Neat. I was in the middle of building a scara a while back but was having issues with the elbow joint so I had the idea just use a linear axis instead. So I basically had a rotating base with z and x axes on top. I called it an inverted polar and much like this design it was capable of printing in 360 degrees around itself. There's a video up on my channel, I've done several iterations of it since then and it proved to be a very reliable and versatile machine up until recently when I left it in my car and the printed parts deformed in the heat.
Haha yes I have actually seen your videos before. A few years ago I actually modeled up an inverted polar printer as well but decided not to pursue it. It is a great design you have going with it.
hey, great to hear from you! this is so cool! i was pondering 360 degree scara mechanics not long ago and here you are again with such an amazing new idea. i wouldn't have thought that motor mass is a particularly big concern for scara mechanics as long as it is close to the centre column... and so it wasn't a major design goal in my thinking, but your solution for leaving them ALL in the base is genius. really creative, effective, interesting, and definitely brings a bunch of additional advantages. any hurdles it brought, you seem to have already overcome. its extremely cool to see it working so well! keep up the great work and please do keep posting videos. have there been any developments with your larger scara? the (was it cycloidal or strain wave most recently?) gearbox looked really promising, it would be nice to see it working with the latest refinements. thank you for sharing your ideas, i really find these videos very cool. until next time, best wishes!
I actually put the large SCARA arm on the back burner due to the complexity of it and having failed many times (I think I did 4 or 5 total rebuilds with all of them failing in some way), so I decided to do something smaller (more manageable) but equally cool so you get this.
Little late to suggest this, but one way to remove some of the cable issues would be to couple its rotation to the 1st arm segment using the shaft you already run up there. This would keep the the cable from being draged around by the the arm. One thin i would be wary of is the spool placement, as it might get in the way especially if you want to use a springsteal strip to keep the wires from drooping into the toolpath when you raise the arm. Looks awsome, good luck!
Very nice! Looking forward to see it perform. For the SCARA arm I've been planning for ages I wanted to go cable drive but infinite rotation is just too good to forego. So belt drive it is, but aren't you worried about backlash? GT2 or GT3? Expected precision at end effector? Lately, I've been trying to figure out ways to get rid of the slip ring rotary contacts. Rejected low power, low efficiency induction coils. Landed on thrust bearing pairs for power transfer. Might also work on bearings designed for axial load, but only with electrically conductive silicone lube. Might. For data, either something wireless and low latency (RF24?), but I'm looking at an optically isolated half duplex serial bus right now (so like RS485), where the twisty turny bit has the photodiode/LED pairs looking at each other through a darkened joint shaft. It would be cheaper than slip rings, less noisy, less wear. Potentially slower, though I think 38400 baud should be doable. Also it still needs to be developed. Sorry for ranting, this has been keeping me up at night.
For a general robotic arm that probably could work. For my use case I'm just using a cheap slip ring I bought on AliExpress (I think it's rated for 2A per wire which is fairly low). I'm running this at a higher voltage (24-28V) as to not put too much current through the slip ring. My plans don't entail using anything with a high power draw I think it will work fine. Your idea of a main power system with digital signals being sent externally would work fairly well I think but is more suited to larger applications (I assume) because you still need some form of conversion from power and digital signal to analog for motors and sensors to use.
@@KoalDesigns Agreed, not everyone wants MCUs strewn along the length of their arm. And stepper drivers aren't weightless. Those 2A wires can also be doubled up for more power and less noise, so there's that. There are 3 points of infinite rotation in my design, so that's what put me on this trail. Thanks for your input!
Working on it currently. Aiming to add custom hall effect sensors for homing and porting SCARA kinematics into Klipper. Progress is very slow due to classes.
Great design! The problem of the wires wrapping around at higher z coordinates should be a non problem if the printer is direct drive, because the filament should be taught enough to spin the spool
Neat. I was in the middle of building a scara a while back but was having issues with the elbow joint so I had the idea just use a linear axis instead. So I basically had a rotating base with z and x axes on top. I called it an inverted polar and much like this design it was capable of printing in 360 degrees around itself. There's a video up on my channel, I've done several iterations of it since then and it proved to be a very reliable and versatile machine up until recently when I left it in my car and the printed parts deformed in the heat.
