Thank you for this topic and the tool. Hoping to build a minimal self supporting portable 6 x 10 frame that can reasonably cut 4’ x 8’ plywood, but still disassemble reasonably quick and be thrown into back of my truck’s 6’ bed.
Thank you!
Great. Was hoping topic would go here… Personally would like to understand how accuracy varies/diminishes at the extremes.
Cool, so, if we can calc/estimate corner accuracy. Then…
. Through either compensation of upright deflected belt length, and/or varying feed rate in the corners, e.g. gcode/CAM post processor that applies gradient feedrate/speed dampening as XY approaches corners.
My concern about this is that it’s going to depend on exactly the point where the motors hit the clamps, small errors in this angle will affect how much the error is.
@arjenschoneveld could you re-plot that with small differences in the angle (say 129 and 131 degrees) and see if it’s a noticable difference in the error?
Here is the same graph, but now on the horizontal axis there is the angle between an arm and the rest of the belt towards the anchor, 0 to 10 degrees.
(I constructed it with 2 arms of 108mm length each, with an angle of 130 degrees to each other, and from the endpoints of the arms the rest of the belts to the anchors. angles between arms and belts equal)
Underneath I gave roughly the corresponding angles you use in the checker. Roughly, because the cornerpoint has deviated.
As you can see, the difference in angle between 130 and 128 degrees does not give a noticable error, at 122 degrees the error is about 0.5mm.
These measurements are all for a frame size of 3048x2440, I expect that other frame sizes have different numbers, but I don’t know how much, don’t expect it to be much.
Thanks, that says that we should be able to (eventually) add this to the
calculations as the exact angle isn’t the critical factor. We probably won’t be
able to completely cancel it out, but should be able to get fairly close.
My head hurts trying to find ideal frame size for cutting a whole 4’x8’ sheet. lol
I have a M4 coming and need to get busy with frame design.
Accuracy, Percussion and No Collusion taking precedent over space savings. I guess we also need to take belt length into consideration.
Am relatively new here, but given what you seem to be after, building 8’ x 12’ frame for cutting 4’ x 8’ full sheets seems to be the common, proven, known to work path.
Am space constrained so going to try 6’ x 10’, hope for the best this/next week, and, hope for future firmware and/or part improvements that can compensate for belt deflection (around uprights) and forces near the corners.
I’m not aware of a consensus, but I can tell you my plans:
I have a limitation in height in my shop, I will get my old hinging Maslow1 frame at about 20 degrees, that should give me 2600 high and 3800 wide, with a belt length of 12’ (which fits well on the spools).
This gives me an angle between the upper belts of 140 degrees, in the middle of the topside. As the M4 is hanging from them, I wanted no wider angle.
It will also give me a workarea of 2440x1220 if I use a minimal angle of 122 degrees for 2 opposite arms. That will give me a theoretical error of max half a mil in the corners of the workarea. Calibration area of 2000x1000 then is within the “safe” zone.
Hi mreih,
there are a range of answers to your question
First of all, it depends on how acurate you want your cuts to be, what kind of projects you are going to do. Large letters or digits or wavy contours do not need the same accuracy as large parts for cabinettry that have to fit together tightly.
In my case, I want to be able to make cabinettry, so I want the machine to be as accurate as it can be.
In order to determine what a good framesize would be, I started looking at the limits of the hardware, which are: the motor arms can rotate a 50 degree angle each. This means that there are zones on the bed where two arms collide with the uprights. That means that an error is introduced in those zones. @dlang thought exactly the same, and made his wonderfull framesize checker where you can see very clearly the safe zone and the zones with possible error. In order to determine what that error would be I later constructed it in CAD, which revealed that it is not a hard boundary but more of a gradient: at first the error is very small , if you go 160mm outside the safe zone the error is 0.5mm (which i find acceptable), if you go 400mm outside the safe zone the error is 2.5mm.
