Aluminium Frame and Linear Sled

Hi all,

I put together a variant on the Maslow / Maslow II kits.

This design is composed of aluminium extrusions and has a linear sled to help with stability, a thin Z axis as well.

The concept is not finished yet, I just thought I would show you the design that I have for now. There are lots of changes that need to be made but it is entirely possible that the linear sled could run on the wooden version of the Maslow frame when I have finished the design.

!

20200512_135258|video
I hope you like this!

Thanks

Dave

6 Likes

There are a few builds like that on the forums, they’re pretty interesting. If I were to do such modifications to mine, I think I’d still leave a way to use it the “regular” way.

What is often seen as a weakness (the sled slides on the surface of the material to cut, thus making it impossible to make “true” 3D things like wings, bowls, etc), is also a strength, because it can accommodate warped plywood without any issues.

Having a machine that can do both would be the best of both worlds.

1 Like

Thanks for your information. I have a bit of work to do to finish this sled project but I am considering three options at the moment:

  1. Standard Linear Sled for the Maslow lumber frame. Some small modifications to the frame would be necessary for the sled to attach to the frame

  2. As shown, a linear sled based on an aluminium frame.

  3. A raised sled option.

  4. Not Maslow option, but vertical full XYZ full 3D twin rail system, will cut thicker and harder materials.

Option 1 or 2 are for this material ply / MDF cutting and are both based on the maslow concept of the sled touching the surface area to be cut, EXCEPT that these sleds will be slightly raised so they do not contact the surface to be cut.

option 1 and 2 PRO’s

  • Comparatively low cost sled kit in keeping with the Maslow ethos
  • Sled is raised above the surface to be cut so the sled does not touch the surface
  • Less friction because of above
  • Sled is a complete linear kit, easy assembly
  • The rails help maintain rigidity and fluidity of the sled
  • Compatible with the chain drive system
  • Designed for 2D cuts only but some minor 3D work may be possible
  • Easy attachment to existing Maslow Frame
  • Potential to slide sled to far left to load cut material… (need opinions on the feasibility of this)
  • M8 bolt on the wheel sets, rigidity guaranteed
  • Option to use traditional GRBL based XYZ controller and Steel insert belt drive.
  • Option for Aluminium frame to be wall based (flat) and pivot when needed to work
  • With traditional GRBL XYZ sled model smaller frames can be made e.g. 4x4ft option
  • UK has access to up to 3M linear rails so no joints needed.

Cons for option 1 and 2

  • The main linear rails would flex if under too much load, hence the need to limit the distance from the cut material
  • Extra cost to implement compared to traditional Maslow kit
    Existing Lumber based frame would need slight alteration (again opinion on this would be appreciated) more details of what would be needed to follow
  • The aluminium frame is more expensive than the lumber frame but potentially easier to assemble
  • USA does not sell longer than 1500mm rails so joints would be needed for the aluminium frame
    There are probably more cons but can’t think of any more at the moment

Option 3

Pro’s
Stronger linear sled rails, allowing full 3D cuts and a ‘proper’ Z axis with 5in distance from bed material

Cons
More expense

Option 4
Not comparable to Maslow for cost but a full vertical CNC kit based on the same linear sled concept. Basically more like a traditional vertical CNC machine but nowhere near the price of the lumber based Green Tree system

I will post more pictures for the options shown above as soon as I get some more free time and once I have modelled the sled option fully for options 1 and 2

Thanks

Just out of curiosity: what are you trying to achieve here? What’s the end goal of these improvements?

@TomD has built a gantry-type Maslow that he calls “Mark II”; his goal was to do some flat-ish true 3D stuff, like a rudder / daggerboard. See here: Maslow Mark II - 3D

If you’re only looking to cut 2d stuff, having the sled not touch the workpiece is actually an inferior solution, IMO. The fact that it touches puts your Z-zero at the surface of your stock, no matter how warped it is.

If you’re looking to build a traditional CNC, but in a vertical orientation, your Y axis will need to use a ball screw of some sort, instead of regular steppers & belts. A stepper will not keep your sled / carriage in place when it’s turned off, which was the reason for going with these “weird” motors in the first place.

Also, I don’t think you’ll get much additional rigidity by going “closer” but not touching.

Good point on the friction, though; you’ll probably get a lot more flexibility on the angles with a gantry setup like this.

