bday wrote:
from my understanding of the fitness number a high one is great. Most have 1
or under so if you are having issues that¢s not it.
fitness is roughly 1/(estimated error) so the higher the number the better.
a fitness of 1 is roughly saying that it thinks the anchors are accurate to ~1mm
a fitness of 4 is roughly saying that it thinks the anchors are accurate to
~1/4mm
David Lang
eamonsca wrote:
Wow, okay I thought that values below 1.0 were ideal. Thanks for clearing this
up. I am still experiencing the belt tug of war that was happening before even
after recalibrating. I think my tool path may be too tight for the small
lettering I am trying to cut.
can you make a video of the problem? my guess is that the machine is flexing
rather than sliding.
These 3d printed towers make things a LOT stiffer.
(these can be set to go higher if your printer allows and you want to mount a
handle/crossbar on top to make it more rigid)
don’t forget to pause to insert nuts at about 10mm)
David Lang
Thanks,
I will definitely print the towers. I think that the machine has been flexing a bit.
Thank you to everyone who has responded. I was able to successfully cut the sign mentioned earlier on out of .25” sintra with fantastic results. I modified my original toolpath to reduce abrupt changes in direction, and this seemed to help. I used the center of the work area for this first cut.
Settings used for these cuts:
I later cut out a large curve out of 3/4” plywood, which went relatively well, although the outer diameter was 1/8” off on each side, and inner diameter was 1/16” off on each side. This still seemed promising.
I then went on to test various tolerances for interlocking slots, depth consistency, and cutting closer to anchor points. The cut was going well until the Maslow moved appx. 8” from the upper right anchor point. I noticed that the motor gear closest to this anchor point was making a clicking noise, and not moving smoothly along with other belts. The resulting cut was a wavy line. I stopped the program to avoid any kind of damage, and to try and figure out what may be causing this.
One thing I noticed was that the angle of belts when nearing anchor points became relatively steep, and I thought that this may be causing the affected belt to slip or catch within the spool. Does anyone have an idea why this could be happening?
@bar @dlang I’m not sure if you see updates in this string. Would it be more effective to create a separate forum post for engagement?
What infill density would you recommend for this print?
eamonsca wrote:
@bar @dlang I¢m not sure if you see updates in this string. Would it be more
effective to create a separate forum post for better engagement?
probably, it’s good to have a separate thread for a new problem.
the clicking tends to be that there is a lot of tension on the belt and it’s
catching on the belt guard.
David Lang
eamonsca wrote:
What infill density would you recommend for this print?
print it with lots of perimeters and top/bottom layers (3-4mm thick) and beyond
that it doesn’t really matter.
David Lang
David Lang wrote:
print it with lots of perimeters and top/bottom layers (3-4mm thick) and
beyond that it doesn’t really matter.
and remember to put a pause in to insert the nuts at about 10mm
David Lang
eamonsca wrote:
Thank you to everyone who has responded. I was able to successfully cut the sign mentioned earlier on out of .25” sintra with fantastic results. I modified my original toolpath to reduce abrupt changes in direction, and this seemed to help. I used the center of the work area for this first cut.
Settings used for these cuts:
- Calibration force - 850
- Retraction force - 850
- Cutting speed - 762mm /min
- spindle speed - 2/6 on dewalt router, appx. 18,000 rpm
I later cut out a large curve out of 3/4” plywood, which went relatively well,
although the outer diameter was 1/8” off on each side, and inner diameter was
1/16” off on each side. This still seemed promising.
it looks like you have anchors at different heights, I assume you set the Z
offsets appropriately when you did the calibration, correct?
then you added your 3/4 plywood, did you change the workpiece thickness? if not,
give that a try (adding height from the anchors and not telling the machine
about it will make the belts tighter)
David Lang
This sounds like it is pushing the limits on how close to the anchor points the machine can comfortably cut. Check out this tool for an estimate of how bit the working area reasonably is for a given set of anchor points
In the picture above, I did have just one anchor point raised by 1.5”. This was a change I made to prevent the belt coming from the highest arm from twisting when cutting close to the anchor point. I later bumped all anchors up by 1.5” because the belts were scraping on the workpiece near the furthest anchor from the cut point. It makes total sense that this could throw off belt tension.
I did not change the workpiece thickness, spoil board parameters, or Z offsets.
Should these values be input prior to calibration?
What distance is measured to determine the Z offset value for anchor points?
I found this tool after making the post. Very helpful, thanks!
I think I found all necessary info here, thanks!
eamonsca wrote:
In the picture above, I did have just one anchor point raised by 1.5”. This was a change I made to prevent the belt coming from the highest arm from twisting when cutting close to the anchor point. I later bumped all anchors up by 1.5” because the belts were scraping on the workpiece near the furthest anchor from the cut point. It makes total sense that this could throw off belt tension.
I did not change the workpiece thickness, spoil board parameters, or Z offsets.
Should these values be input prior to calibration?
yes. Also update to the current version of firmware (the fact that you show
retraction and calibration current limits says you are a couple versions back)
What distance is measured to determine the Z offset value for anchor points?
This is the distance from the height of the belt end when it’s fully retracted
down to the height when it’s at the anchor. The flatter the belts are, the less
sensitive they are to errors in this value.
to make your life easier, the default values assume that the anchors are at the
height of the bottom of the sled, so you can take a straightedge (long level,
ruler etc), lay it on the wasteboard, and then extend it out to the anchor and
measure the distance from the bottom of that to where the anchor will sit. Then
add/subtract that value to the default Z offset for that arm.
note that these values are relative to the plane that the spoilboard is on, it
doesn’t matter of the anchors are level per a gravity level, all that matters is
the plane of the spoilboard (which is why the machine can operate angled against
the wall or flat on the ground, or on a sloped driveway)
If your anchors are all at the same height, then instead of measuring them
individually and changing the four Z offsets, you can leave the defaults in and
use the spoilboard thickness value to adjust them all (positive or negative)
Then when you go to cut, you can set the workpiece thickness value to adjust for
what you are cutting (and if it’s close, and especially if your anchors are
elevated so the belts are close to level, you don’t need to change it for
slightly different thickness values)
David Lang
After getting all software up to date. I input anchor height values and spoil board thickness prior to calibration. After calibration, Z values for anchor points reverted back to defaults. If I change these values after calibration is that still effective?
The Firmware update has already been greatly helpful for tensioning / retracting belts consistently.
I wouldn’t expect find anchor points to change those z settings, however it could be a bug.
Be sure you are updating the anchor height values and then pressing save button up are top of setting screen. If not pressing save the values are not actually updated.
Maybe save and close the value screen and then go back in to confirm they were saved as expected.
Then try running find anchors again and confirm results.
Dano
