.


I tried to keep the weight down on the second laser head for this reason, making it out of
5 series al and removing most of the metal with either profiling or drilling.
It is still heavier that the original head, but not by a lot (sans the motors of course) and is many times more rigid because all the al spacer hardware was machined on my cnc lathe the travelling wheels on the carriage aligned much better the the original plastic ones. over all it feels smoother to move by hand and more rigidly attached to the rail and even though it's heavier, it's running 1.5 x greater Accel settings than the original laser head.
The mechanical part of the build went well, the only snag was that the original laser was switched 12 volt TTL and the larger laser was 5 volt TTL , so I had to make 12 to 5 volt converter and with the board I'm using I had to use an opto-coupler and an 4049 inverting buffer to separate the power and signal ground.


Dave

[Bubba]

Wow Dave, that is some setup.  . It looks like you are on other level with this stuff, and I thought you will be using it for gaskets only.. Nice work. Stay safe. 

[dave benson]

Yeah the gaskets are the main thing that I need in the shop and are
difficult to make when they are  small and or have intricate shapes.
The original idea was to get the laser kit from ebay and use the 15 watt
laser and just toss the other laser in the parts bin, as although installing
it in the mill was pretty easy (it fit in the spindle) there were times when
I wanted to use the mill at the same time and some Images done in the
mill took 10 hours or more, and seeing that video (of the tiles) pushed me
over the edge and I bought the BenBox powered laser.
I had every intention of removing the controller and laser and substituting
the big one.

Fortunately I tried a couple of tiles out on the mill while the laser  was coming
and discovered that the large laser (for doing painted tiles) gave very  little
resolution of the image about 40 percent out of a theoretical 100 percent.
The 15 watt laser for doing unpainted tiles had a resolution of 70 percent.

So depending on the type of method you choose to do your tile, you need to
select either the small laser (good for painted tiles) or the large laser (good for
etching into the tiles surface.

The Gandalf tile came out well etching straight into the tile,and a  clear top coat later
Painted tiles can get you many colours and shades than the above method, however
the drying times (for enamel paint) is at least a week between coats, if you don't wait long enough
before you come to seal them the paint can smudge.
I eventually moved to --> primer -->colour coat-->clear coat-->colour coat--> clear coat ect.

This gave the most clear and defined images and didn't smudge or run the colours
together at all when the final top coat was applied.

The next time I'm at the big box store, I'm going to look at water based paints as I think
the drying times will be much better.

The thing that really interests me is Imaging marble, as I've had the best results
I bought a dozen hex shaped marble pieces and made a few coasters out of them by
glueing some cork on the bottom and I've given all away, except for the three
I completely destroyed getting the laser set up.

Overall the critical part of the process is the set up, you must do to a test tile.
That test tile must be a representative average of all of the tiles.

I've found that, what I thought passed passed for a pretty even coat on a batch of tiles
was far from it.

Eventually I came up with a method (to hold the paint can) and a stance very similar to
David Carridine's character in the TV series Kung Fu when doing the grasshopper stance.

I will play with the Imaging from time to time, but I've been doing this while waiting for some
bits and pieces for the grinder and must get back to that as I have eight turrets finished
to a stage where I would normally machine on the mill, a ratchet wheel and what I want to do
is rough them out on the mill and finish them on the grinder.
 
Dave

[dave benson]

Here's a little video of the twin laser heads working in opposition mode
4 meters\minute,this was just a part of the file being lased where I was tuning
for Accel settings.  The controller can go faster, but there's no point as the lasers
cannot mark even paper at the speed. I was getting good clean text between 2.5 and
3 m/min with the large laser, the smaller laser did better than expected  getting reasonable
text  between 1.4 and 1.6 m/mim.
At the start of the video the heads are ramping down a little, in preparation for a short
machine gun like burst to see how much vibration was carried through to the gantry.
Opps it's 7.5 meg I'll have to upload it to utube and post a link
https://www.youtube.com/watch?v=6Jts0zLzroQ

Dave

[Bubba]

Yep, you can laser raster image without spending a week at doing so. On my machine, it would be a painful process.

Nice setup.

[dave benson]

For those that down loaded the pp+PPT ,I've given you future debug version of the PP
so won't work with the vectors properly you have to set the Arc output to "output as lines"
as it set in the debug version as "Normal"

One thing I didn't say was that you don't have to convert the text to polylines either are fine.

It's always going to be slow lasing with a machine that moves the laser around, machines
built like ours are limited by having to physically have to move the laser head around.

Machines like Ed's K40 are a step up because (although they are using the same transport system)
they only have to move the mirrors and housing around. Lighter equals faster.
Galvo systems only have to move mirrors around, so can be faster, but do struggle with
certain types of Images so they are not a cure all either. Art speaks of this in the video.

Here's a video of Art (of Mach3 Fame) making a fairly cheap system up using galvo's.
https://www.youtube.com/watch?v=zKBA3esQ2jg
There's some good info in there.

I did have poke around his Gearotic site and there's a motion controller designed (free download) to go with the pokeys board that may be worth a look.

Dave

[dave benson]

Bubba, I got to a point a while ago where the pathing had no more optimisations to make
and the laser engraver now can out run the lasers ability to etch even paper, so the only optimisation
left in the process was to see if a lower resolution input Image could produce an acceptable Image
as compared to a hi res one. so I did some tests, where in steps I reduced the resolution to a
point where I could see the Image quality change.

I did a batch of tiles of this space ship in the photo, one takes 1 hr 24 min on the mill and 22 min
on the laser engraver for hi res and 13 minutes for the lo res.

I adjusted down the resolution until, having the tile sat on the table top and from a standing position
looking down at the tiles couldn't see any differences between the tiles.

So for you this may be a way forward as I've found that as little as 30 DPI can work very well.
I use Inkscape mostly, gimp occasionally and I export the Image to the correct size from Inkscape
for example the space ship's dimensions are 70 x 40 mm when I export the Image I set the height to
700 and the width to 400 this is because with my laser's Y step resolution is 0.1 mm
that's how wide the beam is. I do something a different here as I also take into account the
Aspect Ratio of the laser.

It's always advisable to check with the measuring tool what Y step over you actually got as sometimes it can be higher or lower than what you set. for example say I want 0.1 mm step-over in Y and when measured is actually 0.08 then I go back to the height-map generator and adjust the Image size until it's 0.1 .

Dave

[dave benson]

Here's the lo res tile.

Dave

[Bubba]

Here's the lo res tile.

Dave

Dave, that's impressive.