the force doesn’t go away, but if you have say a 1/2" bolt with fender washers (~1" diameter) on each side and between the pieces that force gets spread across a much larger area, so it’s far less likely to give
1 Like
Sorry yes, 6 inches. The handle mounts have a protuberance even when removed, but it isn’t too much.
The handles could be mounted vertically right? There’s nothing magic about the rotational orientation of the router is there? Maybe that would get them out of the way more?
The sled is 18" in diameter? I’m sure I can look this up but does anyone here handily know how far apart the motors are and how high above the work space? I might make a more accurate mockup so we can test the tolerances without having to build multiple physical models.
18 inches is about right. Mine seems maybe .5 inch bigger, not sure if this is by mistake or what the design actually is.
From my GroundControl settings:
440 mm = motor height above top of work space
2900 mm = distance between motors
The motor spacing is around 3000mm and the height above the workspace is arouncd 370mm. YMMV 
In your (GREAT) chart, you called it 3-Bar Balanced.
Earlier in the thread weight distribution was mentioned as an issue with a trial run. Is “Balanced” actual “weight balanced”, or just due to the equal distribution?
All that said, if one of these is more resistant to unequal weight, that could be important, as drag from a vacuum hose or whatever could certainly skew things around during operation.
1 Like
I may have mentioned the unbalanced weight. It doesn’t matter for accuracy, but it may affect your range of motion. In all of these designs if built correctly, when the sled is stopped you should be able to rotate it freely and the bit shouldn’t move. I see this with my version and posted a video attempting to show this last night.
But at least in my design, and some of the other ones, the total range of motion of the arms get very close to the maximum required. So if like me, your sled’s CG causes it to sit 20 degrees rotated when sitting in the center of your work space, then you are likely to have an issue where the arms hit something when moving to one side. I easily fixed this by shifting the weights on my sled when it was sitting in the middle of my work space until it sat level. You can see how my bricks are shifted in an odd fashion to make it work.
But yes, a dragging power cord or vacuum cord, could cause the sled to rotate, but it doesn’t really cause accuracy errors, just range of motion. I haven’t seen any issues with this so far.
the motors are 2978.4mm apart in the stock design (9’9.25", not critical to the machine operation, it just needs to be measured accuratly)
They are 463mm (~18.25") above the workpiece
an accurate simulation is harder than you think, because the chains don’t pivot at the center of the motor shaft, they are wrapped around sprockets that stick out from the shafts, so the pivot point depends on the angle.
There is a simulation built into ground control now. I don’t think it’s been updated to support triangular kinematics, but if you could add that, it would greatly simplify testing (as it supports adding all sorts of errors into the machine)
Thanks, this helped (and sorry I seem to be only “talking to” you as Discourse reminds me :-))
With that said, if you just shift the points “closer” to the centerline (shift left for the right linkage and right for the left unit) wouldn’t you find a point where the red dots are correct?
I’m losing what the difference is which this sketch and the “3-Bar 45deg” in THE CHART 
Or if it doesn’t matter, I’ll hush.
Bar’s done this - there’s a drop-down at the top of the variables to select quad. or triangular kinematics.
Yes, when that happened the center of the sled would be on that gray dotted line, You would then have the 3-Bar 45 degree version.
This sketch was an idea for a different layout that I think we have all decided won’t work.
here is what I think we could do with a top pantograph
onshape parts for top pantograph
note the steep cutouts needed for chain clearance when in the bottom, near corner
onshape assembly for top pantograph
in the onshape assembly, you can grab parts and move them
I may need to fiddle with the lengths.
for those who have been doing the graphs, are there any dependencies between the horizontal and vertical bars? or am I right in thinking that the horizontal ones have to match each other and the vertical ones need to match the sled?
Would not have thought of that.
If this works, one option is the bulk purchase pre-cut components. This could also offer a guide for drilling the holes to get the necessary alignment.
I wonder if a Maslow with the “initial setup” sled would be accurate enough to cut something like that. Hole placement sounds like the only thing that is spacing critical, so …
Maybe even bootstrap yourself a bit - cut basic ones, then use them to cut “better ones”, etc.
I mean this isn’t a highly complicated shape or anything.
Honestly, I expect most of us could get awfully close with standard tools/skills (although I know we are also trying to support people without those tools/skills).
There was one person earlier who posted a link to a place that would laser-cut
the components cheaply. That would make the holes very accurate.
If we used a short chunk of rectangular metal tubing as the spacer, we could get
the holes cut in it to mount the arms to, which would make the only thing that
would need special care in assembly be to mount the spacer at the correct
distance from the bit and lined up with the bit (put a tiny 3/32 bit or a
v-grove bit in the router and drill where the other holes are would solve both
of these concerns)
I like to design this sort of thing so that you use a master link like
these
instead of needing some other bracket, or pin, etc.
1 Like
Something like that would be great from a design perspective. But it should be something easy to remove.
a maslow with bad calibration will cut different sizes of things in different places on the workpiece.
But if you have it cut one part, slide the plywood, and then cut the next part in the ‘same place’ (as far as the maslow knows), it will be very repeatable and very precise, just not accurate to any measurement we know
That is why I asked for the test above with the ‘center hole’ not actually centered. If that works like I think it will, we can define a build sequence that will work reliably,
The biggest problem I see with using wood for this is attaching the chain accurately. It will need to be able to pivot through almost 90 degrees relative to the pantograph bar it’s attached to.
also note that the ‘horizontal’ bars don’t have to be straight, they can be curved to fit around the router, as long as the holes are in the same place
Does which of the “designs” we’ve all been tossing around (say in the chart) change the amount of possible chain to bar angle? Seems like some of them wouldn’t be nearly that much (as the pantograph would be shifting at the same time), but maybe I’m missing a condition.
The video of @krkeegan above (obviously it didn’t go to full extremes), didn’t get much angle at all, even at the edge of the board.