I just adapted a bunch of sisyphus designs for my sand table using Sandify. All those beautiful round designs make me want to build a polar table. I’ve been googling around, but haven’t run across any polar designs yet that don’t require the drawing surface to turn. Anyway, can anyone point me at a good polar design? Thanks!
If you build it, they will come. :D.
I haven’t seen anything like that (I am far from an expert though). I would guess the mechanism in the polar tables is something to spin a rod, and something that rides along the rod. The part that seems most mysterious to me is how the thing can spin 3000 times in one direction, and still have anything attached to it. I wonder if the radius axis is moved by a stepper at the center, so everything that spins could be passive. There must be some way to do that with a couple of gears and a continuous belt. Sort of like the corexy. If you spin just the gear motor, the thing turns and the magnet moves closer, if you do them both, you can move one or the other. I’m not making any sense, but it works out in my head :).
OK, I couldn’t help it, I had to draw what I was thinking.
[attachment file=78011]
The radius motor moves a belt, which moves the gantry. The gantry slides on some track, which is held at a particular angle by the large gear. To change the theta, the theta motor turns, and to keep a constant radius, the radius motor needs to turn the same angle. The mounts connect the track to the large gear, and both motors are stationary.
How would you get to zero radius though? The gantry and the middle motor need to be able to not collide. Maybe there’s something with a string instead of a GT2 belt, so you can wrap it around a pulley instead…
I guess if the gantry was connected to the belt on the “bottom” in that drawing, it could move past zero…
And, you don’t really need the left side of the slider either :). This is really interesting to me (in case you couldn’t tell).
If you put a 608 bearing on the left side, and a 608 bearing out at the end with the normal smaller pulley on the motor, you probably have clearance to get past in a straight line. You wouldn’t have to have a continuous belt, the zip ties could be on the gantry. I wonder if EMT is the right choice for the linear rails? I suppose it all depends on the scale of things. That gear would get pretty big if the linear rails were big. Could one piece of 2020 work?
It’s just like this, but with the arm mounted on the platter, and the middle of the platter is cut out for the arm motor.
With a top/bottom configuration on these rods, EMT could probably easily be used.
Hi Jeff - yes, I think you’ve got the idea right. Magnet goes back and forth on a bar that in turn rotates like a big clock hand. The motion is already built into Marlin, and maybe GRBL too (my present table uses GRBL).
The problem is - how do we get the precision of a cartesian machine into a polar machine? The axis arm that spins is simple enough - it’s basically the X-axis of just about any cartesian - a carriage that slides along a length of profile, or on some smooth rods - or in the case of the ZenXY, on EMT conduit. Put a stepper on one end with a belt or leadscrew and you’re golden. But how do you spin it and keep it rigid? Spinning with precision is fairly simple - another stepper and belt, or even string - but the rigidity is the challenge, in my mind at least.
Here’s an example on youtube of a diy polar sand plotter. It uses some parts from an old printer or scanner. It works, but the precision isn’t there. https://www.youtube.com/watch?v=X7liebJP0JM&t=196s
I was thinking 2040, mainly cause I think I might have a piece lying around.
That build looks like it’s reinvented the wheel (no pun intended). It’s reeling out wire for the gantry, and it looks like it really just needs a stronger magnet to stop the jerkyness.
I don’t think rigidity is going to be that big of a deal, because it’s got almost zero drag.
A GT2 belt is 2mm/tooth, and if you had a 16 tooth pulley, and 1/16th microstepping, that’s 16 steps/ whole step * 200 whole steps/rev * 1 rev/ 16 teeth * 0.5 tooth/mm = 100steps/mm of belt.
If you had a circumference of the big gear at 180mm, you’d have 50 steps per degree. That would be about a 60mm diameter gear, totally printable. The larger, the more resolution you get.
If you had a 500mm linear rail on top of that, 1 degree would mean a change of about 9mm at the edge. So that’s about 0.15mm/step. That might be cutting it a little close, but I think these numbers are worst case. Making the gear bigger would help with strength and rigidity too, so something in the 120mm diameter would about as big as everyone could print and end up with a lot of resolution. 500mm is also pretty large radius, IMO. It would be twice as good at 250mm.
I haven’t got any of that lying around. I do have a bunch of conduit, 5/16" bolts, and 608 bearings though.
Reeling out wire to the gantry - that’s crazy.
I’m thinking of using a slip ring, like this: https://www.amazon.com/GeeBat-Capsule-Electrical-CIRCUITSx2A-monitor/dp/B01MYMM1KT/ref=sr_1_3?ie=UTF8&qid=1543867122&sr=8-3&keywords=slip%2Bring&th=1
You can get them with different numbers of conductors. I was thinking of mounting the steppers and controller board on the spinning gantry, and then feed them with a simple two-wire slip ring.
I think it was Barry that told me that slip rings running any current will have some kind of reaction because it’s constantly making little arcs. I don’t know how long they would last. Maybe it’s plenty of time, and when they fail, it’s not life threatening.
Look at this one:
That video highlights some of what I was talking about with rigidity. Note that the arm, which is pretty long, is supported mid-radius with bearings riding on a raised track. Gives me some ideas of how to support mine.
Darn it, I really don’t have time for this project, but I can tell I’m probably going to do it anyway!
Yea, slip rings are basically consumables. They all eventually fail. The good ones use graphite sliders, and the smoother they start the longer they last. I’ve seen folks try to use the bearings in home made lazy susans pass current, is never ends well.
Do you have any idea how long they would last? I mean, if it’s a 1000 hours of run time, that might be long enough for almost anyone.
Karl, I just got off the phone with your friends, family, and coworkers. We all agreed. This is more important. Netflix also called, and they aren’t going to release any new shows until this is done.
3 Likes
This site may have a movement on it that might help. At any rate, it is cool site to browse thru. http://507movements.com/
LOL that settles it, then!
That is a neat website. If nothing else, it will bang on some pipes and get thing moving in my head. I looked through them, and I don’t see any obvious answers. There’s no corexy machine in there, so we should probably register the url for 508movments.com