Update on the Spherical Parallel Manipulator 

Oh any hardware I incorporate (eg bearings are vital for the SPM to function well) will be super common and cheap, I have good alie sources for bulk-cheap bearings but even if that store dies they'll still be skate bearings and those will still be everywhere. There is no 3d printed substitute for bearings (and I have a background in industrial 3D printing, I've played with every tech out there), the closest you'll get is a smooth metal rod and a hole in the 3D print (bonus points if you use the rod to melt the hole to be a precise fit), with no non-printed hardware the best you'll get is printed rod (ideally high resolution resin/SLA, SLS/MJF/DLM second choice behind that) going into a printed hole - no 3d printing tech produces rolling features well enough to do a ball or roller bearing. That said, I'll try and make a variant with as little hardware as possible too

@@makebreakrepair 3d printing those bearings is totally an option aswell, but this should be up th the user if they want to source standard parts or print the parts 😉

@@makebreakrepair have you tried filament with lubricant inside it? Like ones from Igus?

@@conorstewart2214 Self lubricating filament costs more than bearings (the little ones in the arms are about $0.50USD each) and isn't widely available or easy to print. I started out as a broke maker/tinkerer (now I'm an older, better equipped, broke maker/tinkerer 😅)so I sympathise with the desire to make everything on the 3D printer but there really is no substitute for bearings (note a smooth metal shaft in a hole is a type of bearing, called a 'plain bearing', for most of the industrial revolution that's the only bearing there was... but it needs a very smooth metal shaft to work). Especially with this crazy mechanism, due to its nature it needs extremely low friction on the pivots or everything distorts and locks up (eg if you used a 3d printed rod in a 3d printed hole it'd catch and jam). Don't worry though, I'll make the open source version as cheap and easy to source as possible.

Hell yeah!

Hah I'm with you on that... I made the first one because it looked damn cool even though it's less practical than the obvious solution (ie the dull functional option) - as a result I don't use it for microscopy much but it gets plenty of use as a desk fidget toy 😅

Half the battle is knowing what something is called ;) I had found 2 axis goniometer stages (not cheap, even from china) though they (or at least the ones I've found) all lacked the ability to adjust the work up to the center of rotation - though a small Z stage on top with your work holding attached to it would solve that. They were all still very limiting for what I wanted though (the SPM has similar limitations mind you) in that you can't look at anything longer than the center of rotation (usually 60mm above the surface for the ones I found) eg an endmill. The polar manipulator I showed also exists (and can be made purely from camera accessories bolted together, though that's a few hundred dollars of parts) though I don't know what one would be called in microscopy/industry terms. There's still a reason for the polar manipulator project to exist in that what is out there (I may be wrong, feel free to send me links etc) is an order of magnitude more precise and more expensive than needed for the kinds of low-magnification, inspection style microscopy I have in mind - If we can make one out of 3d prints, two steel rods, a bolt and some bearings then it becomes a $30 tool that anyone can make. Similarly, making one that uses off-the-shelf semi-precision elements for a more serious tool (ie somewhere in the region of $200 of parts? tbd) also bridges a gap - there's no (similar disclaimer, send links if you know of some) medium-precision entry-level microscopy tool for this job, anything I found costs double what my pretty decent digital microscope did. But yeah, 2 axis gonio stages have been in my aliexpress cart for months, could never quite justify the $450 or so but they're damn cool.

Oh yeah, what you're developing here is in a completely different category of precision/cost. My experience is in materials science with ultra precise scientific tools. The X-ray diffractometer that I use cost ~$500k, the goniometer is probably half that cost, and it's accurate to within 1/10,000° across four independent rotators. But I'm looking at reproducible positioning on the order of picometers. What you've shown here looks like it could be adapted for similar use, given precision manufacturing, and improve on the common set-up we use. I'm interested to see where you take this.

​@@johnmiracle7740wow

@@johnmiracle7740 I love precision machines/mechanisms, I have lots of salvaged scientific equipment etc that I strip down to make other things or to understand them better (don't have many examples on my channel, I should post some, though this vid I talk about the differential screw in a microscope fine/coarse feed stage ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-ezp9d5235lc.html) I'd love to play with one of those ultra precise goniometers... know of any getting thrown out? :P

@@makebreakrepair No, usually these machines get used until they turn to dust. When they do get retired, companies usually buy them back as an incentive towards a new machine (like a loyalty discount). You can sometimes find old scientific equipment on public auction sites for universities. Most R1 and some R2 universities will host monthly or quarterly surplus liquidations.

