Ridiculous speed upgrade: collaboration

(Oskar Weigl) #1

Fast machines are badass. Simultaneously precise and fast machines are even more badass.

Shapeoko is already very cool; but wouldn’t it be awesome if it also went at ridiculous speeds? If you also think so, then maybe we can work together to create something next level!

Let me show you some progress I have made already:

This is a modified LitePlacer; the mechanics of which are somewhat similar to the X-Carve, at least when it comes to the rail system. It is retrofitted with 2kW brushless motors from Hobbyking, and a motor controller I made myself.
However, on this machine we are not able to get even close to the 2kW; the limiting factor is the 6mm GT2 belts, they jump teeth on the pulley if the motor torque is too large. Shapeoko is a much sturdier machine, so naturally it would be a good candidate for a more powerful performance.

I don’t have a Shapeoko, and its a bit too excessive to get one just for a fun project like this. Instead, I think this would be a great opportunity to team up and do a collaboration. Projects like these are way more fun to do together with someone else anyway. So if you are interested, then that would be awesome!

I imagine it will go down something like this: I supply the motors, controllers and motor control know-how; you supply a shapeoko, a sturdy table (we will need it), and shapeoko/milling know-how. We play some tunes, get pizzza, and through the course of a series of hack sessions, together we make something truly awesome. When we are done with the project I’m happy for you to keep it: I have a bunch of spare motors and controllers so I’m not too worried.

Primarily I’m looking for someone in the SF bay area, but anyone who is interested please join the conversation anyway!


(Daniel Loughmiller) #2

It’s neat to watch but it sounds like it would only be moving that quickly during fast travels, when it’s not cutting is that correct? Powering through a cut is going to lead to deflection and bit breaking no?

With Carbide Create where it is, optimizing the toolpathing to avoid moving all over the place would be a huge speed improvement for certain types of jobs I suspect.

(Oskar Weigl) #3

I don’t actually know very much about milling: but will the feed force increase if you increase the spindle speed and the feed rate together, keeping the chip load constant?

(William Adams) #4

I believe so — been trying to wrap my mind around that sort of thing at: Tutorial on feeds and speeds

(Oskar Weigl) #5

Thanks for the link to that tutorial. I saw lots of discussion about spindle speeds and feed rates, in relation to chip load and material removal rate. But I didn’t see anything about feed force (maybe that’s not the right term, but I mean the force on the frame and hence motion motors). I could be just looking in the wrong place, again I don’t really know what the terminology is.

(William Adams) #6

The community has a glossary here:


Please let us know if there’s a term you’re looking for, but not finding there.

(Oskar Weigl) #7

Thanks! Hehe I guess it’s kind of hard to do the lookup in reverse: I know the concept, but not the term.

I found the term “cutting force” in the glossary. However, correct me if I’m wrong, but I believe this refers to the force right where the cutter hits the material? The concept I’m trying to find a name for (and hence some equations for) is the force on the linear axes of the machine.
My guess is that the linear force is proportional to the force at the tool: and in the glossary it says that the cutting force doesn’t increase with speed (only the power does, which makes sense since power = force * speed).

Am I understanding this correctly?

(William Adams) #8


I’m not aware of a specific term for what you’re describing.

(Oskar Weigl) #9

How much time is spent cutting vs travelling usually?

(Daniel Loughmiller) #10

Really depends on the project. Optimized to minimize wasted travel, and pulling this number somewhat out of thin air, I’d guess 80-95% cutting. Drops significantly if the software generating the gcode uses a simpler algorithm which has the bit moving all around.