Haha yes I have actually seen your videos before. A few years ago I actually modeled up an inverted polar printer as well but decided not to pursue it. It is a great design you have going with it.
Excellent work
This is really cool; it would be interesting to see how well it does on parts with large z-axis dimensions.
wow that thing looks amazing
This is awesome! Really cool idea, can't wait to see more!
hey, great to hear from you! this is so cool! i was pondering 360 degree scara mechanics not long ago and here you are again with such an amazing new idea.
i wouldn't have thought that motor mass is a particularly big concern for scara mechanics as long as it is close to the centre column... and so it wasn't a major design goal in my thinking, but your solution for leaving them ALL in the base is genius. really creative, effective, interesting, and definitely brings a bunch of additional advantages.
any hurdles it brought, you seem to have already overcome. its extremely cool to see it working so well!
keep up the great work and please do keep posting videos.
have there been any developments with your larger scara? the (was it cycloidal or strain wave most recently?) gearbox looked really promising, it would be nice to see it working with the latest refinements.
thank you for sharing your ideas, i really find these videos very cool. until next time, best wishes!
I actually put the large SCARA arm on the back burner due to the complexity of it and having failed many times (I think I did 4 or 5 total rebuilds with all of them failing in some way), so I decided to do something smaller (more manageable) but equally cool so you get this.
Little late to suggest this, but one way to remove some of the cable issues would be to couple its rotation to the 1st arm segment using the shaft you already run up there. This would keep the the cable from being draged around by the the arm. One thin i would be wary of is the spool placement, as it might get in the way especially if you want to use a springsteal strip to keep the wires from drooping into the toolpath when you raise the arm.
Looks awsome, good luck!
Love it!
Very nice! Looking forward to see it perform.
For the SCARA arm I've been planning for ages I wanted to go cable drive but infinite rotation is just too good to forego. So belt drive it is, but aren't you worried about backlash? GT2 or GT3? Expected precision at end effector?
Lately, I've been trying to figure out ways to get rid of the slip ring rotary contacts. Rejected low power, low efficiency induction coils. Landed on thrust bearing pairs for power transfer. Might also work on bearings designed for axial load, but only with electrically conductive silicone lube. Might.
For data, either something wireless and low latency (RF24?), but I'm looking at an optically isolated half duplex serial bus right now (so like RS485), where the twisty turny bit has the photodiode/LED pairs looking at each other through a darkened joint shaft. It would be cheaper than slip rings, less noisy, less wear. Potentially slower, though I think 38400 baud should be doable. Also it still needs to be developed. Sorry for ranting, this has been keeping me up at night.
For a general robotic arm that probably could work. For my use case I'm just using a cheap slip ring I bought on AliExpress (I think it's rated for 2A per wire which is fairly low). I'm running this at a higher voltage (24-28V) as to not put too much current through the slip ring. My plans don't entail using anything with a high power draw I think it will work fine. Your idea of a main power system with digital signals being sent externally would work fairly well I think but is more suited to larger applications (I assume) because you still need some form of conversion from power and digital signal to analog for motors and sensors to use.
@@KoalDesigns Agreed, not everyone wants MCUs strewn along the length of their arm. And stepper drivers aren't weightless. Those 2A wires can also be doubled up for more power and less noise, so there's that.
There are 3 points of infinite rotation in my design, so that's what put me on this trail. Thanks for your input!
That is genius dude. This is basically a massive format 3d printer, with a footprint of a ender3
Please show us more! I love it!
Working on it currently. Aiming to add custom hall effect sensors for homing and porting SCARA kinematics into Klipper. Progress is very slow due to classes.
nice work. is there a particular reason to have it continuous full rotation? Instead of just moving it back?
Great design! The problem of the wires wrapping around at higher z coordinates should be a non problem if the printer is direct drive, because the filament should be taught enough to spin the spool
Hanzhen harmonic drive gear , strain wave reducer, over 30 years experinece
I wonder if you could couple the movement of the to to the proximal arm movement so that it's not driven by the wires
Theoretically its possible, I was thinking about doing that actually but then decided that it adds more complexity than it is worth.
First
Amazing work