Loading

CoreXY

by frankie, published

CoreXY by frankie Apr 25, 2012

Description

CoreXY is a mechanism for building fast, compact XY motion stages. While it is not a CNC tool in of itself, the intention is that it might be incorporated into 3D printers, milling machines, etc. The design is described in greater detail at http://www.corexy.com . A video is at http://vimeo.com/40914530 .

Recent Comments

view all
Both Marlin and Repetier have preset configurations for CoreXY / H-bot style gantry designs.
Hi, this rocks! have 2 questions.
Which program do you use to cotrol this movement?
i gues it's strong enough for a 3d printer that has no resistance, but could it work on a CNC milling machine?
Thanks!
Do you have a picture perhaps? I am having trouble visualizing what you described. Thanks!

More from Machine Tools

view more

Liked By

view all

License

CoreXY by frankie is licensed under the Creative Commons - Attribution - Share Alike license.

Give a Shout Out

If you print this Thing and display it in public proudly give attribution by printing and displaying this tag. Print Thing Tag
Hi, this rocks! have 2 questions.
Which program do you use to cotrol this movement?
i gues it's strong enough for a 3d printer that has no resistance, but could it work on a CNC milling machine?
Thanks!
Both Marlin and Repetier have preset configurations for CoreXY / H-bot style gantry designs.
It's an Etch-a-Sketch. :)
if you've already got the metal exterior frame, why do you need the bars? can't your carriage ride on the frame?
I was implementing H-bot in Cartesio, but stopped
CoreXY is actually better, and I will implement CoreXY
In H-bot if you move 45 degrees only 1 motor is turning, and if the belt is not perfectly aligned and tensioned  the opposite side of the gantry from the motor will start moving later, resulting in serious backlash
This will not happen with CoreXY
I will keep you updated
jos - in reply to jos
 Edit: the same thing will happen for CoreXY
picture the reference mechanisme pic on your website.
Now if you move the toolholder from top left to bottom right in a 45 degree angle,
The "B" motor will be doing all the work resulting in the "A" side of the gantry to start moving ever so slightly later, resulting in backlash
I can see fab in box in the video and photos.are the plans for that structure open source?
is there going to issue with belt tension
Why do you anticipate an issue?
Any further news on this? Pretty interested in firing up the lasers at work and scaling this up a bit and building one to see where it goes.
If you reshuffle the belts like this maukcc.blogspot.com.au/2011/12/cartesio-update-v04.html you can make the X and Y drive independent which makes control easier. Alternatively you can do it with an H layout like this youtube.com/watch?v=ei4lPk_aM9Y which will simply it a bit, note the tool head is only connected to one of the sections of belt crossing the gantry.
Hello,

nice design :)

Can we use the H layout with the printer firmware?
No, or at least no more so than the corexy design, the H layout is just simpler, but it has the same problem of coupling the axes and so requiring modified firmware.

The layout I linked from maukcc doesn't have that problem.
I've modified the Marlin firmware to support CoreXY - it seems to work fine but I'm not sure what will happen at high speeds (I've run it successfully up to 100mm/s).

github.com/imoyer/Marlin
Actually i was thinking on same thing that it can be converted in to H bot design.also thinking of placing igus or LM linear bushing instead of brass bushings. some thing similar to this. reprap.org/wiki/Kludgebot have you tested h bot config ? i am making this thing and trying out.Have you tested this any 3d printed toolhead (extruder) also what was the load on carriage when getting the testing done ?
The H is actually functionally identical to the corexy design, just simpler.

Llooking at this page corexy.com/theory.html the belt is connected to the carriage in 4 places.

If you detach the top two and join them together nothing changes, because there is now only a single belt it must obviously move the same as it did before.

You can then flip the loop from the top left pulley to the top right pulley (via our new connection) horizontally. That removes the cross over and the inner top left and right pulleys again without changing the dynamics.

And now you have the H layout.
3runo - in reply to Guest
Do you have a picture perhaps? I am having trouble visualizing what you described. Thanks!
Indeed, CoreXY is functionally similar to the H configuration. If you look at the reference section on the website there's a few other interesting bits of information on the H mechanism. Because CoreXY has two belts instead of one I suspect it's bit stiffer than the equivalent H-bot. Of course you could use twice as thick a belt for the H config but that's not as readily available.

I'd need to do more thought and analysis to properly compare the two designs, but in addition to stiffness I'm curious about whether there's an advantage to not needing to clear a belt passing by the carriage and also how the corexy config compares at stabilizing the stage.

The truth is that the
design was conceived as outlined in on the theory page so I'm still learning about it. If you have any further thoughts please send them my way :-)
A video of the H layout in action youtube.com/watch?v=1HxhowShRbE
Cool design.

Have you considered using bearings directly on the aluminum instead of using the linear shafting? Something like this: cncrouterparts.com/linear-carriage-with-abec-7-bearings-p-33.html?cPath=21

Keep up the great work. 8-)
I agree with your philosophy of keeping the moving mass low. I have not been able to combine x and y motor movement with Marlin or Sprinter and still have jogging function. If someone could do the firmware modification that allows us to combine axes, it would open up some great new possibilities. How are you accomplishing the combined x and y movement? Great work!
I've modified the Marlin firmware to support CoreXY - it seems to work fine but I'm not sure what will happen at high speeds (I've run it successfully up to 100mm/s). Jogging works fine.

github.com/imoyer/Marlin
thanks man. great work.
I'm not familiar with the workings of Marlin or Sprinter but a forward step in the X axis requires forward steps on both motors, while a forward step in Y requires a forward step on one motor and a reverse step on the other. If the firmware is written so that taking steps on each motor is broken out as separate functions it should be an easy modification. But in practice I don't know how easy it is to modify Marlin or Sprinter. I use my own control boards so the change was trivial.
so which firmware is used in video of pcb milling
I'm using firmware that I wrote - it isn't as good as sprinter or others but I know how to modify it.
were there any limitation to movements, as i am reading about it it comes to know that its only good for point to point movements,i havent seen any arc movements with this type of mechanics
All CNC machines that I know of operate on a grid defined by the resolution of their positioning system. Because of this arc movements are created by a series of very small point-to-point motions. CoreXY can move thru curves quite well.
my friend and me are working on firmware modification.ask kliment on irc, orth iti told him mant times but he think its not it
I love this idea. I had a similar design planned for my RepStrap printer before mounting cost made be abandon it.

I might still build one in the future...
Moving knots are really great for true/syncing up ones axises as well: cnczone.com/forums/diy-cnc_router_table_machines/51485-make_your_gantry_rock_solid-3.html#post554923 (see the whole thread... or well the forum btw)
Top