Author Topic: SiXYLaser  (Read 23941 times)

Pawpawpaw85

  • Full Member
  • ***
  • Posts: 180
    • View Profile
SiXYLaser
« on: September 02, 2016, 09:58:25 AM »
This is the beginning of a project I call SiXYLaser (Simple CoreXY Laser, preliminary title).

Problem:
Gantry on the laser engraving kit I receive has some flaws, tolerance on profiles are not very good, more resistance on some places than others and suspect it is part of what cause troubles cutting material when using multiple passes. On straight lines it might work OK most of the time, but when the laser is cutting radius, it often fail to track the previous pass properly, creating a wider cut but not deeper.
The laser is mounted too far above the piece to cut and moving it lower creates more unbalance to the mechanics.
Not sure if it would help to change gantry system altogether, but will give it a try.

Goals:
1) Create a cheap gantry system for use of laser engraver/cutter.
2) Increase precision when cutting multiple passes.
3) Use only 2 stepper motors for all control.
4) Reduce cable mess.

Approach:
1) Gantry will be created with easily sourced cheap parts, as well as a minimum amount of 3D printed parts.
2) Linear steel shafts will be used instead of aluminum profiles, hopefully increasing precision.
3) A Core XY type of mechanism for motion will be used, with some modifications.
4) Will try to incorporate the use of cable chains.

So far it's just on idea stage, but have ordered linear shafts, pulleys, belt and other things necessary, and as soon as I get more time, I will start making basic sketches on how to get it all together.

If anyone has suggestions of important features they would like to have on a great gantry system for these low powered laser engravers/cutters, please do post your suggestions!
If someone has a suggestion for a good name then I'm all ears too :)
« Last Edit: September 08, 2016, 07:10:12 AM by Pawpawpaw85 »

ggallant571

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1807
    • View Profile
Re: SigaLaser
« Reply #1 on: September 02, 2016, 01:04:10 PM »
I have observed similar problems and have similar aspirations. I would suggest designing the project with a free host independent 3d modeling environment such as OpenScad.
SARCASM - Just one more service we offer here.

Pawpawpaw85

  • Full Member
  • ***
  • Posts: 180
    • View Profile
Re: SigaLaser
« Reply #2 on: September 02, 2016, 01:15:49 PM »
While that is good in theory; I so far cannot work smoothly on such programs, it'll just be frustrating and wont be productive.
However, if getting a good working mechanics with all dimensions already correct, then "porting" it to a more open environment shouldnt be too difficult for someone that are used to work with those applications.
I will be providing STL files as well as drawings of the important geometry on each 3D printed part for the gantry when the mechanics are working well.

ggallant571

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1807
    • View Profile
Re: SigaLaser
« Reply #3 on: September 02, 2016, 01:31:31 PM »
Throw together a sketch with a few dimensions and I will build the model. You can examine the effort and see if this is something that could work for you.

OpenScad uses a programming scheme where everything is expressed in numerical units make it trivial to change the length, thickness, diameter, etc of individual components. It also allows a component, such as a shaft bearing, to be used in multiple places. It has an export STL option.

OpenScad is quite simplistic. There are other packages such as Blender which provide many for features.
SARCASM - Just one more service we offer here.

Pawpawpaw85

  • Full Member
  • ***
  • Posts: 180
    • View Profile
Re: SigaLaser
« Reply #4 on: September 02, 2016, 02:11:44 PM »
Throw together a sketch with a few dimensions...

That's part of the issue;
Until you got everything modeled and working in a 3D environment, you dont really know of any dimensions or where to put things so that there are no collisions and the gantry moves as intended when the pulleys are turned.

You can probably find some already existing dimensions from 3D printers made with the Core XY principle, but those are most likely over-complex for a laser, and usually work by attaching to aluminum profiles.

ggallant571

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1807
    • View Profile
Re: SigaLaser
« Reply #5 on: September 02, 2016, 07:31:14 PM »
Attached is a possible starting point. Open with generic OpenScad to see the source and play with a 3d parametric design tool.
SARCASM - Just one more service we offer here.

Pawpawpaw85

  • Full Member
  • ***
  • Posts: 180
    • View Profile
Re: SigaLaser
« Reply #6 on: September 02, 2016, 11:01:38 PM »
A Core XY gantry is a bit more complex than a ballscrew/leadscrew setup due to the way motion is transferred, a similar gantry to what we already have is not suitable unfortunately.
It'll be similar to this:
http://frax3d.com/wp-content/uploads/2015/08/CoreXY-Frax3D-31.jpg

That being said, please do continue on your ballscrew/leadscrew gantry! That was my 2nd idea to try if the Core XY fail.
Make sure you have something to eliminate the backlash and it should work fine too!
« Last Edit: September 02, 2016, 11:02:09 PM by Pawpawpaw85 »

Zax

  • T2Laser
  • Global Moderator
  • Hero Member
  • *****
  • Posts: 5366
    • View Profile
    • T2Laser
Re: SigaLaser
« Reply #7 on: September 03, 2016, 04:28:08 AM »
Very nice start Pawpaw, you obviously aren't new to this game.

Pawpawpaw85

  • Full Member
  • ***
  • Posts: 180
    • View Profile
Re: SigaLaser
« Reply #8 on: September 03, 2016, 02:01:29 PM »
Thanks Zax :) I do work with developing mechanical products at work every day, happy to share some of my knowledge and resources with the community :)

I spent an hour today making some rough sketches of the mechanics.

I might actually have solved two things with the Core XY gantry that I have never seen someone do before;
putting all of the pulleys and axis in the same level (Minus a pulley to twist the belt).