True!

i was absolutely thinking of miniature painting

Had a similar idea 😅

A ball vise would likely be better (there's some 3d printable ones too) as you don't really need the object to orbit around a point so much as you need access to all sides.... though I may be wrong, I don't paint miniatures... Either way I'll get a design out soon for everyone to play with :)

Yeah, after seeing the responses the current (vague) plan is to release an open source 3D printable version of both that emphasises cheap for everyone to make, an open source more rugged one using off the shelf metal parts with 3D prints as glue (eg the macro slider attached to a 3D printed base etc) for anyone who wants to make a more serious unit, and, if there's demand, I'll make a small production run of a rugged, all metal unit for lab use/serious work or similar as a product. Genuinely open to suggestions though, I'll make an update vid about it soon

Or, maybe The Boys Surface on top (Jos Leys has a good video animation), printed in conformant TPU (or silicone), so it can change shape as it moves and be dimpled in and out along the central axis. Then maybe put RGB LEDs in each of the three axes for visual feedback of position. Also feels like combining it with reaction wheels could be interesting.

I have a photography background and I understand the struggle of shiny object product photography with in-camera subject isolation 😅 Similar goals for microscopy in that reflective surfaces blur/wash out the image, I'll keep that use case in mind with the updated designs (releasing some soon). I may make a motorised polar variant for the community but for a motorised SPM check out @swannschilling474 who was one of the sources of inspiration for this project, they've done a fantastic motorised SPM that's available to print now ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-vNV2LetkfEg.html

that's a good one patent it quick!!

@@MrAllstar 🤣 im sure its been done before

@@sshh7510But has it?? Post a make and model of one

I plan to release a free 3d printable one, but I might make a more rugged one as a product for labs and such (if there's demand?) I posted an update vid covering cad/files etc, check it out and leave a comment there on which concept you'd prefer and what your use case/application is (people keep surprising me with novel uses) and I'll do my best to accommodate that in the released design. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-wN02itp6-z4.html

That's an insightful thought, thanks for the perspective; I think you're right in that the 'weird cool' one is quite tricky to cad but the polar type is very straight forward for people to make. Current plan is I make a free for everyone version of both as there seems to be demand for it and possibly make a rugged polar one as a product; I'll update everyone soon.

PS love the nails, whether rock n roll or camp 🤘💅

There's some excellent powered ones on youtube if you need inspiration (check out @swannschilling474 ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-vNV2LetkfEg.html) but I'd be keen to see your project when it's ready.

@@makebreakrepair Very impressive, that he was able to fit everything together so tightly! I really like the idea with the inverse gears. I will probably also make a little demo video of mine, hopefully next month it will be finished :D

Exactly, I sorta touched on it in the video but having a universal work table of sorts could allow the community to make whatever work holding they needed and slot it in, the main idea I had in mind when talking about that was: A tiny XY table above a bearing (ie you can spin the XY table) and the XY table would have the workholding on it (tiny vise, ER holder, etc). With that you could hold any object and offset the object such that the center of rotation is where you want it, eg assemetric or non-round objects

This is the first 'alternative to bearings' suggestion I've heard that's actually viable... though can you link me examples that allow rotary motion only (eg the same motion as a bearing)? I know spherical motion is achieved with a magnet (spherical iirrc) and a ball bearing, but for these to work we'd need to constrain all motion other than one axis of rotation per joint... I haven't kept up with the delta 3d printers with magnets thing so I'm not sure if they've done that. That said, the little bearings are only 50c each in a pack of 20 so it may not be necessary.

@@makebreakrepair I was thinking that just the joints connecting to the center hub would be spherical, the other set of bearings on the arm would need to remain bearings for the axial constraint that the spherical magnets wouldn't have(i'm not sure if that would impact the mechanism or not). I do find the spherical magnets to be surprisingly strong and resistance to axial displacement. Hayden arms for delta printers use a high carbon steel ball with a machined delrin cup and a strong magnet at the base of the cup to keep the ball seated. My guess is that pla would hold up well enough to have the socket integrated into the arm with a small hole for a magnet to be glued in place, then the other side of the joint would just need to hold a steel ball and its easy to find them already with threads for a screw or threaded portion to be inserted. Bearings are probably the easier and more robust option but who doesn't love magnets if it is only a light duty positioning device?

