Ok, I've been looking around a bit lately and I'm thinking there's got to be a laser that can be sourced from somewhere to use for laser cutting and engraving with the Rogr/micRo. The gantry is there... It'd only need the X & Y to be in motion. So, what could be so hard about rigging up a sourced laser to laser cut acrylic, delrin, wood? Or etch aluminum, glass, etc?

Seems like it'd be something to look into. For some reason I don't think a DVD laser is going to be able to be pumped enough to cut much more than stryrofoam. Any ideas? Some materials just aren't going to lend them selves all that well to routing.

i am also very much interested in a laser etching/cutting option for the micRo.

Check this out...

http://www.youtube.com/watch?v=gHjGfwf3Aeg

...Now that's just cool.

Or this? Laser welding...

http://www.youtube.com/watch?v=o-kR0c9aFCQ


....I'll add more links after I get home =]

i also want to be able to add a laser to my RoGR. i've got plans for a Router RoGR, a Plasma Cutter RoGR and a Laser RoGR, possible Laser Sintering?

If anyone has knowledge of this please share.. i have none so im starting from scratch.

Ever since I saw this instructable I have been thinking about this possibility: http://www.instructables.com/id/New_007_La...eapon_Revealed/
I wonder if this would be powerful enough to engrave anodizing on aluminum.
Also this baccus61 has a lot of interesting cnc options http://www.youtube.com/watch?v=wi4dKAseDCg...feature=channel

In a word, No.

Lower wattage CO2 lasers are available (which could etch most materials and cut thin wood and plastic), but the really good stuff is pretty hard to get. For the record, aluminum is difficult to etch/cut with CO2 lasers since it's around 90% reflective to infra-red, and requires special coatings to effectively be etched. Though, it is my understanding that anodized aluminum is easier..

A laser is in development for the RoGR; I am aiding Mr. Will OBrien (formerly wrote for Hackaday) for a book deal he has developing high-power laser modules for smaller scale bots like RoGR and micRo.

Most excellent. I was starting to lean towards the realm of this idea might be beyond the scope of DIY.

Making a laser engraver would be a lot of fun! Seeing it melt that electrical tape makes it seem promising... I wonder what would happen with aluminum.

Cutting acrylic would also be pretty darn cool... No more worries when cutting your fresnels.

Looking a little more, "they" say the blue laser from the Xbox360 might be able to do 200 mW, if you search for laser engraver on eBay it looks like the "real" ones are 40 W and higher (200 times more watts).

Implementing laser marking properly (consistent, uniform marks, quickly) is far more difficult that it might seem. It might be possible to do with EMC, as long as the path vectors are preprocessed properly, but it will not be easy. The fundamental problem is that you need to ensure that amount of energy delivered by the laser is uniform across each line. Sounds simple, right? Not in the slightest.

The difficulty arises because the gantry (or more commonly, galvonometers with mirrors attached) has mass, and cannot be instantaneously switched from one velocity to another; it has to be accelerated/decelerated. This means that it will start off slowly, with its velocity changing incrementally until it reaches the desired velocity. Imagine the laser spot on the material to be marked/etched. The laser will deliver a constant amount of heat into the material under the spot. When the spot is moving slowly, during acceleration or deceleration, more heat is put into the material, usually causing deeper/wider/melt marks at the beginning and end of an etched line.

The solution to this is, in retrospect, obvious, but very few people ever figure it out: The laser can only be on (marking/etching) when the spot is moving at a constant velocity. To achieve this, you need to get a "running start", that is, you begin your acceleration with the laser off/shuttered, well before the spot reaches the beginning of the line it is to mark. When the spot is moving at constant velocity, turn on/unshutter the laser when it reaches the beginning of the line to be marked/etched. As you near the end of the line, most people's instinct is to decelerate, to get ready for the next vector. Don't!! The spot must be moving at a constant velocity across the entire line. When the end of the line is reached, then turn off/shutter the laser.

At this point, the spot has overshot the beginning of the next vector. There are some rather sophisticated tricks on how to most efficiently control the spot path - at the fastest possible speed - to line up for the next vector. Also if you want really high quality marks that are uniform across the entire line and you're using a pulsed laser, you'll want to implement first pulse suppression, because the first pulse usually has more energy than subsequent pulses, causing deeper/wider marks at the beginning of a line. We spent years figuring out how to create fast, high-quality laser marking systems using an infrared Q-switched YAG laser (http://en.wikipedia.org/wiki/Nd-YAG_laser), which will mark on just about anything.

If anyone is interested, I'll explain more in another post.

-Brian

First off, impressive. Having that insight you just shared is what'd make this all possible. That movement/laser shuttering problem could all be solved in software. Does EMC support it? I doubt it. But the EMC libraries can be used by additional programs. A "Laser G-code" tool could be made fairly easily, given some well thought out algorithms.

