Hoddem's DIY Linear Actuator

Starting a new build log for my version of the linear actuator design that is floating around here. I have read a few of the build logs (@SeatTime, @SilentChill, @wannabeaflyer2) and they are great. I kind of took it as a challenge to see what I could do from an engineering standpoint and see how it would compare. This may or may not turn into a 6dof simulator, but right now its just going to be about building a linear actuator and documenting every step of the process.

Some directions that I am planning on going
Sabertooth 2x32 or 2x60 driver with kangaroo motion controller
quadrature encoder for feedback
1605 lead screw and ball nut with optional 1610 for more speed
the largest motor I can find, afford, fit
as many chineese linear components as possible to keep the cost down
carbon fiber main rod and anti rotation rods
Extremely compact size, I would like the whole assembly to fit inside a 4" pvc pipe.
probably a steering U-joint for the bottom and a swivel rod end for the top.

Up in the air right now
12 or 24 volts
direct drive or belt drive for a more compact design.

Here is the kicker, I want to 3d print all of the custom parts and minimize custom modifications or machining.
I want to use IGUS bushings, but if they need to be purchased the cost is too high. I have a solution I will share in the future.

Open to suggestions, but keep in mind I can be stubborn and enjoy failures almost as much as success so if you tell me 3 times its not going to work don't get too upset if I push forward anyway.

Here is V1 of the design, Using guesstimates for placeholders for a lot of components.




And I started working on V2 with actual dimensions and adding detail like screws, its only about half done right now though.



And because why not, I did a few renderings. The files are too large to post here so you can view them from the link
https://www.dropbox.com/sh/6z5bhx2myj4muaz/AACm_aAXMw2N9tswby7LtlWOa?dl=0

This Design has a max stroke of about 15.5" and the overall length is TBD, but right now its about 32" without the u-joint.

Once I get my design finalized I will share a full BOM including purchase price/location and complete drawings and 3d print files.

Teaser Shot, I pretty much committed to the project about a month ago when I made my first order of Chinese parts on Ebay. It took a while, but they showed up today and now I can verify all my 3d model dimensions.

Great work @Hoddem !
Your renders not only are great, but also gave some solutions and also simplifies a lot the design, I was thinking about the same in some aspects of the actuators.
Thanks!

@Hoddem I'll be following your journey as I too am about to embark on a linear actuator build having been inspired by the likes of @SeatTime , @SilentChill and @wannabeaflyer2 who have kindly shared so much of their work.

Currently my plan is only for a compact seat mover and just 2 DOF to begin with, but I'd like to think that this could be extended sometime in the future.

Due to the long lead-time on China ball screws I've already ordered 4 x 1605's complete with end bearings. I too was wondering if maybe 1610's would be preferable for increased speed but figure it would be easy enough to retro fit if required once the actuator is built.

In my current design I'm looking to use 70mm x 70mm x 2mm aluminium box section with a 70mm x 25mm x 3mm aluminium channel for the drive belt. The estimated total length of the actuator is approx 285mm.

Here's the preliminary design....

I've sectioned away the wall of the aluminium box section so that you can view the internals, anti-rotation will be accomplished using a simple wear strip and instead of an Igus flanged linear bearing/bush (which are hellishly expensive!) I plan to use an oilite bush similar to pictured below...

Not sure exactly what size I'll require but they come in a variety of sizes as standard and are very cheap, for example Ebay have them for a couple of quid...
http://www.ebay.co.uk/itm/OBF182512...325951?hash=item1e83630e7f:g:b6AAAOSwMmBVyzTi
I'm sure they won't be as efficient as the Igus bushes but for my needs I think they will be fine. I guess time will tell?

I'm currently deciding how to proceed with positional feedback, I quite like the idea of a linear transducer such as a linear resistor or magnetic encoder.
Linear resistors are dirt cheap but probably not up to the job...

https://www.coolcomponents.co.uk/en/softpot-membrane-potentiometer-200mm.html

And magnetic encoders look ideal, but whilst the tape is cheap I'm yet to find a reasonably priced reader...

https://www.machine-dro.co.uk/high-...ar-tape.html?gclid=CK_P7PePh9QCFUq3GwodYb0G8w


I'm also very limited on machining tools and so my design has to take this into account.

