@dlang So can you give us a theoretical example and corresponding X value we would put in the Yaml?
Say for the first arm/belt. The center (or bottom or top or whatever is the best way to go by) of the arm/belt on the sled and it’s anchor point are the exact same height from the top of the workpiece, let’s say 3/4 thick workpiece or 18/19mm with the sled resting directly on it ready to cut.
Thanks again for everyones help clearing this up!!!
Geo
The downward angle of the default means there is more force being applied to hold the sled against the workpiece.
I’m not sure how much of an effect it has, but intuitively it feels that it is important to some degree.
Yes. Thank you @Carson_Barry from mentioning this.
I thought about this as well (especially for vertical frames) and thought about making all the anchors a certain mm below each arm to allow the same “tension/force” per arm.
NOT 4 different ones which is the default and doesn’t make sense doing it that way as it is a very easy adjustment IF we know exactly where/how it was done initially and how its used when changed.
This is easy to adjust, but the need to figure out how @bar is measuring and how the code interprets/use this. So staring from absolute zero and completely parallel belts/anchor points will clear up some confusion and we can go from there.
Thx!
Geo
I did a really poor job of getting my t nuts in the corner brackets, so mine actually vary in offset-from-frame by up to almost half an inch. I’ve made no adjustments to the settings in calibration, but it cuts with less issues than most others have mentioned so far. I’m guessing it compensated for this by thinking the anchors were in slightly different places, as the dimensions it thinks the frame has do not match the actual dimensions.
I’ve been meaning to fix this, but it’s been able to reproduce 0,0 and do multi cut jobs perfectly across power cycles without manually setting home location to a drilled hole, etc., so it’s hard to justify recalibration.
Wow. That’s great to know. Thank you!
Carson Barry wrote:
I’m not sure how much of an effect it has, but intuitively it feels that it is important to some degree.
with the earlier maslow, it worked best if the chains were at the balance point
of the sled and the belts parallel to the workpiece.
now, using a downcut bit would add problems, and the sled/router was heavier,
but if that is the problem, we would just see the sled raising from the
workpiece when you try to plunge
David Lang
Doesn’t the thickness of the workpiece matter a lot? If someone plans to be cutting 2” foam dont the anchors need to be raised by that amount right?
I am trying to find a way to float the anchors to adjust for material thickness to address the fact that I will be either cutting thin plastic or super thick foam
gazinux wrote:
Doesn’t the thickness of the workpiece matter a lot? If someone plans to be cutting 2” foam dont the anchors need to be raised by that amount right?
small angles don’t matter a lot, but the same delta in Z will matter more when
the Z difference is already larger
if you are 600mm from the corner (a pretty small frame)
18mm of height translates into a belt length error of 0.3mm
50mm of height translates into a belt length error of 2.1mm
68mm of height translates into a belt length error of 3.9mm (1.8mm more than 50mm)
100mm of height translates into a belt length error of 8.4mm
118mm of height translated into a belt length error of 11.7mm (3.3mm more than 100mm)
so it’s going to depend a lot on what your z height already is
with the Z height already varying by ~65mm, adding extra to this will affect
some belts more than others
@bar I think I’m talking myself into saying that on a stock maslow, we need to
have a way of entering the workpiece/spoilboard thickness. this could even
explain why people are seeing asymmetrical problems, when the angle to the top
belt gets steeper, the error introduced by workpiece thickness is higher and
belts are tighter.
I also think that your idea of ‘set Z=mecanical zero’ once is flawed, it’s just
too easy for the lead screws to get moved, and I think that the UI is confusing
between ‘set Z=mechanical zero’ and ‘set z=bit zero’
I would suggest that after a boot, the user gets prompted with “I think Z is at
###, accept or home Z” where homing Z goes to the mechanical stops.
how about calling the Z=mechanical zero “square Z axis” instead of ‘set Z zero’
I am trying to find a way to float the anchors to adjust for material thickness to address the fact that I will be either cutting thin plastic or super thick foam
look at the ‘super simple frame’ thread for my thinking. by moving the supports
to the same level as the anchors, you could move the entire frame out as needed
for whatever workpiece thickness. done just right, you could make it so you
stack scrap pieces of your workpiece/wasteboard under the frame corners to set
it to the right height. But even getting it close will take you back to the
“small angles don’t matter much” range
David Lang
Will packing out the spoil board with various foam thicknesses work?
