Thank you! Kinematic control is done by an Arduino mega, and image processing by a standard computer. Nothing special.

Thank you!

Thank you!

Gracias. Esta maquina es una fresadora cnc Sherline modificada.

You're right. I choose this speed due to some not solved issues. As I understand spindle speed for this job should be over 4000rpm. Thanks.

Full units are now available for sale at my site. Maybe I put kits for sale later. Thanks for your interest!

Las modificaciones están descritas en mi blog: automaticartisan.com/blog/

It's real time detection. You don't see webcam because it is just out of the field here, static on the left top. You can see where is the webcam here: instagram.com/p/BzbLaB-ntUz/

Hi!. Cutting parameters on description. I really don't have a reference because for me it's always a trial-error thing that depends on tools, clamping, wood hardness, etc. And please note you want much more than 2800rpm; wood milling isn't really my business.

Thanks!. Well, not exactly; the were made on a cnc mill and then hand finished.

Just check the links on the description. Thanks for watching!.

@@AutomaticArtisan Thank you!

Thanks Pete! Well, my opinion at this point is that Wangsamas firmware (Repetier based, Due (32 bits) ready) works better for a scara and brings more features. The reason I'm using Marlin is because of some minor issues I had with Wangsamas, but I'm almost convinced now that those issues were my fault, so I plan to come back to Wangsamas. Please note I'm using Marlin 1.0.2, as Scara support is broken on 1.1.9 and previous releases. I've not tried Marlin 2.0.x bugfix source.

@@AutomaticArtisan Thanks! I'm definitely keeping an eye on your progress. It'll probably be a few weeks before I get to the firmware portion, so I've just been keeping notes on what might work. I've already got the Wangsamas on my list. Did you by chance try any of the Cartesian-to-Scara coordinate conversion scripts that are out there? (I'm assuming they might work, but not as well as native SCARA firmware.)

Thanks!. My aim is to release the design as open source at some time, but this will not happen soon as my current focus is to finish the first units: close up design, writing gcode, machining last parts, etc. It's always harder than it looks.

Fantastic men!! @@AutomaticArtisan any idea if u can share code?

@@OtroPl4n Thanks!. Well, in this video I'm using plain grbl (no scara support) for just testing. Lately I've been working to adapt Marlin 1.0.2 (it has WORKING Morgan scara support) for this arm, and It's almost done. I hope to share that code soon, within one or two weeks.

@automatic artisan It really looks good! I have done similar project for my Uni last year but ended up to write my own code in arduino because I can't find a open source code that fits my need. May I know did you have any progress in this? Possible to share your codes? Thanks!

@@AutomaticArtisan I imagine the calculations to turn cartesian movement into this eggbeater style is quite a challenge!

Well, there's no problem if your machine is precise enough, which in turns translates mostly on how much backslash do you have on the Z axis. Nominal stepper resolution on every axis is 0,225 degrees, which means submicron resolution. Of course I can't get that precision on my machines due to backlash and rigidity issues. Big, expensive machines, can easily reach submicron precision, though. Thanks for your interest.

Put an end mill in there and try that. That tap is definitely creating the thread. Depth has nothing to do with the calculations for tapping, it's a function of spindle rpm, pitch of the tap and feedrate. This machine looks just fine. It's doing what it's supposed to do.

Rigid tapping by definition is the coupling of acceleration and deceleration of the drive axis with the rotating axis. The axis feedrate = the thread pitch, and the coupling ofthe above accel/decel ensures that it is indeed cutting during the whole speeding up to thread pitch feed rate, slowing down to a stop, then the reverse to get out of there. Generally, you use rigid tapping with cold forming taps, that press the material into the thread profile, indeed not cutting, but displacing. This generates a stronger thread, and no swarf(chips) that you get with a cutting tap. But is does require a LOT more torque and rigidity. Using rigid tapping on a cutting tap is mimicking you tapping by hand, it can be a much better option than floating tap holders, that can get jammed.

I agree wood has a big potential as aluminum alternative on non critical applications. It has great properties and appeal. Structural stability can be a problem though, mainly on big parts.