The World Needs This Tool Changing End Effector Design.
HTML-код
- Опубликовано: 23 апр 2018
- Think you can improve upon this design? Make something, upload it to Thingiverse and send me a link! If I like your design I will test it out and feature it on the channel! Cheers 👍
________________________________________
Check out my Patreon!
/ geardownforwhat
________________________________________
Where you can buy a 3D Printer: goo.gl/gwKB6J
My Favorite Cheap Filament! :goo.gl/FuzECr
For Strong Filament! CF Nylon (NylonX) :goo.gl/a6zeQd
Follow me on Twitter
/ gdfw17
I post most of my designs on thingiverse
www.thingiverse.com/Gear_Down...
Would you help translate my videos? : ruclips.net/user/timedtext_cs_p...
Links to products in this description may be affiliate links. These links do not add any additional cost to your purchase, but the channel gets a small kickback, which helps us a bunch! Thank you! 
Наука
If this will be working in an environment were there is ferric metal shavings, you might want to reconsider the magnets, but that's a minor detail. Cool design.
I think the magnets can be foregone, if the rack is vertical. Let gravity do the work, and all.
ya in that environment he could intergrate the rack that holds the tools and the twist thing into the same part
Good point, though I am pretty sure it would be possible (maybe not with normal FDM printing) to adjust the holes for the locking bearings to the point where they don't fall out, and the locking ring to keep them in could be fixed over them with a spring clip.
You can also provide power with pins, or even though the magnets for your tools :) Essentially integrate electrical connections, think of mil spec auto connectors :p
You can use the magnets directly for power transfer. And mix the north-south polarity to ensure correct position.
I was originally planning on that, but I’m afraid that heating them to solder then will ruin the magnets, and I also would need to find some sort of spring mechanism that would ensure contact during vibrations.
I think a good idea is to have a small pcb on the inside of the head, and have pogo pins on the inside of the tool (I think a 2 sided pcb with vias might be more expensive than 4 pogo pins on a single sided board or wires).
the pcb should have 4 circular tracks (vcc, gnd, and i2c, usb, can, uart...), on which the pins sit, so the orientation doesn't matter, those are fed through to the other side of the pcb with vias.
you can use a slip ring to be able to feed the wires through the 6th axis while being able to rotate it indefinitely in one direction.
given the price of pogo pins, you might even want to double or even quadruple them so you have less chance of a faulty connection
If you wanted to use the magnets for contact you wouldn't need to solder to them, wire wrapping should work.
Gear Down For What? How about a slot just beside the magnets, so you can have a C-shaped wire (spring steel) ‘clamped’ around the magnet? You could solder the wiring to that, with the curve on top acting as the contact spring.
Then again it does feel like there’s a very high chance you can just source some premade gold-plated springs and pads like phones with swappable batteries use, making a more reliable end result
Use separate pins in the center that are not magnets but still use mixed polarity magnets for alignment. You're doing some really cool stuff. You should look into FRC. It's a non destructive robotics competition designed to challenge high schoolers and prepare them for STEM and Business related careers. Here's a video of our robot at the world competition from yesterday. www.thebluealliance.com/match/2018cmpmi_sf1m13 Later in the day we played on Ford Field in detroit. ruclips.net/video/FE0-r8cwR2Q/видео.html
Yeah, I would use a conductive cement instead of solder, but that has it's own problems. There are a few pretty good videos on youtube about soldering to magnets and how it can very quickly demagnetize the magnet, it can be done but you really have to be fast. You have plenty of room on there for a more standard contact.
That's a wonderfully simple design. I love how compact it is and that you had the forethought of electrical connections, very clever!
Really cool - If you make the slots on the tools a ring with a limit stop you could combine locking/unlocking with the tool slot, saving the extra movement from lock position to tool position and back...
After unlocking the detente collar and setting the tool in its holder, a slight twist of your 6th axis would get the magnets to disengage without requiring any force saving wear and fatigue stress over time. Killer ideas!!! Keep it up!!
That is very cool, I like the trial and error part that you showed too. It really helps people understand some of the hiccups that arise during design.
Great idea. I can already see the possibility for making the dock attach / detach point a single point
Brilliant. Elegant and simple, scalable and versatile. Well done!
If you could somehow incorporate a small slip joint that is magnetically aligned you could in the theory have the same locking system that would also allow power transfer into your tool attachments
Very clever having the existing motor also run the lock mechanism.
