Just setup my maslow 4 and made my first successful part. Its very cool! I have a vertical frame, and having to detach, retract all, extend all, reattach, tension on each power cycle isn’t fun for me. I was thinking of ways we could improve the belt homing process for vertical stands and I came up with an idea that I want to see what the community thought.
My idea is this: For the initial setup we continue as normal. Retract all, extend all, attach, calibrate. Once the calibration is done, the maslow knows how much belt is let out on each axis.
We then move the maslow to the top left corner of the cutting area, almost nearly all the way and attach a 3d printed stop on the top left belt. The maslow would then tighten against this stop and record the length remaining, this remaining belt length would be stored as the top left offset. We do the the same process on the 3 remaining belts. At the end of the process, each belt has a physical stop and the firmware has a offset stored for each stop.
Next time we restart maslow, to find our belt lengths, we start with the top left and retract until we hit the stop, we then add the offset and we have the TL belt length. Note while we are doing this we will need to monitor the other belts and slack them if they get too tight, similar to how extend all will allow you to pull out on them, but stop if you stop pulling.
Before I start making 3d prints and forking code, I was curious if anyone had thought of a better way?
I was just going to put 2 holes in the sled to lock it against the spoilboard with screws at 0,0 as a shutdown procedure.
In that instance, a button for ‘save location and power off’ and another for assuring the unit that it didn’t move while off that you could be prompted with when it boots would do the trick just fine.
I don’t know the systems well enough to modify the firmware for the board or the backend for the UI or I would have tried implementing what I described.
Yes, we need to add the ability to have a per-belt added length, and then we
could do what you are saying (and support other things, like anchors further
away that the belts can reach)
it’s been discussed, but other things have taken priority so far.
Brand new M4 user here, but I was thinking about ways to avoid the detach/retract/extend/attach every time I go to use my new toy. Mainly because I am lazy!
I’m not familiar with the mathematical magic that makes this machine work, but I was wondering if the machine could calibrate its own position in the same way it calibrates the anchor position as a startup procedure? Especially now that you don’t even need to set the frame size with the new firmware. Move, measure change in all belt lengths, do this a few times, then calculate where it must be subject to those measurements? It could even use the last known position to speed things up perhaps?
Just a thought I had, but understand that this might not be precise enough even if it was possible.
Thanks to everyone who’s made this machine possible, it’s so awesome to have a 4x8’ CNC router that stashes away in a box in the corner.
You are brilliant because YES…in fact I just added that feature in 0.88
Right now I’m working towards being able to make it so that the machine can load it’s position, but it’s involving a LOT of code clean up in the background to make it all play nice.
It won’t be a perfect system though because we still rely on the belts not having moved (more than one full rotation) since the system was last powered on. Basically if the machine was say taken apart and put back together since the last time it knew where it was, it won’t be able to determine it’s location accurately because the measurements will all be off…fortunately we should be able to detect that automatically and stop.
Sweeeeeet! As a person who’s always against the clock (children!) this is great news. The magic of software that I’ll never understand, I’m glad others do.
Looking forward to testing it out, happy to be a guinea pig if that’s of any use.
