Linear Actuator Design 

Not only is the design genius, the presentation couldn't be more clear. Thank you for posting.

Really hidden tallent with so much depth in science yet unrevealed in public. National geographic should approach.

I've watched this multiple times. It has to be one of my favorite design/build vids. Content and presentation are top notch! Inspirational for engineers and teachers alike.

As I have myself started with designing and manufacturing mechanical systems, I can only imagine the work and experience that was put behind making the system work so flawlessly. Thanks for sharing.

Only found ur channel after catching ur LATHE build. The Mill build to go with the Lathe was phenomenal too my friend. I'm attempting to build my own DIY CNC now. I tried about 18 months ago but didn't really understand the whole process and failed to get it right but now I'm going to have another go .. I haven't watched but 3 of ur videos so far but the ones I have are so interesting to watch. The information is brilliant pal. I hope ur still making these machines today. I will be looing through the channel to see more.. Keep up the great work man..

You Sir are brilliant. The design is simple yet seems extremely effective, the perfect engineering solution for a home hobbyist. Thanks for taking the time to share your work, I have spent a few hours at your website absorbing all I can.

The use of bungie-cord to pre-load the system is an excellent idea to take strain off the motor. I'm surprised it's not used more in robotics for neutral-positioning.

That's what I call superior skills. This video really is super hard to do too.

That was very, very interesting. Thank you. Also, I commend you for not showing yourself screwing in every individual screw. Some people actually do that...

Excellent work, and very inspiring -- also a very fine presentation. I am a retired defense engineering programmer/analyst. It's been a while since I've worked on 6 DOF modelling. I don't think I could do the math right off the cuff, but I do understand the control principles. I think it's amazing that the parts for this only cost you $300 - esp the pulleys. The bungee's to neutralize load is brilliant. The only thing I do not see in the design are damping elements. I realize you can control the acceleration via the pot feedback and software - which makes me thin of an elevator, in principle - but in your case, I bet some mechanical damping would cut back jerk and snap for smoother natural operation. I wonder what your thoughts might be on this... It's refreshing to see individuals actually practicing systems engineering on their own projects. A toast to your ambitions, Sir !

Possible considerations for similar designs. 1) I agree with another comment that the potentiometer should be replaced with an encoder to avoid drift and EMI from the motors while reducing oscillations that are likely to occur under high dynamic loads. This can also be addressed by implementing acceleration curves though I suspect the use of belts was intended to suppliment this requirement. 2) By high mounting (instead of under mounting) you run the risk of a collision of your platform with the legs. 3) As you noted, bungees (springs) do not provide constant force at all positions. F = kx, where F is the force applied, k is the spring constant and x is the extension. Further excursion = reduced weight compensation. You will have zero static load at exactly one excursion point, any deviation from this point will result in a required holding force, though this appears acceptable in your tests (probably due the high ratio of the length of bundee to the traversal length of the acutator). 4) You will likely have difficulty maintaining a level platform unless you include belt stretch compensation for holding forces by measuring the load on the motor, either through a load cell or current draw. This is because your center of gravity will shift as the platform is moved around. Any 2 legs should be capable of acceptable actuation with the entire weight of the platform plus an (arbitrary) safety factor above the expected dynamic loads. This seems like a pretty good home setup where minimal excursion and minimal centroid deviation is required however I suspect it will not work well over the full possible traversal (motion envelope will be small especially fore/aft and laft/right). This is mostly due to the high mounting of the actuators relative to the platform. To experience full excursion, your footprint will have to be quite wide (instead of tall) at which point a larger deflection will be required by the actuators to effect the desired roll, pitch or yaw. Vertical motion will be unaffected by footprint width.

I suggest not using potentiometer unless cost saving is really important. Better to have optical absolute position sensor or simple optical cog-type sensors and zero index (which you already have with your end stop switches). Overshoot can be computed for over various loads if current sensor feedback is used to estimate inertia / mass of load so that servo rapidly slows as destination position is approached. Very cool project - thanks for sharing

I won't be building my own 6dof full motion flight simulator any time soon but it is on my bucket list so I'll definitely be filing this video away for future reference.

Excellent! While I don't see a Flight-Sim in my future, this opens up a whole raft of possibilities. Thank you, John.

You are really cool!!! Machines like you building cost a millions of $, but you make it by youself!!! Amazing!!!

Amazing design, the thing I really like is you can move it very fast or slow, and the materials used in its construction (thick plywood) is actually very strong and easily fabricated and it can move 40kg. This would be amazing in a six axis flight sim.

Kudos to your designing and engineering abilities and especially to your mental discipline and energy to follow through to completion. I truly envy your motivation! Had that once upon a time, but lately that spark has greatly diminished...

your craftsmanship is phenomenal and I enjoyed your thoroughness of design and explanation.

