The servo knows where it is at all times. It knows this because it knows where it isn't. By subtracting where it isn't from where it is, or where it is from where it isn't, whichever is greater, it obtains a difference or deviation. The guidance computer uses deviation to generate corrective commands to drive the servo from a position where it is to a position where it isn't, and arriving at a position where it wasn't it now is.
This has quite some implications... It knows that it isn't in locations it doesn't know about... And therefore knows about the existence of those unknown locations?
@@steveman1982 By subtracting where it is from where it isn't, or where it isn't from where it is (whichever is greater), it obtains a difference, or deviation. The guidance subsystem uses deviations to generate corrective commands to drive the servo from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is.
@@steveman1982 That's a meme reference don't think too much about it, if you want actual reason servo knows where it is then the answer is encoder tells it it's position.
I like the idea of mechanically adding one bit of precision by implementing a binary swich on the rotor. Pretty cool. The same switch can also be used for self-calibration since it defines the semicircle.
@@kacperzurawski5699 well, you would have these noise sources there on 12bit anyway. With 13bit you can measure them more precisely. xd I don't say you gain anything by this one extra bit here, but as a concept it's a pretty cool idea that could be implemented somewhere where having one extra bit would really mean something.
This is really cool! One thing that that wasn't very clear and I'll be a bit pedantic about though is open vs closed loop. A stepper motor can do the same thing with regard to copying the angle of a potentiometer. A better demo would have been to stall your gearbox and show that even though it was interrupted it was still possible to know its position and continue to the target location. I am 100% sure you know this but for some viewers it may not be obvious.
this is the 5th time I'm watching this video, and I'm learning new things every time I rewatch it. [note: I'm trying to build my own servo and I knew nothing about motors. I watch other videos to understand the concepts and come back here]
Cool but you should look up how ball mice work. They have 2 lights sensors and the encoders have slots offset from each other. So it's able to tell exact position and direction of the spin.
Cheap optical encoders do not have absolute positioning. Only relative using deltas. The magnetic encoder he used in the video uses grey code which allows you to know the absolute angle (within 0.1 degrees) even between power cycles. Optical and rotary encoders need to be homed every single time you power cycle since they do not guarantee that the axis was not rotated while powered off.
Not questioning your choice of a hall sensor to figure out your position, I just wanted to let you know that you could also have used a trim potentiometer (german "Spindel Poti"). Those Potis need eg. around 10 turns to change their resistance from 1 extreme to the other. They are used eg. in sailing winches used in model sailboats to move the sailes. The drum which operates the sailes usully needs to make 3 or 4 turnes to move the sailes from the "innermost" position to the "outermost" and needs to be able to hit the same positions extremely accurate. No homing sensor needed, since the Poti does the job without any needed additional help.
Rheostats and potentiometers use brushes which limit their life cycles significantly when compared to this magnet based encoder. They can be used, but its not recommended.
@@SEGnosis My winches operate my sailboat (actually racing in regattas) since more then 5 years without problems. Do the brushes create problems eventually? Sure, they will. It all comes down to how long you need them to operate. As said: mine work since more then 5 years without a problem. If they do create a problem, then they get replaced and all is well again. Homeing with and end switch however would create a problem, since I simply cannot allow the winch to turn until it reached its end position. You see the motor has plenty of power. If it turns the spindle to its maximum position it would crush the whole mechanism due to excessive force. This is so, because the innermost sail position does not correspond with an end point.
@@kallewirsch2263 Sounds like you're using a big pot with big brushes. In the case of this makeshift servo motor. You shouldn't use a pot bigger than the motor. Even if you could.
I feel like this is a sidegrade, not an upgrade, and is only recommended depending on the application of your servo. For a servo that only needs a limited range, sure. But he stated in the beginning of the video that one of the disadvantages of typical RC servos is their limited range. With a hall sensor, you have continuous rotation, which might be a constraint of certain usecases.
for high resolution of feedback(encoder position) single magnet on end shaft isn't enough. use disk with rectangular flat magnets vertically. commercial servos use glass disk either slots with tiny magnets or with optical sensors with slits on glass disk. to store the accurate location even with power losses, supercapacitor with flash memory or always on lipo battery.
