I was hoping someone with Lightburn could answer us about this product. It’s a flying galvo head or a galvo on a gantry machine. It’s supposed to go to Kickstarter this month.
Is it really Lightburn compatible ?.. claiming they have been in touch with Lightburn during the development, any comments? Wonder how some of this is implemented, such as LIDAR Z control, and assume gcode?
Here is a quick video on two machines.. Originally posted by @HalfNormal on the Markers forum.
Here’s what I will say:
They actually are in contact with us. Probably the most consistent contact we have had from a company to support hardware like this. We have not touched hardware yet or seen anything in person.
We are working with them to get started on supporting their hardware. Their team is committed to making sure we have the support we need to support their machine well, but no real work has started yet, as our team is 100 % committed to 2.0 and MillMage launches.
Thanks for the update. It’s nice to hear they’re talking to you..
Any idea of what’s inside?
There are a few videos they have released and their KickStarter has recently launched. More info there and to come.
Already over at the site and watched a few video. Interesting machine. The theory has been bantered around since galvo started becoming popular.
I’d be nice to know basic functionality, such is the fiber using a separate controller than the gantry?.. Just curious.
I put a lot of thought into how to do this, and why.
It’s complicated. It took awhile for me to formulate an answer as to why a galvo laser “doesn’t cut anything but thin materials”. You’d think that, given time, it would just keep cutting further, especially if you keep lowering the Z. It doesn’t do that.
I’m used to CO2 lasers, which have an unusual, very favorable, and not widely understood “magic” property of 10.6um: it is reflected off the walls of a cut when coming in at a low angle of incidence. Once a cut channel starts to form, if successful it forms a small funnel and the beam is “captured” and reflects off the walls until it gets to the bottom. This is why you can get tall vertical, parallel walls with a CO2 laser when the beam is generally diverged and “out of focus” about 2.5mm below the ideal focus and can’t even start a cut on the same material there, but can continue to cut through if it already has a channel starting at the focal point that channels the beam downward.
Fiber laser’s ~1.6um does not do that. It cuts at the focal point, cuts about 0.5mm down at most, but the beam can’t reflect off the walls, it just continues diverging within the cavity it has already cut and won’t be directed into a channel.
But the problem specific to galvos, including ones with a 10.6um CO2 source, is, well, I call it “occlusion”. It’s that the beam is always vectoring in at an angle from the lens. It’s got significant beam diameter up there, too. I know it has a 10mm aperture for the beam coming in, it spreads wider as it passes through the lens glass. I’m just going to think of it as 10mm dia until I can find a real number.
The common 110x110 lens is F160. The mirrors are something like 50 mm above that. The glass does shift the geometry a bit from refraction, but I’m just going to consider the beam as vectoring from a centerpoint 210mm above the work 10mm diameter.
If you try to drill a small dia hole, say 0.5mm, 50mm from the center, the beam is not coming straight down, it’s 13deg from vertical. It will make a 0.5mm hole at the surface, about 0.5mm deep max. And the walls of the hole will actually be slanted at 13 deg.
Lowering the focus by 0.5mm at this point won’t work, not for a small off-center hole. Because lowering the head shifts the hole’s center inward where it crosses the work’s surface before reaching the focal point 0.5mm below the surface. It shifts by 0.119mm. It doesn’t matter if you’re rastering or doing an offset cut or whatever- part of the beam pattern clips on the surface and is lost as reflection or diffuse heat, it won’t cut the surface. Part of the scan pattern will have the focal cone partially clip on the surface. But part of the scan pattern- that intended for the half of the circle furthest from the center of the bed- will make it thorough with no clipping and cut deeper, but at a slant, and you can’t go much further- once you’ve lowered the focus by 2.1mm, none of that 0.5mm dia scan pattern will even pass through the original 0.5mm hole at the surface.
In theory, if you cut, lowered the Z by 0.5mm, and ALSO shifted the circle’s center outward by 0.119mm in the design, it would mostly go through the hole. But it deepens at at 13 deg slant. You could do it several times, but the size of the instantaneous beam where it crosses the surface trying to reach the bottom of the hole gets larger as you shift the focal cone downwards. The size of the cut’s bottom the entire beam can reach without any of it clipping on the surface keeps shrinking.
At first it seemed like the center of the table was an exception- but it’s only an exception to the first part. The centerline of the hole isn’t shifting as we drop the Z. However, the focal cone is still a cone about 10mm at 210mm from the surface. So, like, when I tried to drill a tiny 0.1 dia hole, part of the beam’s scan will clip as we go lower, shrinking the dia of the bottom that the entire beam can reach. Below 2mm deep, even the static moment where the beam is firing straight down X=55 Y=55, the beam’s focal cone is over 0.1mm in dia as it crosses the surface so there’s no point of the entire scan pattern where the whole beam will fit in the hole.
In theory, we could have a straight-walled hole of unlimited depth IF the hole in the center and at least 10mm in dia, and you don’t just cut a circle, you have to do a Fill that removes all the material making a path the beam can fit through.
That’s the best case, in the center. A 10mm dia hole 50mm from the center- even if the software (Lightburn) had control to lower the Z a bit each pass AND did something crazy smart and shifted the beam’s cut outward as we lower the Z, it still ends up slanted and the hole the entire beam can reach will shrink as it gets deeper.
So, the reason I’m stepping through this is, what would a flying galvo actually be capable of doing? OK, say we want to cut out vertical walls in 2mm thick material. The cut will have to be 0.1 mm at the surface. The flying head must be flying for the entire length of the cut, and only galvo’ing back and forth by 0.1mm. Weird.
That is, let’s say I want to cut out a 50 mm dia circle in 2mm thick material, and this is a hole- we’re saving the rest of the sheet and throwing away the hole we cut. So we can kerf 0.1mm inward. The beam would have to have the flying head physically move in a circle 49.95mm in dia, while doing a “wobble” that is perpendicular to the line’s tangent, and it has to start at the design’s line, cut inward 0.1mm, then come back, then do another inward jump but not as far, then another pass less still…
That seems like an ideal… just doing the circular “wobble” would kind of work, but be really inefficient. Actually that would be a pretty poor implementation- Lightburn devs, if you’re listening, I’m already 90% sure that’s what is being discussed, and it’s NOT the right solution.
The circular wobble scan isn’t even right for the current fixed galvo. You need to be able to successfully implement a… I guess we call it a “mapped gradient wobble” that works on our fixed galvos first before you can get real value out of a flying head galvo for depth cutting.
Awhile back, when people started to ask to retask the rotary channel to drive a Z-stepper lower for each pass, I realized this has value but it’s much more limited value than it seems, unless this more complicated cutting algorithm is implemented alongside it.
Yeah, I’m making up terms here. I’ll need some feedback here- does this follow anything you’ve worked out, under any term?
Basically, the top surface kerf must get wider the further you are from the center, and the width and centerline of the cut have to shift as Z lowers, but it must shift differently based on how far it is from the center.
I can draw up some diagrams if that helps.
This is something that I would appreciate as a birthday gift. 
I am very happy if Lightburn team is actually in contact with them.
In fact, it has crossed my mind to attach a fiber cable to the head of a laser diode to remove all the weight of the module that adds inertia to the movement, forcing a reduction in speed so that the motors do not lose steps.
But the idea was stagnant as soon as the question arose of how to connect (perfectly centered and focused) the fiber to the laser diode module.
I did not see where they said they would cut with a galvo. But I do see the advantage of having a galvo on a flying bridge.