The User Guide indicates that the calibration should be done with no router bit and the router chuck nearest to the work surface. I don’t recall if it’s covered whether the calibration is sensitive to factors like workpiece thickness or spoil board thickness and whether those elements should be present during calibration - or if calibration needs to be repeated if those factors would change.
What of these do we need to consider, please?
This is a great question!
The answer is that we don’t really know yet. We just added variables to control for spoil board and work thickness last week so more testing is needed to figure out if they make enough difference to matter.
At the moment I’d say don’t worry about them, but it is an excellent question!
Our layout is like this(Although vertical):
The anchor-points are at the same height as our backplate, and this is also our “calibration-layer”.
When calibrating, the collet barely touches the backplate.
After calibration, we measure from the bottom of the belts at the anchor-points to the bottom of the arms at the router. Still without any router-bit and with the Z-axis all the way down.
Let’s say the measurements in mm for each arm are: tlZ-110, trZ-70, blZ-50, brZ-90.
We do not, however, add any measures for the routerbit. Our last cuts have been done with a bit that adds more than 20mm to the height of the arms when the bit touches the workpiece Z-Home.
EDIT: The collet on our 500W spindle can go further down than the Dewalt. There’s between 15 and 20mm before the Z-axis stops because of the motors.
This will be accounted for. The current height of the z-axis above where z-stop (not z home) was last set will tell the machine to account for that.
There was a bug in how “apply tension” was working which was messing with z-stop, there will be a fix for that in the next release, but generally the length of the router bit is accounted for.
Ah, of course! That’s kind of obvious really. My brain just couldn’t think it out itself 
Thanks @bar
I think a more common term for the bit in cnc machining world is tool length offset?
Dano
length offset is used when you have tools that are pre-set in a tool holder, so
the Z doesn’t change, the machine just changes the effective Z when you change
tools.
only the high-end machines have this (the ones with tool changers), the lower
end machines just touch off the tool to find a Z-zero for each tool, which is
the approach that the maslow does (either manually, or with a probe routine)
David Lang
I am interested in this subject as well. I have a project for which I want to cut some curved profiles out of the edge of a board. The maximum Z-axis travel is just about 2” and I want the final parts to be 1.75”.
My first try was kinda lame. I tried to cut the edge of a 2x4 with two flanking 2x4’s blocked out 8” on either side to make a surface big enough for the tool to run. (setup picture below) The surface was 4 inches higher than the floor-mounted belt brackets. It seemed that the belts were pulling down so hard it added friction to the surface and moved unevenly. The lack of a continuous platform to slide on probably affected the run as well. The shaking and jumping of the tool and the wobble of the insufficiently supported board broke the bit pretty quickly.
My next iteration will replace the printed plastic floor mounts with four sturdy steel posts that have adjustable height brackets for the belt connections. I’ll set up true platforms with a plywood surface to flank the work piece and clamp it in a more stable fashion.
The belts will then be at a more appropriate level to the work. My assumption is that it’s best to set all the brackets to the same height above floor level, as the calibration discussions I have seen mentioned that the various heights of the four spools are included in the calibration calculations. I also assume that the calibration is based upon brackets that are placed at the level of the base/table, or possibly at the level of the spoil board. So for this job, it seems to me I should set the brackets at about ½ inch below or maybe exactly even with the cutting surface.
It’s going to take me a while to get this next phase done. Gotta pick up steel and fabricate the new mounting system. But I’ll post again here in hopes we can keep up a discussion about variable belt bracket heights and best practices.
Eric Rivedal wrote:
My assumption is that it¢s best to set all the brackets to the same height
above floor level, as the calibration discussions I have seen mentioned that
the various heights of the four spools are included in the calibration
calculations. I also assume that the calibration is based upon brackets that
are placed at the level of the base/table, or possibly at the level of the
spoil board. So for this job, it seems to me I should set the brackets at
about ½ inch below or maybe exactly even with the cutting surface.
The calibration assumes that there is no wasteboard or workpiece, and that the
anchors are at the level of the base of the sled. These are the Z offset numbers
(TLZ,TRZ,BLZ,BRZ) in the maslow.yaml file.
As of the 1.09 release, there is now the ability to specify a wasteboard and
workpiece height, so that you can now just enter those numbers rather than
having the change the four offst values to account for yout workpiece thickness.
Having such a thick workpiece without the system knowing it will mean that the
belts don’t reach as far as the system thinks they should and they are very
tight, causing problems like you are describing.
The closer the belts are to level (which would be each anchor at a different
height), the less significant any Z error is (i.e. if your belt would otherwise
be level, a 1"/25mm error in the height matters far less than if your belt was
expected to have a 4"/100mm Z offset and instead has a 5"/125mm height instead)
the one thing to watch out for is that your anchors don’t move. The belts are
under a lot of tension and haivng anchors that match your arm heights, but move
under tension is worse than having anchors with a significant offset that is
properly specified by the Z offsets and wasteboard/workpiece values.
so anchors that match your arm heights anchord to a wall are really good. just
using 3/8"/10mm allthread with nuts to raise the anchor points is really bad
because of how much that allthread will flex under load.
Right now, opinion varies of the value of having anchors below the level of the
arms, some people like the downward pull on the sled, others don’t see value in
it. (I’m in the latter camp)
I have a 3d printable bolt-down anchor design that lets you specify the height
up to the first arm. The first set that I designed and printed were too
flexible, I just got the 2nd set printed and have not yet had a chance to test
them (weather in Phoenix is not good for trying to work outside, 100+ degree
weather and thunderstorms this week)
David Lang
I’m planning either 3 inch or 4 inch square steel tube welded to a 6 inch square 1/4 inch base with four anchors each into the concrete. Maximum 12 inch height, just for flexibility’s sake. Haven’t made a final decision on whether to spend the extra money on the larger tubing or if 3 inch will be sufficiently immovable. More to come on that. Thanks for the feedback - will plan on setting brackets to level of workpiece and calibrate from there.
Eric
3" should be pleanty stiff.
David Lang
@dlang Would you mind attaching a .f3d or other CAD file of the printable anchors once tested? Thanks!
