The Z axis on the Bosch failing to maintain position while cutting, which IIRC was the problem @walter was having. On his router it looked like the positioning teeth on the rack on the router body had tapered edges that were not holding while it was cutting. Not sure if I have the details exact, it was a couple months and half a world away. I think it was the fixed base; I slipped up by wasting all my travel pictures on sheep and scenery and not Maslows
@gero different model Bosch than yours I believe. Bosch GOF 1600 CE router using Bosch MRF01 Router Fixed Base.
Details here:
The rest of the context of that sentence is:
“One idea I’ve been playing with to account for slop, droop, wear, flex or any other variable that might move the router bit away from the tip of the triangle that the chains create, is slotted holes for router attachment so you could move the router up or down on the sled to perfectly align the bit with your chains.”
I actually need to amend that thought process slightly. The sliding router idea does not account for even wear or slop in the balanced 45˚ version, but the good news is that accounting for it should be even easier than sliding the router - it can be done entirely in software when calibrating!
With a linkage whose sled mounted bars are only under tension, any consistent slop or wear at the joints should translate to a longer distance between the motors and the bit, this would put the router bit slightly below the tip of the mathematical (calculated) triangle that the chains form. The router bit will still be “at” the “ends” of the chains but the two chain-sides of the triangle will be longer than Maslow thinks they are. So the simple solution is to calibrate with new accurate measurements. This will bring the calculated triangle tip and the real triangle tip to the same place. Problem solved.
In a compression-tension linkage set up (like the top mounted version) two sled mounted bars are under tension and two are under compression. So any error actually compounds and moves the router bit down below the actual and calculated triangle tip. The problem with this error is that it actually skews the parallelogram that makes the linkages work, so my theory is that some of the error in this type of set up can be accounted for by physically moving the router up on the sled but I believe that it can’t be made perfect with this technique.
The moral of the story is that with a 45˚ linkage version slop in the 8 main joints can be accounted for and cancelled in the software because it only affects the distance between the motors and the router bit (the linkages remain a paralellogram). In the top-mounted linkage version slop can be reduced by moving the router but I don’t think it can be cancelled completely because the linkages are no longer a parallelogram.
Note: this is still a theory. I have not actually tested this with a working Maslow. I have tested adding intentional slop to both linkage designs and what I saw agrees with this theory. I may still learn more and I don’t claim to be the final authority.
Looking at the pictures of @walter the spindle GOF1600 is different then on the GOF2000 I have and I have not noticed any slipping of Z.
To tighten the release clamp needs a sensitive touch though. Fully tight is to much for the motor and to loose the spindle will turn but not move the router. I would estimate that I tighten it about 80-85%.
I was thinking that variable length horizontal arms for the top-mount would allow for corrective adjustments, and I’ve got a design that may work with plywood (this post and the next), just have to find time to try it out.
1 Like
Yes, totally. I believe somewhere on here we have discussed the idea of adjustable linkage bars to account for such things (or perhaps that only happened in my head?). Either way I think it could work. Might take some time to get the quirks worked out but a turnbuckle type adjustment should work if implemented properly. Or the slots like you showed. Or maybe even use a cam in a slot for easier fine tuning? Of course it would be handy to just not have slop in the first place… 
Or just use a design that it matters less with. The 45˚ version can be built with 13" wide bars if you want, in theory it can be made to fit around just about anything you could mount to the sled.
1 Like
k
I also had a look at the MakTec trimmer, I think it could work great on the z-axis that @StephenMcG made
What do you think?
nit-picking here.
This assumes that the wear is even in all directions. I suspect that that will
not be the case, but we won’t know until people run these things for a while.
There is also the asumption that arm deflection is not enough to matter (for all
pantograph approaches)
One take-away I have from the standard kinematics accuracy problems is that
anything that requires the person building the machine to measure something
accurately is a source of error, and anything adjustable requires that they be
able to measure the results.
2 Likes
Yes, that is exactly why I said “…any consistent slop…”
As you noted it will be hard to tell until someone wears stuff out but I assume that things will wear relatively evenly since the materials are the same and the forces are pretty balanced.
It is going to take a LOT of use to wear out 15-plys of birch and unfathomably more to wear out stainless steel!
As for flex I demonstrated less than ~0.2mm flex with 80+ pounds of force on each the wooden linkages. Thats far more force than they should ever see in real life. But you’re right, if flex is enough it throws everything off.
1 Like
Yes, that is exactly why I said “…any consistent slop…”
sorry, I missed that
As you noted it will be hard to tell until someone wears stuff out but I
assume that things will wear relatively evenly since the materials are the
same and the forces are pretty balanced.