The electrical/hardware aspects of the laser are what I still see as being the big hurdle. That pulse problem, off the top of my head, I wouldn't know how to solve for etching, but for cutting that could be another matter of g-code. Start the cut in the center waste if it's a pocket. The outer waste if it's an outline. Just like using a scroll saw.

Given your experience, what would be your take on the actually laser piece? Getting so far as having one in hand, should make the other issues apparent and then solutions could be developed.

If it's going to be cost prohibitive, that's one thing. If it's a matter of solutions not being developed yet, then that just sounds right up our alley.

When making the cam or writing the g-code, you can simply envoke "constant velocity mode". It is highly supported by EMC2. The gantry does have mass, but the motors have a lot of power. In constant velocity mode, the reversal are instant...impressive to watch and makes a nice WHOOMP sound as the moment/energy is transferred to the floor via the frame members.

I've been out of the industry for about 20 years. After a quick check of the internet, it appears that most Q-switch drivers now implement First Pulse Suppression, so that won't be a problem. We had to figure it out as we invented all this technology 25 years ago.


That depends entirely what you want to mark/cut. We used an 80-100 watt Q-switched Nd:YAG laser, which could mark on just about anything (even weld disk head flextures!). However, that is a large investment, and after a quick scan of the internet, will probably be cost prohibitive for most people. To buy the components new looks like it will cost between $10-20 thousand. You might be abled to find used components on eBay for a lot less. Also, realize that this equipment can be very dangerous!


As I wrote above, I've been out of the industry for 20 years, so I don't know what is available in terms of low-power lasers. To size the laser, the fundamental questions that everyone has to answer, are "What types of materials will I be marking?", "How deep do I need to mark?", "How fast do I need to mark?", and "What is my budget for a laser system?" The more powerful the laser, the higher the cost.

-Brian

It is far from simple. Constant velocity is great once the gantry is up to speed, but it still has to accelerate to get to that point. For high-quality marks, the laser can only be enabled when the gantry is moving at a constant velocity; otherwise the energy delivered along the length of the line will not be uniform.

Consider the case where two lines are perpendicular to each other, and orthogonal to the gantry axes, in other words, an "L". The lines will be marked starting at the top left corner, and moving in decreasing X coordinates toward the lower-left corner of the "L". The gantry must be moving at a constant velocity before it reaches the starting point; therefore, the gantry has to move to a point well before the starting point to be marked, so it can be accelerated to a constant velocity. When the gantry reaches the corner of the "L", the laser is disabled.

If the gantry were mass-less then we could immediately stop its movement in the X direction, instantaneously set the speed along the Y axis, and then draw the bottom of the "L" in increasing Y coordinates. Unfortunately, the gantry is not mass-less, so we need to decelerate it, which will take some time and distance, which means that it will no longer be lined up with where it needs to be in order to mark the bottom of the "L". Also, any violent change in velocity will cause the structure to "ring", and ruin the nice straight lines we're trying to mark. So, the gantry will have to be repositioned before we start marking again. We will also need to back the gantry off in the decreasing Y direction to allow it to accelerate up to a constant velocity before enabling the laser when the gantry reaches the lower-left corner of the "L".

If the coordinates of the corner of the "L" is (0,0), then the resultant path the gantry will slew between marking the two lines is a loop through the third quadrant. This is the secret to high-speed high-quality marking. I created a drawing that would greatly simplify this description, but unfortunately I don't know how to insert an image into this post without putting the drawing on some web site and linking to it.

-Brian

For a DIY application, wouldn't simply adding a radius to the corner solve the CV issue? If it's not 100% solved, it's probably going to be good enough...

To insert an image go to "Full" reply window (hit the "More Options" button in the Fast Reply pane).

In the bottom-ish right-ish there's a field/button to choose your file location. Do whatever you gotta do to get your image path in there. Then hit the upload button.
I'm using Google Chrome, here's what it looks like on mine...
Click to view attachment

After the file is uploaded, you can insert it inline using this drop box...
Click to view attachment
Click on the little icon that looks like a piece of paper with a green "plus" symbol.

All your attachments combined must be less than 110K.

Your robot sounds bad.

That is why I prefaced my remarks with the qualification: "for high-speed, high-quality marks". If that isn't important to you, then you don't need to jump through all the hoops, although once you see high-quality marking you'll be really disappointed with poor or medium quality marks.



Thanks!

Here's the magic path diagram. The dotted lines are where the laser is off.
Click to view attachment


We invented most of the technology used for commercial laser marking systems. These systems generally use mirrors attached to galvonometers to steer the beam. We were the world leaders in this area, until a sequence of 3 or 4 unqualified CEOs over a period of 3 years eventually killed the company.
The basic principles that I described apply equally to a gantry-driven robot as they to do a galvonometer-driven mirror system.

-Brian