This should remove any forces that want to twist the shafts and bearings, hopefully increasing precision in the system.
It also allows to use only one belt -> making it possible to adjust the tension on the belt to be exactly equal for both X and Y motions in all directions.

Here is the sketched parts so far:

« Last Edit: September 03, 2016, 02:07:05 PM by Pawpawpaw85 »

Pawpawpaw85

  • Full Member
  • ***
  • Posts: 180
    • View Profile
Re: SigaLaser
« Reply #9 on: September 04, 2016, 12:43:07 PM »
Worked a bit more on some rough sketches today, and it does seem possible to have the Air Assist I previously designed for these laser modules integrated into the carriage design.
It seems also possible to have the laser module attached simply by two screws, easy if wanting to exchange the module.

You may notice that the laser module is located very high up; The idea behind it is that if the whole thing want to shift in a direction during acceleration, the actual rotation of the laser will be around the rods that are located very low, and not like right now on the current gantry very high up, so same degree of rotational error will reduce the projected error on the working area. Hope it works that way at least :)



Next task is to look on how to attach the belt to the carriage in a way that allows easy to tension it. Also will be looking at the part I've seen no one else make before to get the belts twisted around each others with two pulleys; If that fails I will have to be using two belts instead of one long, and the belts in two heights.
« Last Edit: September 04, 2016, 12:46:27 PM by Pawpawpaw85 »

Zax

  • T2Laser
  • Global Moderator
  • Hero Member
  • *****
  • Posts: 5366
    • View Profile
    • T2Laser
Re: SigaLaser
« Reply #10 on: September 04, 2016, 02:43:59 PM »
Very nice!

You would ideally keep the mass centered vertically in the ways (rods) to minimize rotation, although your bearings look up to the task so it probably makes little difference.

Looking forward to seeing more as it develops.

Lob0426

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1622
  • Let the Magic Blue Smoke out!
    • View Profile
    • Richards RasPi Web Server
Re: SigaLaser
« Reply #11 on: September 04, 2016, 03:22:32 PM »
If the drawing scale is correct, you may want to consider the 12mm smooth rod for your Y axis. Droop is going to cause focus problems in the middle of your work area. The X rods probably will be ok as they are doubled up.
Richard
A5 2.5W w/homing switches on L7
A5 2W diode self upgraded from stock 500mw
Printrbot Play. Ext bed, integrated tablet control.
Printrbot Simple Metal
USA

Pawpawpaw85

  • Full Member
  • ***
  • Posts: 180
    • View Profile
Re: SigaLaser
« Reply #12 on: September 04, 2016, 09:24:56 PM »
You would ideally keep the mass centered vertically in the ways (rods) to minimize rotation, although your bearings look up to the task so it probably makes little difference.

Zax, what speed are you usually using when engraving? I only do slow moves when cutting so the offset in mass center hopefully wont matter, but if using high speeds and acceleration it may cause issues.
There are also other things to consider beside mass, like cables that will have different amount of pull on the module at different locations, so even if mass is centered, the moment acting on it will differ, thats why I think offsetting it will reduce this effect, even if the cables are pulling on it, then the little offset in rotation it creates will be small. Having the rods further up, the rotations gets enlarged.
But you may be right, perhaps better to start with the axis further up and lower them if there is an issue, that way it should be more suitable for high speeds and hopefully still work with lower speeds.

If the drawing scale is correct, you may want to consider the 12mm smooth rod for your Y axis. Droop is going to cause focus problems in the middle of your work area. The X rods probably will be ok as they are doubled up.

Both are using 8 mm rods, 2 for X axis and 2 for Y axis. (Reason being to keep the same rods everywhere, and that they are much cheaper than 12mm)
I've calculated a deflection of approximately 0.23 mm on each axis at the center of the beams if the carriage weight is 1kg, giving a total vertical deflection of 0.46mm, do you think that this will be an issue?
The carriage will hopefully be much lighter than that.
« Last Edit: September 04, 2016, 09:35:32 PM by Pawpawpaw85 »

Lob0426

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1622
  • Let the Magic Blue Smoke out!
    • View Profile
    • Richards RasPi Web Server
Re: SigaLaser
« Reply #13 on: September 04, 2016, 10:20:36 PM »
Half a millimeter is not a whole lot, but you may find it is more. I guess it depends on what you are cutting and if it is in the center of your work area. It comes down to the end weight of the carriage, I doubt it will be 1Kg.

Test it if it is an issue. You can print enough material on your corner pieces and Y slides to drill them out to 12mm, if need be. PLA and ABS drill pretty easily and accurately. Once you see what works then you can reprint for the right size.
Richard
A5 2.5W w/homing switches on L7
A5 2W diode self upgraded from stock 500mw
Printrbot Play. Ext bed, integrated tablet control.
Printrbot Simple Metal
USA

Pawpawpaw85

  • Full Member
  • ***
  • Posts: 180
    • View Profile
Re: SigaLaser
« Reply #14 on: September 05, 2016, 01:34:02 AM »
I've now also ordered 4pcs 12mm rods just to have them ready in case the 8mm fail. If not needed I can probably find some other use for them  ;)

Cheapest I've been able to find the parts for the gantry:
1pcs 12 mm rod L=500mm cost ~$10 (shipping included)    2pcs RJ4JP-01-12 cost ~ $10 (shipping included)
1pcs   8 mm rod L=500mm cost ~$4.5 (shipping included)   2pcs RJ4JP-01-08 cost ~$5 (shipping included)

Changing to 12mm rods and bearings will more than double the price of the gantry. Hopefully it's not needed!
« Last Edit: September 05, 2016, 02:17:56 AM by Pawpawpaw85 »