I think having springs lock it in would be better and making every holder double as a locking tool to eliminate having to move each tool to a locking station
This is looking awesome so far! Keep up the good work!
This Is Fascinating!
Very cool. Tip: you could use less magnets and ball bearings if it only served as the latching mechanism. Place thread on both parts to carry the load force.
Nice! Great design!
nice proof of concept. looking forward to seeing more
Pretty neat. Wouldn't it be possible to integrate the the locking and docking mechanisms together? This would make tool changes quicker, and also eliminate the need for magnets to hold the tool for a move to the locking, correct?
You might be onto something possibly
I am guilty of posting before I read the comments already here. This is exactly what I thought, it reduces the number of components, and the number of waypoints to program. Also, the effectors can be mounted on a carousel, as is common on CNC mills tool changers, to reduce and simplify programming.
you should programm so, that everytime the robotarm "finds" a tool, it maybe also recalibrates its position, save x,z,y, so that it is always perfectly calibrated... love the channel, and subscribe
It CNC they use a locking taper and drawbar, you might be able to use something similar
Amazing! Cant wait to see what you come up with! I might need to build one to follow along :)
great design. looking forward to seeing it fully autonomous
It might make a really fun prop for your videos if you made a "pirate's hook" attachment for your robot.
Pretty Awesome.
Wow thanks I have been thinking about this for my robot arm. Ya probably will not include magnets, like some said twist n lock in one location. Thanks bro u are awesome. Keep up the good work esp the heart to share❤️
Print two pieces: the round thing and the funky connector block. Glue or thread the round thing on in a final assembly. I'm sure 3D print plastic has an adhesive for it.
You could use the magnets as electrical contacts and you could embed an RFID antenna in the center to read RFID tags inside the tool heads.
Amazing as always! This is something I can utilize for my Thor six axis robotic arm.
I can't wait to build your's when it's done!
I Love your concept :-)
very sick mate!
Quick release steering wheels use a similar mechanism, but they use a spring loaded sliding sleeve to lock the bearings in place. This mechanism uses a lot less force.
Nice seeing people using a 3d printer for something other than a barefoot trinket maker. Cheers!
pretty cool my guy!
Amazing!!!
I would also add pogo pins to this design, that will let you connect whatever servos, drills, printing heads you may want to have in your tool library.
Love your design work :-)
The suggestion for pogo pins also sound good, they can go against the magnets.
Maybe the tool change station can have a servo for doing the turning in a diffrent version for use in a 3 axes CNC/3D printer:)
I love your Videos!!!
this is cool! but there are push in-push out latches, which will do the job neatly just with a one push! just a thought sharing!!
Brilliant!
You can use the magnets for contacts to run power to the attachment.
Hey man!! Love the design. Something to think about. You could use the detent bearings as electric connections so no plug required. Just a thought.
Put the locking servo on the motor arm side
suggest you look at the E3d tool changer.. they have a very useful tool changing geometry.
six areas of holding seems to be the key.
Also in terms of gearing - and 3d printers.. a major help would be to have a very small stepper on the effector.
yes, but the E3d geometry requires a servo, and isn't really designed to hold much weight.
That is the big problem with E3d's approach.. I would prefer a machine effector designed to manage a big payload and remain ridgid.
I was just about to suggest this. Going to get one of those myself. Ingenious system.
Because you are using detents, you could reduce the turning distance with the locking mechanism. This would allow faster tool changes.
Awesome implementation! You could take it 1 step further and use the magnets in the bottom of your latching mechanism as electrical contacts, which could provide power to the tool for servos.
I was thinking that, but I think soldering to them would ruin them, and it would still need a spring contact to prevent it from arching due to vibrations
A thin layer of dielectric grease on the contact points of the magnets should increase conductivity and eliminate any arcing, but I don't think you'll need it unless exceeding ~10 amps. For attaching wires you could use some type of conductive adhesive, or even an epoxy similar to JB Weld. Or get the magnets with small recessed holes, which will take a small screw to which you can mount wires and tighten with a nut on the backside.
You would be hard-pressed to find a small spring that can apply more force than the attractive power of the N and S pole on those magnets, assuming they are making contact.
You should look up quick change setups on heavy equipment eg front end loader / skid steere
MIND BLOWN!!! 🧠😵
A deeper shaft would be beneficial to help with horizontal load pressures and offsetting tool length. More like a cnc spindle. ( tapered cone)
that's the truest title I've ever seen
How about using the ball bearings on the side to transfer electricity? You could even have multiple functions on the different tolls depending on how many ballbearings are used, and slight vibration shouldn't disconnect them in anyway.