I won't hijack your thread any further and will start my own thread once parts start to arrive and I have finalised the design.
Good luck with your build, I wish you success and will follow your progress

That's Awesome, the more the better. I too started with a compact design using pulleys and belts very similar to yours. I ended up deciding to go direct drive for the first design just for simplicity, but at 55" extended, I fear I will hit my head on the ceiling. Once you start your thread I will be sure to watch. As I said, I think I have a novel idea to get around paying such high cost for IGUS bearings. stay tuned!

Also, What software are you designing in? I am using a lesser known Geomagic Design (Formerly Alibre Design) and for rendering I use Keyshot. I have a copy of Fusion 360 installed and ready to learn, but I keep going back to what I know.

I got a little further on the design tonight. The encoder is now mounted and more hardware has been installed. I still need to get limit switches figured and I have a u-joint on order so I can model that once its in hand.

Here is the actuator without the covers



and here is the actuator with covers



I have uploaded the progress so far to grabcad, anyone who is interested can view the model in 3d right from your browser in grabcad.
https://grabcad.com/library/hoddem-s-diy-linear-actuator-1

The main design points in building these types of actuators are now quite well tested/documented, but your time/money. 4inch PVC pipe should be doable, as the covers on my actuators are 80mm dia. The Igus bushes I used are just bushes and did not include the mounting as per the ones SilentChill 'acquired' at a obviously real lower cost. I manufactured mounts out of carbon fibre to support them. For Carbon tubes I would be careful of using anything else, as it will/could wear/damage the tube - Note. the final design points for these actuators flowed from real testing and allot of time/effort - be careful if you stray. I for one will never get angry if you come up with a better solution, or just enjoy failure - it's your money/time/effort, not mine.

Everyone knows that your actuators are the premier version (seriously, excellent work), but they for the most part out of reach for the average user. I have done a little composite material work way back in school, but no way would I attempt it now. I also understand your concern on the carbon fiber rods and potential wear/failure, I guess I will have to share my plans for that to put everyone's minds at ease. My aim here is to look at it through my own eyes and say how would you do that given the task, obviously I'm not taking ownership for the entire design because it is derived from designs here like yours. As far as my time/money and effort, I have spent hundreds of hours, thousands of dollars and countless effort on my current simulator and have actually logged about 4 hours of seat time. That should tell you which side of this I enjoy more.

Point of no return, I just pulled the trigger on the last few components, Encoder, Limit switches, Sabertooth 2x60! and a kangaroo board.
The old wallet is a little thin right now, but all hobbies are expensive right.

I Promise I will get all of the details up eventually, including a full BOM with pricing and suppliers, but for now I just want to push forward.

Do you think the 4 external rods would be needed? I think that with the 2 carbon guides and the pvc pipe could be more than enough.
Edit: Man, that exploded 3d view of the actuator is awesome, a picture is worth a thousand words!

The way I have it setup right now, the threaded rods are holding everything together against the PVC cylinders. If I remove the rods then the whole system will fall apart, If i remove the PVC I could use only the rods by adding a bunch of nuts at each stage.

I'm still not sold on the PVC idea as is, I think its going to be a real challenge to assemble and troubleshoot. I will probably cut the PVC sections in half to allow removal without fully disassembling, or scrap the PVC idea and change the threaded rod to a larger size like 1/2". I really like the idea of an enclosed and system, I can just imagine someone getting a finger inside one of these while its running.

Any news ?

Sorry, I've been focusing on finishing up my current racing simulator. But I will share some of my progress so far very soon.

ok, so here is my theory on how to save a few buck on the IGUS linear bearings.

I was snooping on the IGUS website looking for anything that referenced carbon fiber shafts and I found this.
http://www.igus.com/wpck/17114/carbonfibertest

and they even make the j260 material in 3d printer filament for prototyping.
http://www.igus.com/product/716

So I was looking at the 3d print filament and the little chat window pops up and next thing you know I have 3 of these little packages at my doorstep (bad picture, the other side says j260 on it).