This is awesome. The only criticism I have is that you need to recalibrate your bed as its too close to the print head and is 'plowing' the first layer. Keep the awesome designs and content coming!
Excellent video... planetary gears... compound gears would give more procission higher monitoring accuracy using agc... automatic gain control... and a digital comparator with positive or negative feedback... would loose the need for a limit switch... I can't believe the number of 3d printer materials being released every week... a very exciting time for 3d printing... Thank You my friend...
Very nice! I think there would be interest from people building flight simulators in using this process to create instrument gauges: if you look at some of the 3d printed servo gauge builds, they are considerably less elegant than this. And although steppers do work, they have their own clunky problems.
Very well done, can't you use, say, an outer race of a suitable bearing to strengthen the outer gear? It would obviously move the encoder gear, but those bearings are quite cheap, easy to open and have excellent strength. The other place you can find those compound gearboxes is in the winches, even the cheap car winches have metal gears in them and often pull 600kgs on 60mm pulley (making it about 180NM if you every would need to build something particularly powerful).
the mechanical switch is viable but a optical method would be far more reliable or even a reed switch or hall sensor with magnet in your large gear set
Maybe you can resin-cast the 3D-printed gears to make them more durable? I do not have the money right now to buy expensive resin and silicone but it could be a fun project in the future.
I don't know how much more durable resin would be. Where the 3D printed gearbox can flex a bit when it skips, I'd think with resin the gear teeth would just shear off and jam up the mechanism. That said, you don't have to give Smooth-On your firstborn child to get started with mould making and casting. Everything you need can be found at your average hardware store for pretty cheap: silicone caulk, plaster, epoxy (J.B. Weld might be interesting to cast with). I'm sure there are plenty of UA-cam videos that can explain the "how-to" of it better than I can.
AMIGO O BRIGADO POR DISPONIBILIZAR ESSE LINDO E MARAVILHOSO PROJETO. ONDE CONSIGO O DIAGRAMA ELETRÔNICO DESSE PROJETO ? NO DOWNLOAD NÃO TEM O DIAGRAMA ELETRÔNICO DESSE CIRCUITO QUE ACIONA O MOTOR. FRIEND THANKS FOR MAKING THIS BEAUTIFUL AND WONDERFUL PROJECT AVAILABLE. WHERE CAN I GET THE ELECTRONIC DIAGRAM OF THIS PROJECT? THE DOWNLOAD DOES NOT HAVE THE ELECTRONIC DIAGRAM OF THIS CIRCUIT THAT DRIVES THE MOTOR.
Watching you hit your bed with Aquanet was painful to see, reminds me of my days of chasing down failed prints. Get a flexsteel bed with a PEI sheet, I haven't had a print come lose ever since, I never prep the surface either
I am curious as to why you chose a mechanical switch in the gearbox assembly, instead of a tiny magnet and a hall sensor? The switch will wear out pretty quickly I would imagine.
Do the two outer Rings have different number of teeth? I'm a little confused how the transmission is achieved because they seem to be the same size and and the caseing und output shaft therefore shouldnt move indipendently (or maybe i'm just missing something obvious ^^).
How small do you think you could get one of these? They'd be incredible for tiny animations for model railroads or christmas village projects! :) Thanks!
The servo knows where it is at all times. It knows this because it knows where it isn't. By subtracting where it isn't from where it is, or where it is from where it isn't, whichever is greater, it obtains a difference or deviation. The guidance computer uses deviation to generate corrective commands to drive the servo from a position where it is to a position where it isn't, and arriving at a position where it wasn't it now is.
makes perfect sense
UNDERAPPRECIATED COMMENT RIGHT HERE
Beutiful
Bahh I wanted to make this comment haha
War thunder meme right here
THE SERVO KNOWS WHERE IT IS.
IT KNOWS THIS BECAUSE IT KNOWS WHERE IT ISN'T
This has quite some implications... It knows that it isn't in locations it doesn't know about... And therefore knows about the existence of those unknown locations?
