CoreXZ

I would be quite happy to add 3:1 or variable ratio CoreXZ support to RepRapFirmware (for Duet electronics, but currently being ported to RADDS and Alligator); however I won't do it unless there is someone who has a CoreXZ machine and can test it.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I would be glad to test it. I personally have 10 and tons of my students have them also.

---

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

Quote
nicholas.seward
I would be glad to test it. I personally have 10 and tons of my students have them also.

OK then, I will implement it in the next release. When do you expect to have a CoreXZ machine with a Duet, ready for the firmware? Also, am I right in thinking the following:

1. A CoreXZ machine is like a CoreXY machine with the Y and Z motors swapped and the Y and Z axes swapped;

2. The 3:1 ratio on the Z axis means that X movements are calculated as for a 1:1 CoreXY machine, but for Z movements I need to move both the X and Z motors 3 times as much as for a 1:1 machine?

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I will order a duet setup today. As to when I will have it done, probably before Christmas would be a safe time frame. I know I have to build a few before then. I don't see myself replacing electronics on working machines.

1) correct
2) correct
(capitals are motors and lowercase letters are directions.)
If X and Z both rotate clockwise one rotation when looking at their face-plate then the x will move a distance equal to the circumference of the pulley.
If X rotates CCW and Z rotates CW then z will move a distance equal to circumference/3.

Bonus points if you can allow for any rational ration. For instance, I have considered making x 1:2 and z 1:3.

---

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

I've made the firmware changes, and added XYZ parameters to the M667 command to allow axis scaling factors to be set, see [reprap.org]. These changes will be in my next release.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42
I've made the firmware changes, and added XYZ parameters to the M667 command to allow axis scaling factors to be set, see [reprap.org]. These changes will be in my next release.

This is now released in version 1.07j at [github.com].

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42

This is now released in version 1.07j at [github.com].

Thanks for doing this! Do you have any more information about the RADDS port?

The RADDS port has been done by Dan Newman. He isn't making the binaries public until he has done some more work, but I understand he has made binaries available privately. Or you can build the binary yourself from his github sources. He started from my fork, but I don't know whether he has integrated my latest changes yet.

RADDS/Due has some significant disadvantages compared to the Duet: it has no Ethernet port (so no web interface), the SD card is accessed via the slower SPI protocol, and you don't have software control of the stepper motor currents.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thanks for the info on RADDS - I'll look into it.

Quick question: When you load the XZ spools, do you mirror the wraps, or wind both spools the same direction? (Does it matter?)

Quote
scottyo
Quick question: When you load the XZ spools, do you mirror the wraps, or wind both spools the same direction? (Does it matter?)

It doesn't matter. Once together you might have to flip stepper wires or change stepper direction in firmware but that is typical.

---

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

Has anyone got Nicholas' MarlinXZ firmware to work with Ramps 1.4? I am having the same issue as MacMarty. XZ motion is as expected, but I get no Y motion. I changed the motherboard from 65 (Azteeg X1) to 33 (Ramps 1.3/1.4). I checked pins.h, but that looks correct.

I could run the RAMPS14Test firmware and get motion from all steppers. So I think I am missing a setting somewhere in Nicholas' code. Any help would be appreciated.
Thanks,
Scotty

It's bizarre, but changing Y_ENABLE_ON to 1 in Configuration.h did the trick. (Leave the others as 0.) That shouldn't be that way I don't think, but it worked.

My adaptation of Nicholas' CoreXZ is working very well (I call it "QuorXZ"). I am extremely pleased.

Here it is in action:
QuorXZ Video 1
QuorXZ Video 2

Thanks for your design, input, and firmware Nicholas!

Congrats! That looks very well built. Looks like you have lots of room to push up the performance based on your rapids. I wouldn't be suprised if you couldn't print 2 or 3 times faster.

---

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

Quote
nicholas.seward
Congrats! That looks very well built. Looks like you have lots of room to push up the performance based on your rapids. I wouldn't be suprised if you couldn't print 2 or 3 times faster.

I already ran a print using Repetier Host's 200% of my basic settings (making it 80mm/sec), and the quality looked just as good as the original. That surprised me quite a bit. (Sure shakes the table though.)

Change the acceleration values down until the table is okay.

You also want to speed up normal moves but not the rapids.

Edited 1 time(s). Last edit at 09/25/2015 10:03AM by nicholas.seward.

---

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

Thanks! I'll give that a try and report back.

Hi all, I'm working on Smoothie{ware,board} support for CoreXZ. You can see a preview here: [github.com]

These changes build successfully, but there are probably bugs and bad signs in there. I've just finished my CoreXZ build and will be testing this on my machine tonight.

