I have a sad, embarrassing thing to admit. I haven’t made a single thing since cutting out my permanent frame. You know why? It’s not because I don’t have ideas or dreams or plans. It’s because as I watched the final frame pieces being made with the temporary sled and I became scared of the “wobble”. That is, the router tilting back and forth based on where it’s at on the plywood. I keep telling myself the temporary sled is far inferior to the final sled but I still can’t bring myself to risk ruining a piece of plywood.
Because of that I keep holding out for a nice solution to that sled wobble. @bar I’m super excited about the ring design, and I love what others are doing with laser cut arms but it seems like the whole community is still working through finding a commonly held best practices for higher accuracy.
I hate that I’m doing this, but I just cannot get past the frustration.
So I need your help. How can I contribute to the community and work to solve this “wobble” problem?
I want to make sure that I better understand. “wobble” is the fact that the sled tilts to a different angle at different points on the work area so at the left side of the sheet the sled is tilted to the right and vice versa? This tilt moves the position of the router bit and can make a ‘straight’ line from one side of the sheet to the other not actually straight.
In theory this tilt is calculated by the internal math (or ‘kinematics’) and it is compensated for, however we have found that there are many variables in the equations which do this compensation so it is very difficult to calibrate. An error in any measurement can effect the straightness of a long straight cut in ways that can be hard to use to determine which variable needs to be adjusted.
The community solution has been (exactly as you said) to move towards a system where the chains virtually meet at the router bit. In this kind of system the tilt of the sled doesn’t matter because the sled tilts around the router bit so as the sled tilts the router bit does not move. Also, because the sled is free to rotate in this way, it very rarely tilts at all. The calibration process now asks which system you are using and tells the firmware to use the appropriate math.
I agree that the whole community is still working through finding a commonly held best practice, and I expect that we will be for quite a while. While traditional CNC machines have been around for decades, we’re pioneering our approach every day and improvements are truly being made on a daily basis. At this moment I would say that the best practice would be to use one of the linkage kits available. As much as possible I try to make all of the improvements in the software which we can all easily update, but in this case physically changing the way the chains attach to the sled makes a big difference and is absolutely worth it.
@bar, yes, I’m referring to sled tilt when I say “wobble” - mostly because I couldn’t think of a better term. But “sled tilt” is a much better description of the problem.
Thanks for the recommendation on ordering a current linkage kit type solution. I think I’ll go that route, and then see if there’s any way I can help with feedback on future revisions of the design.
I’ve been tempted to try the stainless steel ring you used initially since I live close to where they’re shipped from but I doubt I could account for all the variables. For me it’s probably better if I stuck with something that others have already proven to be effective.
I’m sorry I can’t give you a 100% tried and tested best practices answer, but down the road when we have that answer you’ll get to look back on it and know that your feedback is how we found it
I’m getting excited again! Just moments after my last reply I ordered a linkage kit and will set to installing that, then cutting all the fun ideas I’ve collected in the past year. Wahoo!
@bar, yes, I’m referring to sled tilt when I say “wobble” - mostly because I couldn’t think of a better term. But “sled tilt” is a much better description of the problem.
That sled tilt is a feature of the original design, not a drawback or something
to be afraid of. unfortunantly, it means that it’s hard to measure everything
accurately enough
Thanks for the recommendation on ordering a current linkage kit type solution.
I think I’ll go that route, and then see if there’s any way I can help with
feedback on future revisions of the design.
I’ve been tempted to try the stainless steel ring you used initially since I
live close to where they’re shipped from but I doubt I could account for all
the variables. For me it’s probably better if I stuck with something that
others have already proven to be effective.
The nice thing about the ring/linkage approach is that you eliminate almost all
the variables. you are left with two
make sure the router is centered in the ring (or movement of the linkages)
what is the distance the ring/linkages add to the effective length of the
chain (and this is getting worked out in the calibration step)
The other comment I’d make is that unless you are having really bad problems with accuracy and you really need high precision for your projects, get building something that may be less than perfect. You might find that its good enough or at least learn a little about how to get the best out of your Maslow. Remember that perfect is the enemy of the good.