@@steveman1982 By subtracting where it is from where it isn't, or where it isn't from where it is (whichever is greater), it obtains a difference, or deviation. The guidance subsystem uses deviations to generate corrective commands to drive the servo from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is.
@@steveman1982 That's a meme reference don't think too much about it, if you want actual reason servo knows where it is then the answer is encoder tells it it's position.
I like the idea of mechanically adding one bit of precision by implementing a binary swich on the rotor. Pretty cool. The same switch can also be used for self-calibration since it defines the semicircle.
It won't be 13 bit instead of 12 because of :
innacurate magnet placement
output noise
gear backlash
@@kacperzurawski5699 well, you would have these noise sources there on 12bit anyway. With 13bit you can measure them more precisely. xd I don't say you gain anything by this one extra bit here, but as a concept it's a pretty cool idea that could be implemented somewhere where having one extra bit would really mean something.
This is really cool! One thing that that wasn't very clear and I'll be a bit pedantic about though is open vs closed loop. A stepper motor can do the same thing with regard to copying the angle of a potentiometer. A better demo would have been to stall your gearbox and show that even though it was interrupted it was still possible to know its position and continue to the target location. I am 100% sure you know this but for some viewers it may not be obvious.
Then I'll up your pedantism. A stepper motor may skip steps, and you will have no way to correct it, hence it's an open-loop system.
@@dthe3 Isn't that what the comment said ? or do I not understand pedantism
@@DemsW Pedantry ;)
It is a word that means you have a fetisch for feet.
@@DemsW Gremlin technically left that point _implied,_ and "for some viewers it may not be obvious"
this is the 5th time I'm watching this video, and I'm learning new things every time I rewatch it.
[note: I'm trying to build my own servo and I knew nothing about motors. I watch other videos to understand the concepts and come back here]
its super impressive how his basically stock ender 3 prints such good prints respect
Your calibrations on your printer must be hella on point
Looks like an equatorial mount or a tracker of some sort you're working on. In any case, that gearbox has huge application potential.
your channel is dope. i work with ABB robotics so this is right up my alley
That was pretty awesome bro 👍 & I don't like the torque test because you heard that sound from the gears inside xD
You got a new sub bro ;)
Awesome man! Great design. I'm glad to have found your channel
Very clever. I like the choice of music too.
Awesome video, no idea why I am just finding this now. Great stuff!
Cool but you should look up how ball mice work. They have 2 lights sensors and the encoders have slots offset from each other. So it's able to tell exact position and direction of the spin.
Cheap optical encoders do not have absolute positioning. Only relative using deltas. The magnetic encoder he used in the video uses grey code which allows you to know the absolute angle (within 0.1 degrees) even between power cycles. Optical and rotary encoders need to be homed every single time you power cycle since they do not guarantee that the axis was not rotated while powered off.
Awesome project! Thanks to yt algorithm for recommending your channel
This is so smart and so so neat!
Impressive mate. Amazing project!
Awesome work and fantastic video! It was a pleasure to watch.
Not questioning your choice of a hall sensor to figure out your position, I just wanted to let you know that you could also have used a trim potentiometer (german "Spindel Poti"). Those Potis need eg. around 10 turns to change their resistance from 1 extreme to the other. They are used eg. in sailing winches used in model sailboats to move the sailes. The drum which operates the sailes usully needs to make 3 or 4 turnes to move the sailes from the "innermost" position to the "outermost" and needs to be able to hit the same positions extremely accurate. No homing sensor needed, since the Poti does the job without any needed additional help.
Rheostats and potentiometers use brushes which limit their life cycles significantly when compared to this magnet based encoder. They can be used, but its not recommended.
@@SEGnosis
My winches operate my sailboat (actually racing in regattas) since more then 5 years without problems.
Do the brushes create problems eventually? Sure, they will. It all comes down to how long you need them to operate. As said: mine work since more then 5 years without a problem. If they do create a problem, then they get replaced and all is well again. Homeing with and end switch however would create a problem, since I simply cannot allow the winch to turn until it reached its end position. You see the motor has plenty of power. If it turns the spindle to its maximum position it would crush the whole mechanism due to excessive force. This is so, because the innermost sail position does not correspond with an end point.