X and Z axis reduction will be easily configurable.

Edited 1 time(s). Last edit at 09/30/2015 01:20AM by jnj.

Has anyone hacked Nicholas' code to get MIN endstops working?

I've been fiddling with the configuration files because I wanted to use auto bed leveling, but although my probe and limit switches showed to be TRIGGERED properly, nothing was stopping the stepper motion. Then I had a big "DUH" moment where I realized that the actual code had to have been changed, because triggering either X stepper or Z stepper to stop wouldn't actually stop both at the same time.

I'm guessing this is not trivial to do, but I'm not afraid to code if pointed in the right direction. (@Nicholas: Do you have the original Marlin tree you modified? I could capture a diff, and that would at least get me started - Thx.)

NOTE: I don't mind manually homing, but it would be nice to have at least the Z Endstop do the right thing so I can enable Bed leveling. Any hints? (On a side-note, why would Y Endstop not work as normal?)

On another side-note, does either the Smoothie or Duet code mentioned above take auto-leveling and/or endstops into account?

Edited 1 time(s). Last edit at 10/03/2015 07:27AM by scottyo.

Quote
scottyo
On another side-note, does either the Smoothie or Duet code mentioned above take auto-leveling and/or endstops into account?

The Duet code supports 3, 4 and 5-point bed compensation. I don't know exactly what you mean by "taking endstops into account". The Duet code gives you a choice of a maximum or a minimum homing switch for each axis, or you can use the Z probe in lieu of a low end switch where that is practical (e.g. Ormerod printers use a switch to home the Y axis and the IR proximity sensor to home X and Z). Once homed, the printer will refuse to exceed the high and low axis limits specified by then M208 commands in the configuration file, unless you choose to override this protection deliberately.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42
I don't know exactly what you mean by "taking endstops into account".

I am using RAMPS and Nicolas' version of Marlin. I might be making a wrong assumption about his code (and probably asking the wrong question). I am guessing he modified the CoreXY code for his CoreXZ implementation, and needed to somehow modify the X and/or Z minimum endstopping routines. Since leveling on the RAMPS depends on the Z-endstop, and both X and Z motors have to stop, I can see how that would require different considerations than a Core XY machine. I've not yet dug into the code very far (I've only grepped for COREXZ to find the files involved), and it might be more obvious once I do. But if not, or if I can't get it working, I might break down and buy a duet or a smoothieboard for my printer.

I do not remember what version I used.

---

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

Quote
scottyo

Quote
dc42
I don't know exactly what you mean by "taking endstops into account".

I am using RAMPS and Nicolas' version of Marlin. I might be making a wrong assumption about his code (and probably asking the wrong question). I am guessing he modified the CoreXY code for his CoreXZ implementation, and needed to somehow modify the X and/or Z minimum endstopping routines. Since leveling on the RAMPS depends on the Z-endstop, and both X and Z motors have to stop, I can see how that would require different considerations than a Core XY machine. I've not yet dug into the code very far (I've only grepped for COREXZ to find the files involved), and it might be more obvious once I do. But if not, or if I can't get it working, I might break down and buy a duet or a smoothieboard for my printer.

Should be no problem. When running my fork of RepRapFirmware in CoreXY/CoreXZ/CoreYZ mode on the Duet, when any homing switch is triggered during a homing move then all motors are stopped. This has implications if you want to set up the homing files to do simultaneous X and Y homing, as I describe at [reprap.org].

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

In case anyone is interested, the Smoothieboard code has been tested and merged into the main Smoothieware repository. There is also some sample configuration that you can put in your config file: [github.com]

It's working beautifully on my CoreXZ

Re: endstops: I don't use them, but they will work just as well as they would on any other Smoothieboard-based setup (i.e. just fine)

Has anyone got auto bed levelling working with Nicholas's firmware mod? It doesn't work with the Marlin 1.1.0 RC2 version of CoreXZ.

Edited 1 time(s). Last edit at 11/09/2015 06:35PM by Gina.

Looks like I shall be changing my CoreXZ to a cartesian arrangement. A few extra pulleys but I can still retain the 3:1 Z drive.

Repetier supports CoreXZ but I don't know about auto leveling. (Note: I have leveled my CoreXZ 1 time over the last year. That includes about 12 car trips.)

---

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

RepRapFirmware supports CoreXZ with up to 5-point auto bed compensation. It doesn't yet support grid-based bed compensation.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I posted an issue about this on Marlin github and it's a verified bug. Link here - [github.com]