Cool concept. You may have mentioned it, but have you given thought to loading and swapping out 4x8 sheets? May be as simple as moving the gantry out of the way, just be sure the rails are long enough.

Thanks for your comments, much appreciated, and yes, moving the linear stage out of the way is one of the modifications the Maslow Lumber frame would have to have to use the linear sled, the aluminium frame would have this extra width built in.

Thanks agaoin for your interest and your points, they are very much appreciated.

Well first off I am a real fan of the Maslow CNC concept, low cost being the ultimate aim. However the world of CNC has changed a great deal since the first vision for the concept. GRBL electronics & software are now prevalent and low cost, certainly unless I am mistaken a good quality GRBL board can be purchased for around $30.00 and GRBL software in many circumstances is free. So the first aim is to introduce a lower cost electronics kit, coupled with a good software source.

The chain drive concept fitted the time period of the availability for hardware to create a low cost CNC machine, even at 8x4 area. Now there are available many other low cost ideas for linear rails. I am a fan of openbuilds and I do create my own CNC designs based on the openbuilds concept.

2nd aim, well, it is now entirely feasible to create a linear sled that comes in at around the same cost as the Maslow sled, but removes a lot of the restrictions placed on the machine and with respect to the Maslow concept introducing a more readily available source of hardware will encourage better design ideas that ultimately arm the user with an easier to use machine. Low cost is feasible also.

Addressing your other points, the Maslow II forum post was where my interest was piqued, this is a very similar design to a machine that I had made for a client only a few months ago, but this was a ‘traditional’ 8x4 horizontal machine. As part of the design and with a vertical machine in mind for the future I designed a linear sled for this client that I could use on smaller machines such as 4x4, the sled would fit both, even a 4x2 bench machine could utilise the sled, no further design required.

Then I saw the Maslow II post and I realised that a similar less equipped, less technical, lower cost linear sled could be designed to fit the Maslow timber frame quite easily. So this is were I joined the forum, with the intention of discussing this idea.

My client only wanted to cut 2D stuff, from ply and MDF, and she had no need for a typical Z axis, she was only ever going to cut a maximum of 1" (25mm) width materials, which means a much lower Z axis can be applied, possibly even one with a similar principal as the Maslow sled, one which slid on the surface of the material.

This brings me to your next point, that the Maslow sled touched the work area, so Z is always zero. I differ with your view that this is a better solution, for a couple of reasons, the first reason is that if the material to be cut is warped in any way then the edge of the sled will lift with the warp, thus lifting the tool at the same time no matter how small the warp and a large warp would tilt the sled and the tool. A heavy weight or spring based sled which pushed the Z onto the material may make some difference but that is the wrong approach, that is adding to the complexity of the machine needlessly. Finally, the Maslow sled is deliberately large diameter, so the base can slide over existing holes or pockets without dipping into them. A raised linear stage can be much, much smaller so again aids in lowering costs, so for this reason I dismissed the sled touching the bed idea.

Your next point about the vertical axis is correct, a standard stepper motor would indeed not be locked if there was a power cut so a belt or chain drive would not be an option unless a gear motor was used. Unless a lead screw was used, such as a TR12x3, the motor would not move, the sled would remain in position in the event of a power cut. Once again another hardware option is available, a low cost TR12x3 anti-backlash Delrin nut is available so expensive ball screws do not have to be used, and in time, yes the Delrin nut would wear, but can easily be replaced with another low cost nut. So on the Maslow Linear sled I will incorporate a vertical lead screw. Yes there may be issues with balance & weight of the sled head with gravity but a balanced pulley system could be introduced to combat this or a more powerful Nema 34 motor used, I would need to investigate further for this.

Rigidity would be built into the linear stage so aiding the Maslow sled to move in a smoother fashion giving more precise cuts, there would be no judder or friction kick steps on a raised linear stage.

Another point is that the Maslow timber frame is now a proven design, and with a wheel based (or other linear slide method) linear sled this timber frame can still be used. So for those who want to keep this low cost frame to create a large CNC cut area the frame can be slightly adjusted to incorporate a more standard linear XYZ stage. For new users the complexity of the Maslow V design sled and associated chain drive system would not then be an issue.

Of course I am open to having all of these points discussed further, that is the point of my post.