Nice work. You have put a whole lot into your design. Optical encoders can be used in place of the POT. You don't have a proportional feedback control on the shaft. You can add one just my printing it on your printer and an optical interrupter. Also, the groves in the timing belt is deep enough to be used as an encoder. The valleys of the timing belt can be painted white and a forward facing or reflective interrupter can be use. Nice work

This is aka, a SUPER powerful printer-style linear actuator.

Pretty spiffy. I went ahead and downloaded the file and added a cross bracket to the motor mount to minimize the tilt issue. I will work on it a little more this evening and e-mail you a copy within the next few days. Brilliant configuration, especially for a wooden case as opposed to aluminium.

Vintage mechanical keyboard, That's the AGE of keyboard that made me want the one I have today.

Your videos are like an open learning program.

Thank you for sharing your project. I will use this as an example of how to document a project for presentation. It is a skill that I need to practice at.

I am sure i can come up with a better design...but i am glad to know intricacies of other people's design.

Holy crap this dude is some sort of genius

To stop the tilting motion just add a 1/8' thick backing plate to the rear or the motor mount. Then use a top half wooden or metal bracket the fits the diameter of the motor and bolt it too the bottom half. Or you could just make the motor support block wider. And then bolt the top half over the motor support block.

Those smaller belt wheels look like ones on car water pumps that are usually discarded when changing timing belts. They can be easily pushed out with shop press. Those would be almost free compared to €14.50 a piece. Also car timing belts could be used for power transmission, even used ones if going really cheap.

Bungee as counterweight is genious idea, but aren't you worried about rubber degradation? Would same design be possible with springs?

Looks incredible Roland! I hope this project is still going strong!

mount a potentiometer on the bungee L bracket, and record position of L bracket. Use a PID loop to eliminate error over time. Great project overall.

Very well done! Thanks for putting up a great video and more importantly, for sharing your knowledge. It is appreciated. By the way, don't let the envious trolls get to you.

My father built a linear actuator with chains in a somewhat similar way, except that the output of the actuator was on the side of the cogwheels, rather than on top of them.

I don't understand any of this stuff, but it's just so ridiculously cool!

Amazing! I think that works like faster as Diadic acotuator but, with 10 times more power and course. Congratulations!

I used normal AC mains wire for the motor leads. It's normally good for 20A or so. Since average current is low, it is not a problem. The motor controller is my own design, see my website under Motion Platform III and IV "electrical drive" but commercial motor drivers like MD03 are fine too.

Хорошая работа! Я думаю что при использовании косых шестеренок шума было бы меньше.

Very nice implementation, Roland! Isn't it funny how the "haters are gonna hate"? Apathetic spectators of life are full of "better" ideas - but no action. Keep up the excellent effort. I'm off to visit your website now.

Great job. Why so many Dislikes - not cool. Never hesitate to share you work!

Thanks for posting Richard.. We all appreciate it. Pretty cool..

Great inspirational video, now I am so glad I did not throw out that old Nordic Trac ski machine as I want to take it apart and tinker

Wat een Nederlandse accent zeg, ik kon meteen horen dat het een Hollander is. Voor de rest een prima project.

So much better than free to air news thx for sharing

Great project, look forward to seeing the final build!

A motor driving a worm gearbox coupled to a lead screw is also a viable solution with the added benefit of anti-backdrive.

Скрупулёзная, тонкая и очень дотошная работа, чтобы всё было на высоте! много крутых и точных деталек скоро будут выполнять свою работу! фанера конечно выглядит немного хламовато, ну да может быть она будет покрашена.

I like the use of bungee cords. Nice going!

I noticed some tap holes on the motor front plate. Maybe you could use a bracket to prevent the motor moving forward by utilising those taps. However it would exert a great amount of force on the screws holding the clamp.

This is extremely cool. Way to go on the whole design!

Man there are some clever buggers out there Nice Work

I enjoyed seeing the current during testing ty!

Wow! All I can say is ... "BRILLIANT!" Amazing job.

Thats impressive and functional. Just a thought: in my experience, bungee cord doesn't last long under load. Most rubber seems to oxidize at a very accelerated rate when left in a stretched condition.

serious comment, nice actuator, if I had the time I would love to build an adjustable computer desk using a slightly toned down version. I love the speed though. I hate waiting for the desks at work to adjust. this would be like going from a sitting level to standing in 1 second. Just need to remember to not leave my coffee on the desk before pushing the button. Might need to gluew the keyboard and trackball in place as well. Seriously I hate waiting that much, although I might add in a bit of a reduction in speed. Move up about a foot in 2 seconds instead of instantly.