Like @dlang said I don’t think small increments will matter much (1 but or 2 inches yes). IF we can figure out the correct zero Z anchor heights for each anchor/arm and correctly enter them into the code.
@dlang it’s interesting to see the math on accuracy due to Z though there is also the mechanical stability of the sled when accommodating super thick materials (with fixed anchors that don’t float up to meet the ideal depth for the specific workpiece).
For some jobs this will work, yes
Reading through this thread, I think it’s clear that a preferred means to run this machine would be to have the belts parallel with the work surface at all times, and intuitively this makes sense. This being said, this is challenging as most of us have “fixed” anchors in the Z-Axis relative to the frame by which they are attached. Furthermore, it seems that depending on the spoilboard height as well as your cutting material thickness, this can potentially create some additional issues and concerns.
Although somewhat involved, one solution I may experiment with is generating sets of anchors that have built in Z-Axis offsets that account for the spoilboard thickness plus standard material thickness’ (3/4”, 1/2”, 1/4”, etc). Depending on the material thickness I plan to use, I swap the anchors. Upon swapping the anchors, however, this would require a recalibration of the machine, but at this point, I would recalibrate WITH THE CUTTING MATERIAL INSTALLED ON TOP OF THE SPOIL BOARD. This would, in my opinion, ensure the most accurate calibration for that particular material that you are cutting with and anchors you are using in their relative positions. By performing this process, I would also believe it would reduce tipping of the sled as is an issue I saw while performing my first calibration and also what is being mentioned here. In theory this all seems to make sense, but unfortunately for us that have vertical/upright frames, we have to contend with gravity. This means that the more offset in our anchors, the more bending/deflection it will experience and that could be catastrophic from a load/stress perspective.
After just thinking about this, a threaded rod of 10mm diameter and some jam nuts might do the trick of acting as the anchor pin and account for the adjustability problem. It might also help with deflection as well as that is fairly thick and may withstand most of the deflection.
Mark Thomas wrote:
Reading through this thread, I think it’s clear that a preferred means to run
this machine would be to have the belts parallel with the work surface at all
times, and intuitively this makes sense. This being said, this is challenging
as most of us have “fixed” anchors in the Z-Axis relative to the frame by
which they are attached. Furthermore, it seems that depending on the
spoilboard height as well as your cutting material thickness, this can
potentially create some additional issues and concerns.Although somewhat involved, one solution I may experiment with is generating
sets of anchors that have built in Z-Axis offsets that account for the
spoilboard thickness plus standard material thickness’ (3/4”, 1/2”, 1/4”,
etc). Depending on the material thickness I plan to use, I swap the anchors.
Upon swapping the anchors, however, this would require a recalibration of the
machine, but at this point, I would recalibrate WITH THE CUTTING MATERIAL
INSTALLED ON TOP OF THE SPOIL BOARD. This would, in my opinion, ensure the
most accurate calibration for that particular material that you are cutting
with and anchors you are using in their relative positions. By performing this
process, I would also believe it would reduce tipping of the sled as is an
issue I saw while performing my first calibration and also what is being
mentioned here. In theory this all seems to make sense, but unfortunately for
us that have vertical/upright frames, we have to contend with gravity. This
means that the more offset in our anchors, the more bending/deflection it will
experience and that could be catastrophic from a load/stress perspective.
anchors flexing will do more damage than the Z offsets being off. But the maslow
calculates for the Z offset and Z axis movement. you just need to tell it what
those Z offsets are.
you should not have to recalibrate when you add thickness, just change the Z
offsets.
David Lang
Mark Thomas wrote:
After just thinking about this, a threaded rod of 10mm diameter and some jam
nuts might do the trick of acting as the anchor pin and account for the
adjustability problem. It might also help with deflection as well as that is
fairly thick and may withstand most of the deflection.
try it with the firmware version that included the test for frame flex. I think
you will be amazed at how much flex you get from the 10mm threaded rod.
David Lang