@@kallewirsch2263 Sounds like you're using a big pot with big brushes. In the case of this makeshift servo motor. You shouldn't use a pot bigger than the motor. Even if you could.
I feel like this is a sidegrade, not an upgrade, and is only recommended depending on the application of your servo. For a servo that only needs a limited range, sure. But he stated in the beginning of the video that one of the disadvantages of typical RC servos is their limited range. With a hall sensor, you have continuous rotation, which might be a constraint of certain usecases.
Amazing video with outstanding results. You gained a new subscriber here
for high resolution of feedback(encoder position) single magnet on end shaft isn't enough. use disk with rectangular flat magnets vertically. commercial servos use glass disk either slots with tiny magnets or with optical sensors with slits on glass disk. to store the accurate location even with power losses, supercapacitor with flash memory or always on lipo battery.
This is awesome. The only criticism I have is that you need to recalibrate your bed as its too close to the print head and is 'plowing' the first layer. Keep the awesome designs and content coming!
Beautiful project.
Excellent video... planetary gears... compound gears would give more procission higher monitoring accuracy using agc... automatic gain control... and a digital comparator with positive or negative feedback... would loose the need for a limit switch... I can't believe the number of 3d printer materials being released every week... a very exciting time for 3d printing... Thank You my friend...
1:35 that song is fire, i checked DivKid's channel and oh boy! He's a madman
"3 axis turd" i was confused for a while
Cool. Nice set-up.
Congratulations!
Pretty good job mate
The gear box knows where it is, I knows this because it knows where it isn't
Nice mötör! Great work!
The gearbox knows where it is, because it knows where its not.
Very nice! I think there would be interest from people building flight simulators in using this process to create instrument gauges: if you look at some of the 3d printed servo gauge builds, they are considerably less elegant than this. And although steppers do work, they have their own clunky problems.
Nice project !
Nice work. 👍👍👍👍👍
Thanks! Also possible to record and playback.
Very well done, can't you use, say, an outer race of a suitable bearing to strengthen the outer gear? It would obviously move the encoder gear, but those bearings are quite cheap, easy to open and have excellent strength.
The other place you can find those compound gearboxes is in the winches, even the cheap car winches have metal gears in them and often pull 600kgs on 60mm pulley (making it about 180NM if you every would need to build something particularly powerful).
It counts 👌all industrial machines have DC servo
Is there any chance of a wiring diagram please?
the picture at 3:00 is wrong. u compare the input with the output. (input minus output). the difference is fed into the controller.
great quality video
why the magnet for the encoder it's like slided and not the poles facing the encoder?
Cool projects! keep it up!
I like it! Telescope base in the future?
the mechanical switch is viable but a optical method would be far more reliable or even a reed switch or hall sensor with magnet in your large gear set
can you make a video on modifying one of the cheap ones into being able to do this?
For a little more money you could instead have a FOC actuator using a pancake brushless motor. Probably a superior choice. But nice video!
This is avesome educational and entertaining content ^^
Gives me motivation go design my own :) Thanks a ton
Creative video, thanks :)
Who’s channel do I have to subscribe to around here to get a linear servo design? Awesome project
how does a servo do it without homing?
Very cool!
Maybe you can resin-cast the 3D-printed gears to make them more durable? I do not have the money right now to buy expensive resin and silicone but it could be a fun project in the future.
I don't know how much more durable resin would be. Where the 3D printed gearbox can flex a bit when it skips, I'd think with resin the gear teeth would just shear off and jam up the mechanism. That said, you don't have to give Smooth-On your firstborn child to get started with mould making and casting. Everything you need can be found at your average hardware store for pretty cheap: silicone caulk, plaster, epoxy (J.B. Weld might be interesting to cast with). I'm sure there are plenty of UA-cam videos that can explain the "how-to" of it better than I can.
"The gearbox knows its position at all times. It knows this because it knows where it isn't."
can you make servo using off the shelf gearbox and make attachment for encoder? that will be awsome
could u share which slicer and settings you use?
This is an awesome video. It unfortunately justifies me in buying more components though....
Hi that also servo from stepper motor.?