The 0.5mm error transtates into a 122 degree minimum angle between two opposite motor arms (in the CAD drawing I made, with the default framesize)
If you use this value in @dlangs framesizechecker, you can determine a good size for yourself.
And this leads to point two: how big is the space you have available for the machine?
In my shop I have a height restriction, so my height will be 2600mm, which then translates into 3800mm wide. You can go less wide, but then you should go higher.
And let me stress that this is theoretical. I don’t think I’ve seen anyone to actually measure their results, and my machine is not up and running yet.
I have a Large space on a garage floor. I’m wondering what would be the optimal size given that I probably have an excess of space. I think the limiting factor will be the length of the belts? Is it recommended to trim the belts when you get to your final size? Also wondering if having long belts will affect anything : belt stretching, weight of it hanging.
I have a Large space on a garage floor. I’m wondering what would be the
optimal size given that I probably have an excess of space. I think the
limiting factor will be the length of the belts?
correct.
Is it recommended to trim the belts when you get to your final size?
it’s acceptable to trim the belts, but there is no reason to do so. The belt
length is measured purely on what’s fed out so it doesn’t matter how much is
left on the spool.
Also wondering if having long belts will affect anything : belt stretching,
weight of it hanging.
we don’t know. not enough testing yet. Since concrete anchors are relativly
cheap. I would go large and you can add some more to go smaller if you want.
I am about 218.5, and 217.5 at the angles - so pretty close to square - I think it will be good enough.
Big but I’m interested to try it out. At the corners of a 4x8 sheet, I’ll be near the edge of the belt length.
I’ll update when I get anchors in.
Please do.
@bar, your frame calculator gives a different result than mine, mine shows the
top center as being good as far as the arms go, you show that it gets into the
yellow there.
Took all of an hour and a half to lay out with a laser level.
With the studs in it is 0.5 inch out of square which I think I can live with.
I will probably grind the top of the threaded rod down so that I can seat it a little better with a wrench but even just by hand there’s little to no play. Now I’m thinking about maybe a delrin bushing to go over the threaded rod and in between the eyelet of the belt - might make rotation a little easier.
Hello, looking this over, it appears to me, that if we’re not space constrained, and we’d like to try to cut a full sheet with high accuracy, then something like 160"x160" would be good, if not near ideal.
The entire work piece ends up in the “green”.
I have some newb type followup questions to this thinking.
On maslow4_frame.html: what do the “orange” bands indicate vs “yellow”? which represents angle constraint/hits vs ?
Assuming i’m getting about 14.5’ from each belt, how can the tooling show the corners as remotely green? The “belt length required” says 165, which is within the 14.5, but what i mean is that there is no way the maslow can exceed 14.5 either, so If you’re interested, it seems that it might be good to enforce a max-belt-length constraint in the visualization to avoid indicating that there is viable cutting area.. This is just a thought, as I’m not entirely sure of what i’m talking about, but it would seem that the outlying green areas currently could be a little misleading if someone didn’t double check or use “common sense” etc.
During calibration in this setup, where can i expect the maslow to need to travel? just over the work piece area, or up into corners? (this is mainly a question of under-layer/spoil setup).
I plan to lay this out on the floor: So, a) how close to vertical/horizontal numbers… 90/0??
It seems that the belt securing bolt heights are not considered here, but are somewhat in the yaml.. (seen in some other post that escapes my recollection). Which values correspond with which corner (if you happen to know)? My plan was to align the bolt height pretty close to the corrosponding “arm spindle” height back on the router stack. reasonable idea?
my garage floor is multi-angle sloped for concrete pad draining, but i expect i can shim the spoil to be stable. Are there any concerns/gotchas i should consider before proceeding setup/drilling holes etc.?
7.any other “hey, try this/think about that” before i should do anything else?
. Through either compensation of upright deflected belt length, and/or varying feed rate in the corners, e.g. gcode/CAM post processor that applies gradient feedrate/speed dampening as XY approaches corners.
