For a while now we have been talking about some of the limitations we have found with the current Maslow 4 design. We are still working on making the current design (perhaps with small modifications) work as well as possible and learning which problems are really problems and which are just minor annoyances or not important at all. Nevertheless, it’s good to document ideas for an improved Maslow 4, or perhaps a design so different it can’t be called a Maslow 4!
So if you have some ideas, major or minor, post them here!
I have been thinking about the interference issue where the vertical supports get in the way of the belts, creating the red zones in dlang’s diagrams. I can think of two solutions:
Increase the diameter of the arm assemblies and put the vertical supports inside the spools. This would open up the possibility of higher capacity spools to allow longer belts.
Make the vertical supports telescoping and attach them to the ends of the arms. They would no longer be attached directly to the sled and no longer at 90 degrees from each other, but would rotate with the arms, so maybe we would need 4 rather than 2 z-axis motors. Attaching this assembly onto the sled could get tricky. I think this idea could get very complex very quickly.
I also had some thoughts about the arms heights positions on the M4. This is for a vertical use only.
As it is, the top left anchor attached at the to the top level arm (the highest on the of all four arms) and the top right anchor attached to the second arm from bottom. With some usage, it seems that the worst direction due to tilting for the M4 is going from the center diagonally towards the TL Anchor point.
It is better going from center diagonally towards the TR Anchor and lastly, no problem going toward BL or BR anchored, gravity helping!
Would it be better if we could swap things around to have the TL anchor attach to the lower arm and motor (of course motors would have to line with there new anchor point?
So we would get:
TL anchor= bottom most arm
TR anchor=second from bottom
BL anchor=Third from Bottom
BR anchor=Fourth from bottom (or top arm)
That way, we would have less angle on the belt when pulling to go upward. Going down should not be a problem due to gravity
Just a thought, no sure if this can be adjusted on the actual version or if it should be review for futur version.
make the arms longer so that if you have anchors that make the belts level, you
can fully retract a belt with the belt sstill attached to an anchor
make the arm pivots attach to the sled so that Z movements don’t require belt
movement
move the steppers to the top of the machine, so homing Z moves you away from the
workpiece
move from route to spindle motor (better availability of 52 or 80mm standard
size spindles that use er11 or larger chucks
move electronics from being directly against spindle/router (reduces
interference, allows for more spindles/routers, avoids blocking cooling airflow)
much easier if you have a fixed column rather than guides.
we need to see long-term wear, but there may be a need for bearings between
spool and arm and/or arm and whatever it pivots on
for the Z movement, consider 3x lead screws connected via a belt to a single
stepper.
Add a PCB Mount subassembly that connects to the vertical supports, and not to the router. This would then completely open up router/spindle choices. So a bit like what dlang referred to. PS. I am absolutely going to do a 3D model for this and print it.
Better yet, make this easily demountable in some way, plus a quick(ish) release for the router / spindle. Then changing a bit is just → demount PCB support, release pressure on router/spindle → slide out router/spindle → change bit → put it all back together.
I really suspect this proximity is causing some of the current (pun intended) issues in stability. I wonder if we could shield the board from the motor its sitting on in some way. distance would help. maybe md8n’s model can do this (although the current cables to connect further would need to be longer)
@ronlawrence3 If you can get a measurement of the height from the top of the top router clamp to the top (highest point relatively) of the Dewalt (whatever model it is) router. Then I’ll know the minimum clearance needed by lots of people, and therefore also minimise any need for longer cables to the Z-steppers.
~ 75mm from the router top clamp to the bottom of the board (65 ish to top of router). Thsi will depend on how seated the router is in the clamp too, so ± 5m or more. There may be a little wiggle room up from that but not too much or the lowest motor cable BL won’t reach.
I’ve figured out that any PCB mount can’t be attached to the vertical supports because there’s just not enough cable length from the motors back to a fixed elevation PCB to do the job.
Therefore, I’m going back to plan A which was to attach them above the uprights with the linear bearings, so the whole arrangement rides up and down together, which leaves only the cables to the Z-Steppers to worry about.
That can be adjusted by users right now. Restack your arms, rearrange the z heights in the setup menu, and recalibrate. I think this would be a great experiment to see if it improves sled stability and accuracy of the calibration.
That can be adjusted by users right now. Restack your arms, rearrange the z
heights in the setup menu, and recalibrate. I think this would be a great
experiment to see if it improves sled stability and accuracy of the
calibration.
In theory, you should not even need to recalibrate.
An issue I’ve noticed now that I’ve had to change bits often:
The router can easily accidentally be rotated in the clamps while tightening the collet.
In other posts/threads, I’ve talked about being able to assemble the Maslow without a router in it, by having a hollow cylindrical core that everything actually sits against instead.
If the inside of this cylinder were threaded to match the router, it would be possible to mitigate the distance it can rotate. It may also be a large boon to be able to remove the router to swap bits, or even swap to a different tool altogether, and to be able to leave the Maslow mounted and ready for the next job.
Beyond this, it would facilitate having easily swappable “cores” with much less design investment required per concept since they could have a standardized attachment to the rest of the system. The only issue I see there is the positioning of the circuit board. I would feel more confident in swappable cores if it were opposite the dust collection system mounted on the vertical stand off on that side. This may also help distance it from sources of EM interference by moving it away from dust collection and the internal circuits of the router itself.
Another thing that came to mind today as I was disassembling was that I could loosen the clamps and remove and replace the router, so this is almost even feasible in the current configuration. It’s just that putting in a threaded tube without making wider spools/clamps would preclude the current recommended router. There would at least be an easy way to make any ‘cores’ compatible with both the original and the wider design, even if the original loses compatibility with the original router.
The other thing I noticed is that I almost didn’t have to detach the linear rod supports to take the spools out, but they just don’t fit between the vertical stand offs. Both assembly and dissassembly/maintenance would be easier if these spaces were wide enough for these to fit through.
In other posts/threads, I’ve talked about being able to assemble the Maslow
without a router in it, by having a hollow cylindrical core that everything
actually sits against instead.
strongly agree
The other thing I noticed is that I almost didn’t have to detach the linear
rod supports to take the spools out, but they just don’t fit between the
vertical stand offs. Both assembly and dissassembly/maintenance would be
easier if these spaces were wide enough for these to fit through.
I am wondering if we really need the supports above/around the lead screws. If
we can get away without them, then it becomes easier to assemble/disassemble
In mine, I broke one of them off, but haven’t run it enough to say that it’s not
a problem. (I actually plan to cut off the ears for the top clamp above the lead
screws to give the arms just a little more movement range)
Not as easily, because the potential screw holes are already in use.
I did come up with a ‘cap’ to go over the end caps for the linear rail supports, but I realised they wouldn’t work because it wouldn’t move with the whole router assembly, and would therefore need longer cables for many of the bits and pieces
(although the current cables to connect further would need to be longer)