Could use the magnets for positioning and a solenoid to lock tool on
You could also set some spring loaded contact in the alignment magnets and use them as electrical contacts for powering the end effector. You could also nest a resistor network and have a measurement routine to identify the effector, so that the controller will know which end effector is on and how to control it.
I'd stick the identification part to i2c or uart, so you can control the end in addition to knowing which one you have..
That would mean that you would have active components out on the end effector too. And for I2C, you would already have to know the address before you can talk to it, unless you just query all the addresses and wait for a response so you can know. With the resistor network, you can just put a pair of resistors across the voltage rails, have the microcontroller analogRead() the voltage that comes back from the voltage divider, then it would know what is attached. With his current design of the interconnect, that would leave 2 pins that could both be used for PWM which would likely be used to drive a pair of servos, and the microcontroller can easily know servo drive limits based on that analogRead() voltage from the resistor network. As it is a robot arm, it is probably better to leave as much weight off the end effector as possible and sticking more active components (ICs or even another microcontroller) would just waste capability.
we could have the best of both by having 5 pins : vcc, gnd, id, data1 and data2.
the id pin (i.e voltage divider) would then dictate which protocol to use.
an alternative would be to specify a slot for each tool, so you can do away with having to id the tool, but that means you need to configure that..
the id could also be in the slot so the software knows where to find each one.
a cool thing would be to put the different slots on a belt, and have a barcode reader read each tool's barcode.
though, I would prefer a digital identification over simple analogue, because I have used analog circuits, and even if in theory there is a resolution of 1024, in practice at the end of a wire and through one or 2 connectors, you can reliably distinguish about 10 (more if you can afford to take measurements over a second an average them.) Also, the added weigh is maybe 10 grams at most, so the impact would probably be minimal.
maybe the set could be separated in 2, let's say 0..511 dedicated for passive, and 512..1032 for different protocols.
the address problem in i2c can be fixed by either using a fixed address for all tools, or using a discovery algorithm (i.e. sending a packet one address at a time and waiting for an answer.)
Have you already designed a 3d printable 6-axis? I would really love to see that video. This is seriously cool.
Nope, it’s a work in progress
Maybe you could design a tool holder imbued with latching capability so that the robot pick up the tool and lock it in place at the same position...
Thats an awsome idea. Would it have been possible to use the original flat plate design the same way? Meaning without the servo but lock it manually. An exposed release button or lever to release or engage like your twist lock apparatus. The twist lock is a very interesting design. There are many variations of quick locking tools in construction equipment none like this tho. It could have huge possibilities.
Pretty awesome! Now you just need some pins and rings inside so u can transfer power and/or other signals. Tho at this point it sounds like you're already talking about something like that XD
You should look into using a spring-latch Polymagnet from CMR for your catch and release mechanism!
Wouldn't it be faster to combine the latching and grabing in one motion ( basically a latching mechanism for every tool). Also, how about a build in (into the middle of the latching mechanism) connector for the tool servo (this would save the step of connecting the tool servo with the robotic arm and save time). Befor i forget it, very nice video, i love the simple and elegant way this mechanism works without taking to much space and an additional servo, great :D.
Why not use an aperture ring lock? The ring when rotated fits into a grove in the tool piece and locks in by using a ball bearing like a ratchet handle and socket. This looks like the long to wear on the tool from the ball bearings would cause failure sooner. The aperture would cause a heartier hold that would only put wear on the ball bearing during tool change instead of operational load.
Have you made the stl files available for this design. I looked on Thingiverse, but couldn't find it. I like what you have done so far. Great job!
Call it "dock and lock"
Amazing idea bro. I tried searching for the design files at your website but couldn't find them. Are they available somewhere?
You can use the magnets themselves as electrical connections
Your first design is very very similar to how a skid loader changes buckets.
imagine far in the future, a sentient AI worker robot. You can tell he's a technical engineer robot because his torso is covered with a dozen specialized HANDS.
You walk up and shake hands with him, and he really quick switches to his special handshaking hand :P
"Because in the future I plan to put in little tiny connectors..."
Literally two minutes after I thought about that. My version was that wires are connected to each magnet that touches the utility module's magnets and allows flow of electricity.
Good ol' HS-50s
Try a dissolvable support material
could your tool holder hold and lock all in one also your magnetic hold could supply current. also
The only 'real world' issue I see is contamination of the magnets.