So here is my theory, I purchased some cheapo Chinese flange linear bearings and I want to replace the guts with a custom printed bushing. That should bring the cost down significantly over the complete flange bearing direct from IGUS, but I still get all the benefits of the IGUS material.

The first thing I had to do was dissect a Chinese bearing to see what the guts looked like.
This is a double bearing, notice the wiper seals are actually molded onto a type of snap ring that holds the cage in place. its pretty difficult to disassemble these without destroying the wiper seals.



I measured the little bearing cage and did a little 3d modeling and printed up a part out of PLA, it took a couple of tries but I was able to get the dimensions pretty close. I actually went on the IGUS website and downloaded the 3d model for the 12 mm linear bearing and copied the inside profile over to my custom part, so now I have the same inside dimensions as the igus bearing.








Here are the actual bearings that will be used in my actuator disassembled and ready for new bushings.



And the new bushings ready to be inserted. These are still PLA for now, I want to get the design perfected before I use up my sample material.



I ran my carbon fiber shafts through each bearing and they were a little snug at first and then loosened up nicely, I suspect that the pla will wear very fast though based on how easy they loosened up. I am using my bench vice to press these into the housings, so no retaining clips are needed. Also the tolerance on the carbon fiber is not nearly as tight as a linear shaft, I can actually feel the high and low spots throughout the shaft when i run the bearings across them.

on my actuator design I need to flip the ball nut around, so I was digging for a piece of tube or something to keep the balls lined up and it hit me. I quick printed off an adapter that matches the ball screw so no more worrying about losing the balls. I printed this at .3 as fast as possible so it looks a little crappy, but it works great.



I was able to easily remove the nut and re-position it.

So I started assembling the guts of the actuator, I had finished the 3d printed parts a couple of weeks ago.

Here is the ball nut mount with the ballnut and the two 12mm linear bearings (just the housings at this point).



and the 25mm shaft clamp that the carbon fiber rod will mount into.



I mounted up the parts for the upper an lower plates and then began assembly (the bushings have been pressed in at this point).

First the ball screw



then the ball nut



then the 12mm shafts



Then I inserted the 25mm carbon tube, but before I put it in I added a bushing to the end that will be clamped to give a little more rigidity. It was a snug fit, but not overly tight. the ID of the bushing is 16mm so it matches the ball screw, well see if this causes any issues in the future. also I may add a little bushing the the end of the ball screw to prevent any potential whip inside the carbon fiber tube, If so I will print it out of the same J260 material.






And here is the finished assembly, of course I don't have any pvc pipe on or any of the threaded rods.



I grabbed my cordless drill and ran a few strokes, everything was very tight the first 2-3 passes and then It loosened up and ran pretty smooth. I am pretty amazed at just how strong the assembly is without the pvc or threaded rods. Of course I still need to get the motor setup, but this will give me enough to do some preliminary testing with my drill. Also I have a pretty cool Idea for how to test this with the pvc installed and still be able to see whats going on inside.

I made a few videos of the early test runs

Here is just running it back and forth with no load, the drill is running at 2000 rpm max.


Here is a load test, superficial because I am using a drill and not the actual motor that will be installed. but it gives a good Idea of how strong this thing is mechanically. I'm 210lbs for reference.
I ended up pulling out the 12mm carbon rods, they were not meant to hold the whole structure together so I expected as much to happen eventually. With the PVC and threaded rod install though, this thing will probably be bullet proof. Again I am just amazed at how strong the carbon fiber is.


I'm thinking of ordering a 10mm pitch ball screw just to see, it seems a little slow right now.

I really like the look of your actuators matey they look really good

3D printing the bearings is genius, totally thinking outside the box there !! and the little ballscrew insert is a brilliant idea that would of saved me loads of time when I screwed mine off lol. If only I had half a brain

That's some nice work - yep I have a car modding background so was very aware of how strong and light carbon fibre can be if used properly. I'm still a bit old school - so most of my stuff is handmade, but you have got to love 3D printers - well done.