Yeah, micro endmills are really hard to inspect, and even small ones (2mm+, my arbitrary delineation between micro and small) are tricky

Apologies for not putting giving you credit sooner; I only shot the original video for friends and forgot about it, a few months later and it has half a million views somehow?! If I'd known it'd get such an audience I'd have made sure to mention your project from the start; I felt guilty that a one day project and a 1 minute video with zero production quality went viral without any effort from me (i sent the vid to about 5 friends... I have no idea why the algorithm gave it any attention) and got more views than the incredible project that inspired it. (For anyone else reading this: neither swanschilling or myself invented the SPM, it's been around for decades, swanschilling made a well documented, 3d printable, motorised (with full IK!) one that inspired me to make one) I saw an SPM somewhere else on youtube (can't remember which, was ages ago now), realised it would work for the microscopy task, then while searching for a printable one I found yours. I didn't have any of the bearings etc you had on hand but your vids helped me understand that designing one is fairly straightforward (axis of all bearings meet at a point) so I knocked one up. Your project is damn impressive and you've put a lot of work into your videos, anyone looking for a motorised version I send straight to you because you've basically made a fully fleshed out product free for people to make. Also, IK and motion sim in Unity? I'd love a tutorial vid on how you set that up (or some resources you can point me to?) I'm super impressed... I know just enough to do James Bruton style IK of an independent set of serial polar joints I'm not sure how you'd tackle something like an SPM with all the parasitic movement etc. You should definitely do a V2, there's been a lot of people in the comments for my original vid suggesting it be used for 3d scanning (ie a turntable that tilts), I think that'd be a great use case.

Yeah, my main use case too, checking tiny end mills was the original reason I made it. I think I'll just make both as open source 3d printable tools, and work on a more rugged version for lab/microscopy/more serious work. Watch this space, I'll update people soon

Yeah... I think I'll make both for everyone, it seems to be a roughly even mix of people wanting to get stuff done and people wanting the weird toy. I'll update people soon, doing some first draft designs now

Interesting, I dabble in optics/lasers/related (still learning), what lenses do you clean under a microscope?

Laser Optics and lenses. Defects of the size above 4-6 Microns can become problematic

Ah, serious laser-ing, like clean-room level? I'm unsure about the off-gassing of the various filaments but one could substitute the bearings for clean room rated ones

Interesting use case.... the current vague plan is to release a free, 3d printable design for the polar and spherical variants for anyone to make, and possibly (if there's demand?) I'll make a small run of rugged/professional grade polar ones for labs and other serious users. Based on your comment (and a few others that are similar) I think I'll add the ability to motorise/control/automate the polar variant (both the free and the rugged one), but I won't make a motorised SPM as there's already an excellent project that is doing that better than I could, check out @swannschilling474's channel, their project lends itself well to automation and can be printed now ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-vNV2LetkfEg.html

If instead of tool holding, you put work holding in the middle, this could be a 4th/5th axis add-on for an existing 3-axis machine for small parts - if it could be made rigid enough

No, it never gets stuck

You are correct for my design (it's intentional, the arms are elastically averaging each other to give higher static accuracy), but it doesn't have to be the case - another project I found in my research was 'The Agile Eye' project by a french university, it's a high-speed mechanism for rapidly aiming a camera, check it out ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-gVlqjA1EKE4.html

@@makebreakrepair Nice, that was a way more "smoothly" controlled one than I've seen before.

There's actually an open source project doing just that... can't remember the name at the moment though. I think it was designed to go on an Ender but could be adapted to anything

For microscopy (outside of high magnification etc, where you need expensive positioning systems from the start) it's not really an issue as you're more interested in vibration damping (ie the subject shouldn't shake while you're looking at it) and that any sag due to gravity is consistent and smooth (ie when you go from subject -45 to +45degrees the subject shouldn't sag out of frame suddenly when the load changes direction). I'll have an update video soon but I've already printed and tested a polar variant under the microscope and it's stable enough for some decently high magnification :) (and before the '3d print everything, even the bearings' brigade arrives: that's only possible with actual bearings ;)

I'm not sure how - pick and place is usually planar with a rotating tool head, i don't know of any components that need to be placed at an angle relative to the PCB... Or I don't understand your question?