Small neo-magnets can get 'fuzzy' with bits of iron and steel over time.
Your magnets, while they do not carry a load, depend on a close face-to-face engagement.
The ball bearing locking ring might have trouble locking if the effector's position is off due to grit on the magnets.
Aside from that, cool beans!
Greg Gallacci would a thin cover help? Ferrous-fuzzy may still collect, but one can do occasional cleaning by lifting the cover off and dusting off the fuzz.
This is a Lot like a "real" toolbar chancer, only they are operated by air pressure..
Should just use the magnets for the electrical connection like a magnetic pogo connector
How are you going to pass the electrical from the tool?
Cant you inckrporate tool changer and holder in one?
I'm trying to make a tool changer for my router cnc using magnets. Thanks
Rather than a separate 'unlock' spanner, then park, use a tool park where you enter, rotate to unlock, then withdraw, leaving the end-effector behind?
But could something like this, assuming it was made of metal and scaled up, hold a spindle and handle the forces of machining materials like metals?
where can i find the stl files ? ,i didnt found any stl files about this project in thingverse
You sir are a genius. You should be working for NASA or some other tech company that needs design innovation. Have you considered getting a degree in mechanical engineering or a Polytech certificate in mechanical engineering technology? I have met engineers who don't have one tenth of your ability.
I really don’t want to spend the money on it
Hrm, how does the locking pall work with needing to use the 6th axis? I.e. how would this work with a screwdriver attachment? It's neat that the lock-unlock force needs to be applied to the collar, but hrm. Looking at 4:54, I don't see how rotational pressure on the part won't be transmitted. (Though I suspect that's very much me not seeing it.) Is the notch shown at 3:02 the only thing that is preventing the internal thing from rotating?
Also, very very neat design. At 8:37 (sooo neat) I'm reminded of a sentence from an old old short story (Pirates of Zan, if you're interested) that noted that there needed to be a button pushed down as a vibration check before rotating an airlock latch. Would that help?
11:11 I think the test for this, instead of a screwdriver would be a tool-changing tool-changer. (Or maybe just an extendo-stock that has a normal tool mount on one end, and one of those rings on the other. If the robot can change the far part without loosening the near part (or shearing any plastic) I think that'd be good enough? Or, instead of needing a vibration check, what if the electonics port extended out in a hex-head (with the same clearances as the outer circle, as you note) so that if the tool wanted to rotate it would be pushing against that printed hex head instead of the tiny notch?
But you assume that your 6th (usualy small and not powerfull) axis can lock and unlock the end effector?
Look at e3d new tool changer ises kinematic design very cool
Why not just build a collar around the green outer locking part that holds it in place statically. Then the middle is free to spin
sir can u send me the stl files , i didnt found them on ur thingverse page
could you use dowel pins instead of ball bearings? Like just round the ends of them and have springs instead of magnets. I think this would make it stronger because the detents could be deeper
Jack gee dubs Or just nail - no more magnets around then...
Why not use the magnets themselves for the powering/control for the tools?
Magnetic contacts
Why not use the magnets as the electronic connector points? You shouldn't be driving anywhere near enough current to damage then. You also reduced complexity on the mounting process
is that detent going to ware out?
Probably, but it’s easy to replace that part if it does.
Too many NEO magnets will prevent it from being used around small ferromagnetic objects.
Also, it will collect small metal crap in all the wrong places and will fail in the most inopportune moment (Murphy's law).
Patent that shit right quick
why not just build that latching mechanism into each tool cradle?
Fuckin' magnets, how do they work?
STOP STEALING MY THOUGHTS YOU DEVIL GENIUS!!!
Nathan I'm calling the thought police... A thought crime has happened
Nasa wants to know your location
Electro magnets are a thing
Or a solenoid for a locking mechanism. This would mean constant power isnt required making it safe under power failure
Make one as small and cheap and lightweight as this mechanism and I would be happy to use it
And don't use electricity.
what might be cool if the design used printed magnets that when twisted offered greater support, like in this smarter every day video: ruclips.net/video/IANBoybVApQ/видео.html
of course that would be expensive, but it may reduce the physical wear and tear that comes with your current design.
Can't you just use a screw design, and have the end effector screw the tool on. That way you wouldn't need the fine detail for the balls that might wear down quickly, nor the magnets that could intervene with the environment.
Check out ATI or Schunk tool changers.
Kind of Did this in my milk and lathe 7 yrs ago. Check out @mikeshobbycorner