Those FDM prints look very precise.
are you using an unusually small nozzle and layer height?
I achieve a high quality FDM print on these parts with a .6 nozzle and .30 layer heights using a Prusa MK3S
Awesome !
Damn this gearbox is also a position
Hey does anyone know how to figure out how to fix a ender 3 from having thermal runaway errors?
AMIGO O BRIGADO POR DISPONIBILIZAR ESSE LINDO E MARAVILHOSO PROJETO. ONDE CONSIGO O DIAGRAMA ELETRÔNICO DESSE PROJETO ? NO DOWNLOAD NÃO TEM O DIAGRAMA ELETRÔNICO DESSE CIRCUITO QUE ACIONA O MOTOR.
FRIEND THANKS FOR MAKING THIS BEAUTIFUL AND WONDERFUL PROJECT AVAILABLE. WHERE CAN I GET THE ELECTRONIC DIAGRAM OF THIS PROJECT? THE DOWNLOAD DOES NOT HAVE THE ELECTRONIC DIAGRAM OF THIS CIRCUIT THAT DRIVES THE MOTOR.
It's very easy to make those rc servos turn a full circle like a DC motor. A quick modification and is done.
Why do you print with rafts??
I know this is really late , but what size nozzle are you printing with ? Is it a .4 mm or smaller to get the tooth detail for the gears
Great videos for only having 6.99k subs (sorry for bad English)
Do you have a wiring diagram or know of any info videos on it? I bought it based on this video and am struggling with it.
Great video and design, wonder if drone motors would work this way too?
nice work! subscribed...
Watching you hit your bed with Aquanet was painful to see, reminds me of my days of chasing down failed prints.
Get a flexsteel bed with a PEI sheet, I haven't had a print come lose ever since, I never prep the surface either
Even with nylon?
I've always hated that servos need to be controlled by PWM. Means the controller is constantly busy feeding it signals at precise timing.
What kind of spray do you use for your 3d printer ???
wow! congreatulations...good job..... question: ¿which program for make robot parts use?
What size magnet did you use?
Вот подобные видео я люблю.
Amazing
I am curious as to why you chose a mechanical switch in the gearbox assembly, instead of a tiny magnet and a hall sensor? The switch will wear out pretty quickly I would imagine.
If by quickly you mean in twenty years, then probably.
@@dieselguy62 Only 20 years? That thing could sit on a shelf for at least a few hundred before the metal bits corrode.
A gearbox is never there, Frodo Baggins. Nor is it here. It arrives precisely where it means to.
It knows this, because it knows where it isn't.
By subtracting where it is from where it isn’t or where it isn’t from where it is
I am very impressed by your servo and stepper gear box. Do you have a link for the bracket or arms used for holding or using the gear boxes?
is that an ender 3 pro or a regular ender 3?
Nice servo! Would really like to know your print parameters too cuz those prints look awesome!
Is there a tutorial for this any time soon?
Are you building a camera platform?
I was thinking antenna tracker
Do the two outer Rings have different number of teeth? I'm a little confused how the transmission is achieved because they seem to be the same size and and the caseing und output shaft therefore shouldnt move indipendently (or maybe i'm just missing something obvious ^^).
The large bearing looks expensive but isn’t really in bulk and 7N/m is 5.2 lbs/ft
Awesome. What's the approximate cost per unit, before labor?
very nice. im interested in speed... how fast can it be?
hi you video is amazing can you make a video about the basics about gerbox and thanke fore this informations i hope you can complete
Hi!
Do you have the code but using a stepper motor with the AS5600? Thank you.
How small do you think you could get one of these? They'd be incredible for tiny animations for model railroads or christmas village projects! :) Thanks!
Can you make a super small servo motor？
is that the alan wrench from a snow chain for a car?
hey michael, congratulations! are you planning to make any video evaluating if apply any load on it? like if it's easy to break
It took me a few moments to realize that you didn't actually mean to say "three-axis-turd", but still that's what I heard.
What Brim do you use? It's massive but I like it lol 🤣
What printer do you use?
It would seam to me that the motor could turn ot with less gearing, hence faster, with out losing power.
Cool🔥