My ultimate main aim is to help the guys out there who want a low cost, large area, wall based CNC cutter to be able to build something that is based on a traditional XYZ system but is still affordable and maybe not as complex to get their heads around as the chain drive system

looks nice, but all that aluminum extrusion is going to be expensive to buy and very expensive to ship.

Okay, so if I understand correctly, your aims are:

  1. Leverage the wider open-source CNC community by moving to more “generic” firmware
  2. Support that move with a more “traditional” XYZ kinematic system.
  3. Achieve this with a machine that’s still vertical, to save on shop floor space.

Regarding point 1, there’s already a GRBL-based firmware that’s compatible with the Maslow chain system; that’s what MakerMade will use with their new M2. It’s interesting, as it allows usage of more common host software, not just the Maslow-specific GroundControl or WebControl. See here: https://github.com/ldocull/MaslowDue

I believe point 2 & 3 are really what’s difficult, as you’ll have issues with gravity that are absent in flat CNC machines. Maslow solves this by using worm-drive motors, but they are non-standard, hard to find, and expensive. They also require specific motor controllers, so you can’t use any off-the-shelf CNC controller.

Some solutions could be to use a worm-drive motor on the Y (vertical) axis, or a ball screw system. Either of these has the potential to make that axis much slower than the X axis. Another solution could be to use a counterweight, but that increases inertia a lot.

Having a gantry system does however liberate you from using gravity to keep the sled on the work surface, so you could go 100% vertical, and simply have the wall itself be your frame. Add a thick spoilboard screwed directly into the trusses, have the machine flatten itself by facing the whole board, and you have a very rigid base to go from that should be 100% true.

Last point, about warped surfaces. You’re right that significant warpage could lift the sled up, but that’s not what I was referring to. Most of the plywood sheets I’ve bought up to now have had a slight bow on the long side. The Maslow accommodates this very well, and I don’t have to screw down my work too much as a result. Having a gantry would force me to use a lot more screws to hold my work and make sure it’s flattened out.

if the workpiece is not flat and you ride on the surface of the workpiece, you
have the problem that the sled may tilt a tiny bit (but will still be
perpendicular to the average of the surface)

if you have a gantry and the material is not flat, you are cutting perpendicular
to the bed and the workpiece is not parallel to the bed so you have the same
angle problem, plus you have more of a depth problem.

If you are going to switch kinematics models, go for coreXY

consider a pulley and counterweight system to reduce the effect of gravity on
the sled.

the maslow orm gear motors are more expensive than NEMA 17 steppers used on
common 3d printers, but incredibly cheap compared to steppers with similar
torque capacity.

There are significant advantages to a closed loop system, especially when
cutting and fighting gravity.

I would love to see grbl extended to support DC motors and encoders rather than
open-loop steppers, there are a few forks floating around that do this, but
forks are not maintainable in the long run.

David Lang

Looks like a very neat project. Do you think by doing this you can get away from having an angled frame and mount it straight to a wall?

I am in a dilemma with my Maslow. It is pretty compact for what it is, but still takes up nearly all my free garage space. I am looking to get another CNC, this time horizontal, but would like to keep them both, so I am considering rebuilding the Maslow in a wall mount setup on the other side of the garage. It would be great if it was a low profile and totally fixed to the wall and did not need to swing out.

This is entirely possible, and will be an option, but you do need a slight angle, maybe 5 degrees, this is just to help stop the parts falling out of the material when cut, unless you add use tabs.

The Maslow sled could not have a lesser angle, well, it could but it would probably cause problems. The angle and the associated gravity and weight of the sled is what keeps the sled on the surface of the material. Lessen the angle and you lessen the gravity eventually getting to a point where the tool bit will just push the sled away from the material instead of cutting.

As long as you are able to utilise a timber frame to house the linear sled then this project will be a feasible alternative for you to consider.

Thanks for replying. The aluminum frame is an optional extra that will be offered, and yes, it is far more expensive than the wood frame but for those (me!) who prefer bolts and nuts to sawing wood its an option.

However, to address the cost issue, the linear sled I am designing would also be able to be mounted to a wooden frame. The frame would need some adjustments but not many. See attached one my beta concept designs, this design would be for new users again as an option because this sled design is at the higher cost end because it is fairly solid in comparison to a lighter sled design I have in mind.

The blue items are what need to be added and the red items are what can be taken away from the Maslow frame.

