Adding kinematicsType for XY machine

[Ned]

Hey guys…

Im messing with my maslow and am trying to turn it into a really cheap XY machine using as much original maslow gear as possible, but attempting to get the 1/64" accuracy we get in the middle, across the entire sheet.

I know this has been talked about before etc, and i dont wanna start a discussion about it here but wanna talk about the code.

I read through the whole lot last night, and as far as i can tell, to do what i want (and make it usable for others, as thats the open source idea) i should create a new kinematicsType for XY, add a new function to the kinimatics code to turn XY into chain lengths, and thats about it?

The triangularInverse function has some easy enough maths to follow for all the chain length compensation etc, so i can just copy all of that.

Are there any pit falls i should be aware of? i havent spent a lot of time looking at the code, and im staill building the frame so wont have any time to look at it until tomorrow i guess, after i (hopefully) finish the frame.

Just putting it out there, is there anyone who knows the code intensely well that feels like adding XY to it? it seems like it should be pretty easy for someone who knows whats actually going on and would be cool to have it work once i finish the frame. No biggy if you cant, seems like it should be reasonably straight forward.

Frame wise, imagine a DIY panel saw that moves in X and Y with some motors and chains basically… Im hoping it gets rid of the had maths, and adds accuracy across the board for not much more money. ANd if it doesnt work, i’ll build a normal maslow and use the frame im making now as a panel saw

[blurfl]

I’m not certain, but I think that with these step-direction controllers to drive the gearmotors you could use LinuxCNC software to run the gcode. There really isn’t any need for the Maslow firmware or GroundControl in a machine that doesn’t use quadrilateral or triangular Kinematics.

[Ned]

From my understanding the maslow code interprids normal gcode and its the arduino that converts to funny angles, so should be able to use everything but have it not convert to funny angles and just drive it as a normal xy machine?

[dlang]

allof the triangular calculations would be meaningless if you are talking about
a traditional XY kinematics.

But if you are wanting to build a more traditional machine, instead of modifying
the Maslow firmware to support different kinematics, you would be far better off
modifying grbl to support motors/encoders instead of steppers.

grbl has a much more mature g-code parser, it’s got a good acceleration planner
(something maslow really needs), and it’s got a much larger community supporting
it.

grbl also supports both traditional and corexy kinematics.

We have an ongoing desire to try and leverage/merge the maslow specific things
into grbl to take advantage of it’s features (and support in CAM software), so
work to support it advances the long-term goals of the Maslow project.

David Lang

[bar]

I read through the whole lot last night, and as far as i can tell, to do what i want (and make it usable for others, as thats the open source idea) i should create a new kinematicsType for XY, add a new function to the kinimatics code to turn XY into chain lengths, and thats about it?

That sounds right to me! Having an option to choose regular XY kinematics from the settings drop down would be a few more steps, but is still pretty straight forwards.

Please make a PR when you get it running, I think it would be great to give people the option to play around with making different types of machines. As much as the Maslow hardware isn’t designed for an XY machine it could be a fun an easy way to play around with building one

[kf2qd]

Would be a way to upgrade - a flatbead machine with a bit more stability and not have to change the software system

[TheGasse]

Hi Ned,
i know the post is old - but my question is if you have done changes in the kinematics.cpp for using the Maslow Hardware as a lienaer X-Y System. We are actual trying to get such a System running. Are there any hints to use the Maslow HW for a X-Y System.
Thanks Greeting from Germany

[Orob]

This is purely an observation comment with no depth other than a “what if” scenario. Having been though most of the code… are you wanting to use webcontrol or groundcontrol with it? If not, what if you started with marlin or one of the other firmware distributions and use the encoder and motor modules from this work so you have the dc motor and encoder pportions instead of the stepper functions?

[Ned]

Changing the code was easy. Maslow works in XY until the final stage where it calculates the hanger plotter thing. I dont remember the exact code changes, but I basically just commented out the bit that made it calculate the triangle stuff, and that made it ‘just work’
But I never finished it. Part way through I found the Lowrider 2 from V1 Engineering, and asked myself why I was redesigning the wheel, and just built that.
I then used the maslow frame as a plywood storage rack, and used the Z axis i designed as a router lift for a router table

[TheGasse]

We want to use the Webcontrol-Interface. We just addad some Lienar-Guides for X and Y Achses because we also want to use the System for leveling massiv wood in similar dimensions. It would be also possible to hang the sled to the Linear-Guides but from the contruction point of view its easier to drive the sled directly in each axis direction with the chains and we also can save space (hight) by the frame. But in generall it will be a upstanding milling machine with all the original Maslow Components and just adding the Lienar Guides wich will give us the freedom that the sled don’t has any contact to the milling object. We will try to modify the code and see if we get the System up an Running. But if there are further ideas i am open to it - I’m not the programming crack but i hope i can handel it

[TheGasse]

We also managed it and now the Software is “just working”. X-Axis is already moving

Y-Axis will follow tomorrow.

[madgrizzle]

Any chance you can post more pictures of your build? Looks nice.

[mooselake]

SBR rail? I can’t see any seams when zooming in, must have been pretty spendy just for shipping. Wonder if you could wing that with conduit or fence rail pipe keeping with the maslow low cost tradition.

[madgrizzle]

There’s a lot of flexing with conduit and tubing at those length though maybe it could be calibrated out. Otherwise, you really would need to support the conduit/tube along the span. Maybe a bunch 3D printed support block that can mimic a SBR rail?

[mooselake]

That’s kinda what I was thinking, or maybe get the swamprats a TIG welder

[TheGasse]

Yes sure, i will keep you updated during the buliding progress
Today I was building the Y-Axis and the prelimanry Sled for the Router this will be improved later also with some Rails. But for now i will attach the Z-Axis directly to the router … task for the weekend …

@mooselake - yes you are right i’m using SBR-Rails - shippment inside of Germany and price was quit Ok. Originalplan was to build a “normal” Maslow. But i was also searching fo a Tool to planing wood (oak) in very big dimension so the idea came up to upgrade the Maslow by Rails. But yes it breaks with the idea of a cheap system.

[madgrizzle]

Looks nice. Make sure you do something to prevent the saw dust from getting in th bearings on the bottom rail. Maybe some brushes or “blower” or something. Lots of dust will be dropping down even if you have a good vacuum port on the sled.

[lamvpr]

Nice job!!!

[TheGasse]

There is a litte update of my building process.First I installed a technical update to avoid problems with the long lever of the x-axis, a shaft was installed and the axis is now driven synchronously at the top and bottom .

All cables are guided in drag chains. A RPi with a touchscreen is running Web control. Last missing part is the distribution Box :-).

Slowly but continuously progress.