Don't worry, I have a good cheap bearing source everyone can use and I'll design it to not use exotic bearings and huge chunks of shaft. I'll keep everyone updated on progress etc as it happens but yeah, key goal is easy and cheap. The arms in the original use flanged skate bearings (F608s), the flange makes them very easy to design for with 3D prints. 10 pack is $9 here (not sponsored/no affliation) www.aliexpress.com/item/32814352400.html

I'd say it's likely not the best option (if we're talking about the wacky armed SPM) as the resistance to motion (when you're moving it by the work table rather than the arms) varies greatly depending on arm positions, it'd feel weird basically. You also don't need its key feature - orbiting around a virtual point. I think a ball joint with hall effects would work better for your application.

Good information. That all makes sense, thanks. Having thought about it more I wouldn't be surprised if durability were also a problem.

I thought about that, I love flexures and I've designed a few simple ones (nothing complicated, still learning about designing them) but I'm no expert so I may be wrong but I think they wouldn't work well here... that said I do think it might be possible to make an adjustable center variant by strategically adding pin joints... it works in my head but I'd need to prototype it

@@makebreakrepair It would be ambitious, that's for sure.

...and quite possibly a folly, but that's how I learn 😅

I am also interested in this.....willing to pay for it also!!

@@raymondcuda5906I’ll pay double what they offered!

Alright, lets start a bidding war! (I'm broke AF so definitely open to making a product people would buy, but either way I'll release an open source printable one for people :)

Yeah, after seeing the responses the current (vague) plan is to release an open source 3D printable version of both that emphasises cheap for everyone to make, an open source more rugged one using off the shelf metal parts with 3D prints as glue (eg the macro slider attached to a 3D printed base etc) for anyone who wants to make a more serious unit, and, if there's demand, I'll make a small production run of a rugged, all metal unit for lab use/serious work or similar as a product. Genuinely open to suggestions though, I'll make an update vid about it soon.

Yeah.... I built the first one because it looked cool even though I knew there was a more practical way to do the same thing :)

@@makebreakrepair I just had a very dumb idea... Can you tell me how back drivable it is? And is it prone to gimbal lock that way? I want to put a joystick on top of it 😁. Add three servos and you have full 3 dof force feedback.

It's very back drivable, but the felt force varies based on arm position so it would feel weird I think

@@makebreakrepair hmm, true but with some clever math probably correctable. I'm thinking about how to design a 3 to 6 dof force feedback joystick for years now. However since I am apparently a masochist I set one of my design goals to be that all motors have to be fixed to the base. Because why easy... One day I may even have an idea convincing enough to build an actual proof of concept lol.

That's an interesting project.... Are you familiar with 6dof cad 'mice' (3dconnexion space mouse and similar)? I have one, I haven't taken it apart (yet....) so I'm not certain but I think it works by having the knob you manipulate held floating by a few pairs of opposed springs and senses movement with hall effect sensors tracking the deflection or twist of the knob. If you take that concept, and add electro magnets in strategic places (which will likely mean not using hall sensors for position....) you could give a sense of force feedback, at least for XYZ translations, not sure how you'd do rotation force feedback though. Another idea: take whatever works for existing 3dof (roll,pitch,yaw) force feedback joysticks, then mount that on a small xy stage with opposing pairs of pull solenoids (ie the type where their default state is out and they retract under power) attached in X and y. The springs of the solenoids will passively self center the joystick in the stage and could provide the force feedback under power. That'd get you 3 rotational and 2 translational force feedback axis, adding z gets tricky.... Just a random thought though, I'd be curious to see what you come up with if you make it

I'm working on the design of both versions at the moment, I'll update everyone soon. There will be open source, 3d printable versions (don't worry, I'll release CAD data as well as STLs) of both styles for people to adjust to suit their needs

I have a ball vise, the issue is the center of rotation is inside the ball, ie tilt the ball and the object moves out of view - not an issue for engraving etc but doesn't work for microscopy or similar. If you're suggesting making a hollow ball vise with the work holding in the center of rotation then... that's an interesting thought, though making a smooth, large sphere isn't easy (smooth enough for microscopy at least). If i've misunderstood your comment though please elaborate, the comments section has been full of interesting ideas.