The blue extensions to the left allow the sled to be out of the way when loading materials on to the bed.

All comments are appreciated.

1 Like

Many thanks for your continued communication.

More or less, yes, those are the initial aims.

I was aware of the MaslowDue, but my hands are way up in surrender here, I know absolutely nothing about schematics, or electronic design, just reading this ( There are 3 places that must be cut and one place where 3.3V is patched over as shown here) puts me off. I know PCB designers who would quite happily produce a PCB but I would not have a clue what to send them to enable this. So, this is another reason for including an off the shelf GRBL board, for people like me that just want the board to work when I wire everything up. Wiring I’m OK with, schematics, no.

Points 2 & 3 is where I am carrying out most research, but a viable option is something like this: McMaster-Carr which I believe I found reference to on this forum but can’t find the original poster so apologies to that person up front.

Holding the work down is one of the constant problems with a horizontal CNC so I totally appreciate that the Maslow sled excels here.

Thanks David.

Yes, totally in agreement with this statement.

I don’t think coreXY is suitable for larger machines, this is IMO but I am open to comments on this. You should be able to see a beta design on this page showing a very early (adopted from an existing horizontal machine design) that demonstrates a proposed belt and lead screw system, but as Emile has pointed out, the vertical TR12 leadscrew would be much slower than the horizontal belt. However sometimes you have to accommodate these limitations.

I posted another option McMaster-Carr

Please bear in mind that I am in favour of those who prefer to use the Maslow motor to keep them. Stepper motors are my preferred choice and should be an option for the linear sled, that is what I am proposing, at the end of the day it is a choice to be made.

This is entirely possible, and will be an option, but you do need a slight angle, maybe 5 degrees, this is just to help stop the parts falling out of the material when cut, unless you add tabs.

you need tabs to prevent them from falling anyway

The Maslow sled could not have a lesser angle, well, it could but it would
probably cause problems. The angle and the associated gravity and weight of
the sled is what keeps the sled on the surface of the material. Lessen the
angle and you lessen the gravity eventually getting to a point where the tool
bit will just push the sled away from the material instead of cutting.

but if there is a gantry, you may be able to not worry about this. I say may
because it may also be that there is enough play in the rail attachments that
your Z will be a bit unreliable

I don’t think coreXY is suitable for larger machines, this is IMO but I am
open to comments on this. You should be able to see a beta design on this page
showing a very early (adopted from an existing horizontal machine design) that
demonstrates a proposed belt and lead screw system, but as Emile has pointed
out, the vertical TR12 leadscrew would be much slower than the horizontal
belt. However sometimes you have to accommodate these limitations.

I’m curious why you think it’s not suitable for large machines?

I posted another option McMaster-Carr

I cannot get this link to work

Please bear in mind that I am in favour of those who prefer to use the Maslow
motor to keep them. Stepper motors are my preferred choice and should be an
option for the linear sled, that is what I am proposing, at the end of the day
it is a choice to be made.

steppers without feedback are going to be significantly larger and more
expensive (and require larger, more expensive drivers), you aren’t going to be
able to use NEMA17 steppers with the normal plug-in drivers like the CNC shield
uses.

David Lang

I can assure you David, any play will not be associated with the rail attachments or the Z on the frame I have in mind. There may be some downwards Z movement when the sled is in the middle of the 1.5M sled due to sled weight but that will be negligible and only be on the lowest cost rail option (the sled will be designed so other more rigid rails can be used if needed) and even then, if only cutting 2D that will not matter. The openbuilds linear system is one of the most stable linear slides available for low cost machines, once properly set up there is no play. In fact when put to extreme forces such as on a traditional horizontal 8x4 system the X rail bends first, the wheel based gantries stay where they are. This project is for light machine cutting of ply, mdf or similar materials cutting at comparatively low speeds, so I am very confident this project will do the work it was designed for,

CoreXY is unnecessary, there would be an awful lot of belt on a large machine. There are simpler solutions that achieve a similar result IMO.

Link to constant force spring

I have worked with Nema motors for years, for this project there would be 1 of Nema 34 driving left to right motion, but I am inclined to consider a Nema 23 for this as well, 1 of Nema 23 driving the vertical motion and a Nema 17 or 23 driving the Z, they are more than powerful enough for this project. The Maslow linear sled will retain the current setup, there is no need to change that.

Hello,

hows the progress of your frame?