Excellent! Here is a man of great mechanical mind. If I may be so bold as to offer a suggestion. I noted on the bungycord "Eye Bolts , that an extra force has been applied. I am speaking of the friction between the eye bolts and the bungy cord when it passes through the eyelet. I do wonder what the result would be if you were to add " pullies" where the bungy meets the eye? Hmmmm! Maybe you could achieve a smother operation. I will stay tuned to see if that idea has merit. Good evening and good luck.

Thanks for choosing components from Taiwan

you do it like a scientist. thanks

This guy is a genius

Excellent vedio, the highlight for me are the socks.

Wonderful! Very helpful Waiting to see more from your awesome lab!!

My random foible is to avoid micro switches, potentiometers and DC motors - in favour of stepper motors with optical encoders (I won't dedicate the motors / drivers - but make them 'snap in' - so I can move them between; simulator, parallel robot, plasma table, etc.)

impressive, beautifull, and AWESOME actuator. congrats !

Great Project ! I will suggest to modify design to be able to tension belts. This will let you avoid tilts on stops. as well will improve longevity of your belts in general. Probably you will need to buy sonic tensioning meter for this and, unfortunately they are not cheap :( .

Great thinking and most simplistic execution...Superb..!

pure genius! very well done - thx for sharing.

Very cool. You should make yourself a patreon or something and get some money back for all this cool stuff you're putting out.

What a great job Sir. I'm amazed about the construction

Very nice job. Thank you for sharing!

This is a wonderful, well explained and well thought out video.

Beautifull job . I am just wating for the whole thing being built.

Thank you for the video Roland. Enjoyed it. I like your work, quite a lot.

Very Well, Built - Good Job from Chicago

Nice job. Clean and Efficient.

A+ for ingenuity, D- for application. Hydraulic / air cylinders are much cheaper, easier to use and more durable. The cost of a decent pump + control valves / hoses would be off set by what you save on the 6 actuators. Sorry for downing your design as it is very interesting and shows a good deal of thought and effort in its design. But not the design I would choose for its intended purpose.

Thank you for sharing in such detail!

heb je de stop sensoren in serie geschakeld? paralel is toch makkelijker?

I would like to see and long-term evaluation of the resiliency of this design

I think you may find you have a bit of a maintenance issue with your teflon bushings. The 3D printer community experimented with bushings instead of bearings for a while, and found that they just didn't hold up to the repeated motion, even with the tiny loads a 3D printer presents. Of course, replacing those cheap teflon bearings every now and then might be a pretty good trade for not having to buy 12 expensive linear bearings.

Super .. Thanks for sharing the video.. Can also be used to lift a bed..Bed in ceiling My best regards to you from Patagonia Chile

Great project Roland ! Keep it up...

i wish this man was my father

Fantastic video! Very informative. Very well designed.

Thank you very much, very informative, verycomprehensive, very systematic.. bravo. thanks a lot... loved it..

Thanks for sharing this video and welcome to Taiwan

Thanks a lot for this explicit video !

Mr. Roland, I salute your Intelligence!,Thanks for sharing

Basically 6axis micropositioner, but without pistons. great.

excelente you did a good job! most people must be like you....

Wow, this was a very good quality video.

Nice design! I'm not sure about the bungee cords though. It wins regarding low cost, but some surplus air cylinders with a self relieving regulator could provide a more consistent counter force. Your actuator design could be used for a number of things. :-)

Very nice, would love to see the final simulator setup in action with these actuators in motion.

I was just thinking about designing one of these, cool design but I have yet to finish the video

Awesome work Roland.. it seems the two belt tensioners for the motor drive belt could introduce slack over time and the eyelets, where the bungee passes through would tear the bungee over time (like my broken shoelace : ) ).. in my huble opinion but aside from that.. excellent work and very inspirational thank you!!! Subbed and God Bless my friend : )

Very Cool Project ! Thanks for Sharing !

Have you considered using pneumatics instead ? You could replace all of that with an air cylinder and a proportional valve. If you're lucky you can find some used ones on auction sites

Please, more videos like this..!

Excellent amazing what you can do at home when you put your mind to it, thanks for the presentation.

This is such an inspiring video for someone just starting out. Can I ask some advice please? I am self-teaching control theory and electronics, some 12 years after I graduated with a mechanical engineering degree. Our control lecturer was sadly not the most ... understandable. I have a few projects I would like to create that will require control loops with motors and rotational plus linear movement, and be able to respond to external forces. My maths is now rusty too - do I really need to understand all the integrals of "e to the minus ess tee" laplace / fourier / matrices, or will something like matlab simulink handle all that for me? I am not averse to putting in time on the maths if its truly vital.

I'm a little bit confused about the bungee cord. What role does it play in the function of the actuator? Really cool stuff thanks for posting

Congratulations! Excellent work.