Modified CoreXY Geometry

In case you missed it, I built a prototype that used the CoreXY geometry in the XZ plane. I call it the CoreXZ. This post is only tangentially about that. CoreXZ works way better than I expected. However, as expected it has bad Z resolution and consequently it has some subtle Z banding. 1 fullstep corresponds with .28mm layers which is way bigger than I would ever want. I could just get smaller drive pulleys but I would like a fullstep in the Z direction to be about .1mm or smaller. That would be a tiny drive pulley and I would give up my speed in the X direction. (If anyone is wondering why I would want to do this instead of the normal slow screw drive Z, I want a quick Z so I can experiment with non-planar printing.)

I just had a brainstorm that would allow me to get 3 times the resolution in the Z direction when compared to the X. With my implementation it takes about $4 more in hardware. The gantry will be heavier but the mechanical advantage should counteract that easily. The Z should still be plenty fast. (Additionally, my current pulleys store just enough string to allow for this design.) The biggest con I could find was that my gantry will be 88mm wider.



What do you guys think? Is there any reason I shouldn't go this route?

Edited 1 time(s). Last edit at 05/05/2014 10:37PM by nicholas.seward.

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

... with adding wire length you'll add elasticity too ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

It seems reasonable. Other than widning the gantry, you can leave off the extra hardware and you've got the previous CoreXZ, so it seems pretty safe.

It might be better use of space if you could fit the new bearings outboard of the existing ones. There ought to be enough clearance.

This is a good arrangement for more Z resolution and keeping the same X speed. It mimics the functionality of other designs that use dual Z lead screws to have a separate Z resolution, but it eliminates the high cost of precision lead screws, as well as the Z periodic errors of low cost lead screws.

The old block and tackle?

This gives you mechanical advantage at the price of length of travel?
Do you NEED to help out the pull of the motors?
Complicated, more sources of error?

I don't think timing (positive drive) belts work well like this?

Must -- only pulleys on motors should be toothed pulleys !

The 90 deg contact on drive pulleys is NOT good.

I don't understand combining X Z axes?
unless one will slice in that plane?
X, Y move around a lot Z moves very little

@VDX: Surprisingly this only increases the length by about 50%. From experiments with my proto, I haven't been able to identify any artifacts from elasticity. I tried to test with the tension on the low side. No problems. From that experience, I would expect with normal tension I should be fine. (I am using 80 or 100lb test string. I can always up that too.)

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

@cozmicray:

The old block and tackle?
You know it!

This gives you mechanical advantage at the price of length of travel?
I still get all the length of travel that I need. I stated above the spools have enough room to deal with this design.

Do you NEED to help out the pull of the motors?
No. I stated above that I would like some improvement in the Z resolution.

Complicated, more sources of error?
Possibly but I doubt laser cut holes and perfectly round pulleys will introduce much in the way of error.

I don't think timing (positive drive) belts work well like this?
I am using string.

The 90 deg contact on drive pulleys is NOT good.
The string wraps around the pulley several times until it get to the end of the pulley and then goes through the center to the other side and wraps back to meet where the other string is. Like I do on Wally and GUS.

I don't understand combining X Z axes?
unless one will slice in that plane?
X, Y move around a lot Z moves very little
As I stated above, I am experimenting with printing layers that are not parallel to the XY plane or aren't planar at all. Besides, this removes a stepper when compared to a Mendel, has a gantry that can't get out of level, doesn't have screw artifacts, and moves the x stepper off the gantry. Oh, and it is cheaper. What is not to love? (Bottom line: I don't want the Z axes to be mostly stationary and I wanted to address the things I find annoying about the Mendel style bots.)

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

With this arrangement you may also find that your steppers have enough clogging torque to hold the axis Un-powered. This would be a good improvement!

nicholas

I feel like this is added complexity for a minimal gain. Would it not be easier to switch to a 400 step stepper to make .14mm your resolution instead of .28mm?

---

WWW.ZATOPA.COM - Your Place for high quality 3D Printing Filament and accessories

@jzatopa: Switching to the higher resolution steppers would be good. I hesitate to rely on them. If I can get by with standard 200steps/rotation steppers that would be optimal. That also doesn't solve my "gantry falls when the power is cut" problem. (The real reason I hesitate is that I have 100s of 200step steppers.)

I don't buy that it is that much more complex. I have to laser cut 8 more holes and bolt in 8 more bearings. Stringing it shouldn't take any longer.

My real hope is that I will have enough static friction to prevent a fall as konwiddak mentioned above. I doubt it but the gantry shouldn't fall as fast.

Cost:
$8 and minimal time

Benefit:
3 times the z resolution
slower gantry falls
works with more steppers
looks cool

Bottom line: this may be a dumb idea but it will be fun to test out. :-)

Edited 1 time(s). Last edit at 05/09/2014 05:13PM by nicholas.seward.

---

ConceptFORGE
Wally, GUS Simpson, LISA Simpson, THOR Simpson, Sextupteron, CoreXZ

A counterweight could be used to help stop the gantry falling - it could even be set up to make it rise upon power failure!

Or a constant force spring (a coil like a tape measure)