Size limits on the frame

Is there a size minimum on the frame? I have a rather small on the sacrificial board 2 ft x 4 ft. The frame is 4 x 6 . Every time I put a .75 board under it I start having faults. I also have the same fault when I put the z axis all the way up for a dry run. Just seeing if anyone else has seen the same issue. Here is the serial log. I have ran calibration multiple times with the z axis bottomed out and it will run beautifully.
Maslow-serial.log (5.9 KB)

1 Like

Hmmm is this on a Maslow 4 or 4.1?

Generally you are on the smaller side of what can work, but it shouldn’t be impossible.

How does the tension in the belts feel?

Ross O’Connor wrote:

Is there a size minimum on the frame?

yes, something like 400mm square

I have a rather small on the sacrificial board 2 ft x 4 ft. The frame is 4 x 6
. Every time I put a .75 board under it I start having faults. I also have the
same fault when I put the z axis all the way up for a dry run. Just seeing if
anyone else has seen the same issue. Here is the serial log. I have ran
calibration multiple times with the z axis bottomed out and it will run
beautifully.

take a look at the frame calculator http://lang.hm/maslow/maslow4_frame.html and
see how far out of the green area you are going.

you should do the calibration with a board the thickness of your workpiece in
place (and check that the Z offset values in the maslow.yaml file are sane)

with such a small difference between your workpiece and your frame, you may be
running into problems with the angle of the belts (in the Z direction) and may
benefit from moving the anchors to where they are closer to level with the arms.

David Lang

1 Like

@bar it is 4.1. The belts are tight and have a steep angle. I had increase the extension by 10mm to even get them on. Which on a side note, we really need a way to manually over ride and release the tension if there is a problem. It is rather difficult to get them off when full tension is applied. I am guessing this is a safety feature to keep it from dropping. Maybe a release button that lets out a mm or so each time it is pressed.

Overall this would be awesome information to share on the frame building page. I was under the impression it was a lot more open to sizes.

@dlang your calculator helped. It took me some time to understand it. If it is going to the public could you do a quick walkthrough on a video? Also, I get a bad gateway when I click the link. I think this is because I am on my work laptop currently.

I am thinking of putting longer bolts in with a double nut to hold the belts more in line to keep them closer to 90 degrees based on what you said

The “Release Tension” button should do exactly that :grinning_face:

The steep angle seems somewhat concerning. How steep are we talking?

Understood but it would not do anything for me in this state. Fan didnt kick on or try to move when I pressed the button. I figured it was a fail safe of some sort.

I am really not sure. My guess is 30 to 45 degrees. It was almost at the top of the z because I had a router bit in and was doing a dry run. I could set it up and measure it this weekend if it would help.

I’m not completely groking the tool…
How bad is the distortion in the peach and yellow colored areas?

Ross O’Connor wrote:

I am really not sure. My guess is 30 to 45 degrees. It was almost at the top
of the z because I had a router bit in and was doing a dry run. I could set it
up and measure it this weekend if it would help.

that sounds like enough to be a problem (it will flex the arm down and add
friction making it harder for the arm to pivot)

see what you can do to raise the anchor a bit (and change the Z offset to match
what you change)

david Lang

Dave Covert wrote:

I’m not completely groking the tool…
How bad is the distortion in the peach and yellow colored areas?

each band in those areas is 1 degree of error on the belt. how much that changes
the length will depend on how long the belt is. just above the graph there is a
table that shows what the effective belt length would be if it’s supposed to be
1000mm

David Lang