The boards are fundamentally different based on different designs.
With the Nano V3 you’d have to coax a 3D printer board to work for a laser. So likely no GRBL support and certainly not from factory. I saw that there was some work done in porting grblHAL but not certain about the state of port. This means you’d likely be stuck with a Marlin based firmware which is not ideal if you’re building a for purpose system, especially if you want to limit tinkering.
Is there something specific with the Nano V3 that’s compelling?
Nothing specific, just making sure I’m not missing anything as I’m not that familiar with the boards and their minute details.
Thanks for the explanation.
I got some recommendations and narrowed down my choices to two boards (which is what I was looking for) - MKS DLC32 and Ruida 6445. And I’ve got a month or so to make my decision.
Unless there’s some other board that comes highly recommended that I haven’t seen yet…
Is there a good forum for DIY laser machines? Want to look at discussions of stepper motor choices, etc. As I’ll be building a significantly larger than normal machine, I want to make sure my hardware choices are up to par but at the same time not over built
These are very different classes of solutions. Using a Ruida for a dedicated diode laser is not playing to its strengths as they’re really designed for RF and electrically excited CO2 lasers. Additionally it could be argued that they’re sub-optimal for the purpose of driving a diode laser although adequate. For example, Ruida would require external stepper drivers. This might be moot if you have high-end demands, however.
How big are we talking? And what material are you using for the frame? And what kind of drive mechanism are you planning?
I’ve seen many but not aware of a particularly good site for this.
I’m building this on top of a Quick CNC Swift 1313 - 1300mm x 1300mm work area (+ extra Y travel for ATC system). So overall I’m expecting to use something like 1400-1500mm long 2060 extrusion for X rail and 1600-2000mm long 2060 extrusion for Y rails. I’m not 100% sure on sizes yet as I don’t have the machine and haven’t planned out everything yet. I might also step down to 2040 extrusion if I can find a decent priced custom producer.
For drive mechanism - I’m looking at belt drive for X and Y. X motor probably attached to the laser plate, and Y motor(s?) at the back or front and linked via a metal rod to both Y belts. NEMA 17 for X stepper and probably 23 for Y stepper.
I like the idea of keeping the moving gantry as light as possible (which will be hard due to the size and use of a “80W” diode), and since I’ll probably end up going with NEMA 23 for the Y axis, the stepper will have to be stationary.
I’m 99% set on closed loop motors (unless the controller board doesn’t support the alarm function).
Okay. I don’t see anything too exotic in what you’re doing. The belt runs will be quite long, however. Don’t have experience with anything quite that long. I suspect X-axis will be fine. I’m curious how Y axis will behave.
The mass of X-axis will be inconsequential. Gantry mass could be significant especially with 2060s but a typical NEMA 23 should be hefty enough to address. I’d be concerned about the left and right sides of the Y-axis staying aligned, especially at those lengths.
I don’t think standard GRBL has any provision for closed loop motors. There’s nothing keeping you from using a hybrid driver as it should manage this internally.
I’ll say in general there’s relatively little value in going to a closed loop model. You’re dealing with relatively low mass and no resistance from the material that you’re burning. So this is very different from CNC operation where material impedance is practically expected. With lasers, if you’re missing steps there’s something fundamentally wrong with your setup or you’re just driving too fast. You should never lose steps in the normal course of operation.
I had my previous spindle CNC running on closed loop steppers and was able to get much faster speeds before stalling when running the same motors in closed loop vs open loop. I think it has to go with the way that the drivers deliver power to the motors but not 100% sure. All I know is I could easily stop the spindle carriage at around 1000 mm/min in open loop and could barely stop it at 2500 mm/min in open loop. (Same motors, same drivers, just configuration in the drivers)
And I don’t think the alarm would have anything to do with GRBL - just want it to stop the controller. But I guess I could always wire something up with the emergency stop.
What kind of issues do you see happening with having a roughly 3.5-4m belt?
I’d like to be and to run quite fast (at max 8000 mm/min for MKS DLC32 or faster if it’s a different controller)
Did you measure if you were actually reaching these speeds? Trying to get my head around how/why a closed loop motor would be able to go faster. I could see how perceived torque would be higher but surprised that actual top speed would be higher.
You may be able to wire the driver to a pin on the controller that accepts a halt condition when steps are lost. Even Z endstop could potentially work.
My thoughts are around belt stretch and maintaining proper tension. Concerns may be unfounded.
The biggest factor in this will be mass of the laser head. You can pretty much rule out any kind of motorized Z-axis mounted on the head at least. Drag chains would be a hindrance as well. But there are plenty of commodity machines today that claim to run at those speeds at least on the X-axis so see this as perfectly possible.
You definitely want a 32-bit solution or better, though. 8-bit solutions like Arduino Uno based solutions will be speed limited.
Just in terms of max speed… it’s a tradeoff for sure as you need to come up with a different way of managing wiring. In any case you’ll need a fairly rigid setup to get you clean burns at really high speeds. But you need to think like a sports car engineer… much easier to strip out weight than add power.
I’m not familiar with the variety of belt types so not sure about this. I’ve dealt mostly with commodity belts at much smaller scale.