I’m wondering about that. If someone commonly works near the center, than the
contact points in play there will wear more than the contact points that are in
play near the corners.
It is going to take a LOT of use to wear out 15-plys of birch and unfathomably more to wear out stainless steel!
I think it’s going to depend in part if the joints are against the threads of
bolts or against smooth surfaces. I agree with you if we are talking smooth
surfaces.
While you are creating your kits from very high quality plywood, I think it’s
worth pointing this out to folks cutting the parts themselves who may use lower
grade material.
As for flex I demonstrated less than ~0.2mm flex with 80+ pounds of force on
each the wooden linkages. Thats far more force than they should ever see in
real life. But you’re right, if flex is enough it throws everything off.
same note here about the material.
I am not trying to put down your efforts or kit, just trying to be clear the
advantages and drawbacks of every option (including the drawbacks of my
kit/design)
2 Likes
Oh totally! I’m sorry I didn’t intend to sound defensive. I realize I probably did sound that way though. I didn’t think about working in one area over another and how that would affect wear, I was only considering that the overall force on each joint should be consistent across all joints. You bring up a good point with that.
And yes, my early prototypes showed that threads against wood can dig in; but thankfully since nothing ever makes even half a revolution it sort of built its own threads instead of continuing to cut in.
I’m in no way trying to cut down any ideas or concepts; just trying to weigh pros and cons. Sorry for coming off that way. I truly appreciate all you’ve added to the linkage discussion/design!!
I was just trying to make clear that I am not either. I know I have a tendancy
to look like that, and someone else here said that I seemed to pick on ideas
only if posted by them, and so especially in this case where I posted a
competing design, I wanted to make clear that I was not trying to cut down any
ideas.
P.S. did you get the kit I shipped you?
2 Likes
I figured all linkages are evolved from the same idea (a pantograph), I’m not worried. The first (weighted) sled mounted linkage test I ever did was with a top mounted design, it was sort of my go-to because I like how far out of the way it gets, it was just much easier and faster for me to make the 45˚ version partially because I was using wood and didn’t have a good way to account for potential compounding errors. They all (including bearing, rollers, rings etc.) have their own advantages and disadvantages, I figure it’s best to get all the information possible into the hands of those who need it and allow them to make the best choice for their specific use. For some it will be a ring, for some it will be a linkage, whatever works best is best!
I did get the linkage kit! Thank you so much! I’m sorry I forgot to ever post an update about that, I kept meaning to message you! I’ve been too busy to do much of anything lately but I plan to put it together as soon as I can! The box was quite mangled and pieces were sticking out the holes in the box but miraculously nothing was actually missing (as far as I can tell)! The parts are beautiful and I can’t wait to try it out! Laser cutting in general is awesome but even more awesome when cutting metal! I love it!
1 Like
I have access to a pretty nice CNC router at work. Would it be worth while to precision machine a sled with accurately spaced mounting holes for the linkage kit and/or ring mounting? How much of an effect does hole placement for the linkages have on the overall tolerance stack up and resulting accuracy?
I’m willing to machine one out of wood or plastic and send it to someone for testing. I don’t have a kit yet, but could make a sled for myself also in advance of receiving a kit.
Would also be very nice to make it heavy enough to eliminate the bricks, although shipping becomes an issue. Anyone know the target weight of the sled + bricks?
I have access to a pretty nice CNC router at work. Would it be worth while to
precision machine a sled with accurately spaced mounting holes for the linkage
kit and/or ring mounting? How much of an effect does hole placement for the
linkages have on the overall tolerance stack up and resulting accuracy?
the mounting holes need to match the arms, but that can be done by simply
putting an arm on the center of the bit and drilling through the arm to make the
holes in the sled. (although you have to watch out for damaging the wood arms if
you do this)
Would also be very nice to make it heavy enough to eliminate the bricks,
although shipping becomes an issue. Anyone know the target weight of the sled
- bricks?
Sled + bricks + router is in the 20-30 pound range.
David Lang
1 Like
It’s just been revealed (science!) that sheep can recognise faces. If anyone ever teaches them sign language, there will be more than a few nervous Australians 
2 Likes
Sheep learn to associate faces with whoever has the feed bucket, or bottle for lambs. We raised them for almost 30 years
2 Likes
FWW - I teach all my dogs by hand signals. A) It means they have to keep looking at me while working. B) I’d never need to yell over someone else.
1 Like
I have their trimmer and their 1/2” router, which is built pretty well and well priced. The trimmer is light-duty only.
1 Like