What to take another look at poll? What should I work on

Here are the two key printed parts I designed and/or remixed for my wall mounted panel saw:



The poll has been added to the original post at the top of the thread…

I found this listing…


SKR 1.2 pro

Try adding it to the pins file, as I_MIN_PIN, (I am pretty sure it works even if you add it to the config page), https://github.com/MarlinFirmware/Marlin/blob/11f98adcce51f06a85ef0dd64e6e6b16139a34bd/Marlin/src/pins/stm32f4/pins_BTT_SKR_PRO_common.h#L71

I do not see any other place to define it.

Hi Riley. I think FluidNC is a continuation of Bart Dring’s ESP32 Grbl. It will work with Grblgru. I suggest trying to sign on with the GrblMega5x controller in Grblgru if you are using more than 3 axes. Otherwise if only using 3 axes, the Grbl V1.x controller will work. If you use Grbl Mega 5X to try to sign in to a 3 axis controller board like an Uno, you will probably get an error since Grblgru is looking for several axes in that instance. Nothing is damaged of course.
Grblgru has also added support for GrblHal. I’ve found this to be a very nice version of Grbl. I’ve been able to program the GbrlHal firmware onto a Red Mach3 USB controller board. That will control 4 axes. The screw terminals and low price are appealing. It has a 32bit STM32 running the motion control. Naturally Mach3 won’t work on it anymore ha ha.
I’ve also programmed GrblHal onto a MKS Tinybee V1.0 board. This is 5 axis, ESP32 chip = 32 bit as well. This one doesn’t have screw terminals.
The prices of both of these control boards off AliExpress is quite low, often similar to an ESP32 or STM32 development board but with all of the protective bits already built in. Eventually the faster processor speed will be useful for calculating better curves and complex movements.
GrblHal has several Grbl commands that were excluded for memory space reasons in 8 bit versions of Grbl. Those include Grbl versions for Unos and Mega 2560 boards. For example, I’ve put the backlash compensation to work on a Metal lathe I’ve adapted for Ornamental Turning work. FluidNC does not include Backlash compensation as I recall.

I also have a 6 axis ESP32 based controller board off AliExpress. This has Bart Dring’s ESP32Grbl I’m pretty sure. Screw terminals and a good price too. What you get depends on how many axes you want to control potentially.

This may seem too far afield, but I built an MPCNC about a year ago, and now I’m “converting” to a LR3. I think the MPCNC is great at small sizes, and near the origin, but I did find the size somewhat limiting for the projects I’d like to do.

I also found the nut-traps frustrating for tool mount changes. so it left me with a dedicated tool, on a smaller frame. I did eventually use an adapter, but that just slowed my feedrates further.

However, I really enjoyed using it as for a diode laser. What about a MPCNC - Laser edition? I think this would benefit from some interesting engineering choices. How to get more speed, prevent dropped steps, and potentially have options like a rotary axis. I might benefit from a bit more height without the worries you’d run into on a router-based MPCNC.

Lasers are becoming more affordable and mainstream. It also might open the door to engineer an integrated enclosure, with options for smoke removal, and appropriate safety windows (obviously the builder would have to match those to their laser wavelengths.) Maybe the MPCNC-Laser edition could even cut the enclosure pieces similar to the LR3 struts?

I guess the question becomes whether it can be engineered to bring an advantage over the existing cheapo frames for these lasers. However, as you add enclosures/etc. their price climbs quickly.

Just a thought.


I have a couple thoughts for 3.1. Have to try a couple things first though before I start recommending something that makes it worse…


Well that seemed to work, testing this weekend

1 Like

OK, got a few TMC 2209’s added to the board. Thanks Ryan for helping me with the I_MIN_PIN problem.
With the 4:1 driven 4th axis I have, the steps.mm ended up having to be changed to 35 for a 360 degree rotation.
One step closer to the rotary lathe.
FYI: I modified the 515S V1 firmware for this. I will only be using X,Z,A(R)

1 Like

Here is one version of that. It is certainly not yet optimized for usability (bit changes are tricky) but works well and does not eat up a mm of the workspace. New clear dust shoe design!

1 Like

Yes, please!

1 Like

Kinda comes down to a decent tube, power supply, and mirrors.


Not necessarily – there are some excellent all-solid-state lasers out now (combine the beam from several blue powerful laser diodes) that can ride on the Z assembly, so no big glass tube or complicated mirrors to align. For example: https://youtu.be/1b6Uu0Yy2FQ?si=gipjDNolXzPICAqQ


Hmmm, you are right. A primo with a light infill core, and a Z axis that is hand adjusted so no stepper weight. That might be an easy light laser setup.

1 Like

Would that be able to cut acrylics and light colored/clear glass, etc? (My laser knowledge is limited)

The drawback of the Diode lasers is that it struggles with acrylic/glass/etc. This is where my interest in the CO2 laser mostly lies.

Any reviews I’ve seen of lasers other than CO2 that can cut them, the heat is too much to get a clean cut and it leaves melty/distorted cut line.

I think it would be cool to have a design that supported a CO2 laser, and it opens up possibilities to compete with a market of expensive machines. There absolutely could be room to build with a different mount for a diode. Seems like both would benefit from a focus on speed and flexibility.

And besides, who could pass on a name like MPCNCO2?


I’d like to see a redesigned Z carriage on the MPCNC that’s more rigid, easier to shim/tram, easier to swap holders, and possibly available in all-aluminum.

I’d like any access to screw headd be beveled so the v1 screwdriver will work on all of them… or at least most of them. (3dp v4 core xy mounts?)