Hoddem's DIY Linear Actuator

Looks and works great! How incredible tool is the 3d printer for Diyers!
I'm getting my actuator parts made in delrin, possibly with teflon bushings, still getting quotes (the first one I got was about us100 per each lathed part...) I'm still trying to do it my way, but the costs are getting too high... Anyway, your 3d parts will be of great help and a great advance for the future actuators made in the forum.
I will use polished stainless steel center shaft and 2 solid 12 or 16mm hard chrome shafts with threaded ends, it will weight a ton, but I don't think I will be traveling much with the sim

Thanks, If anyone wants the file let me know. I will post it eventually, but I generate my files in high resolution so the file is about 125mb right now. if anyone wants it I can reduce the file size and upload it.

3D printing is great, but just like any new technology you have to learn how to use/apply it. I still print things several times to get it just the way I like it.

I have never really worked with carbon fiber, but wow. I ended up buying a 25mm od x 20mm ID carbon fiber tube for the main tube and it is so light and just rock solid. It will be interesting to see how the actuator holds up under a side load.

Go with what you know! my build reflects what I'm most comfortable with. I debated using hollow linear shafts but they are ridiculously expensive, so far I'm looking at $365 per actuator and that number will probably climb a bit. it really starts to add up, especially if you end up with 6.

Don't forget the 'Achilles Heal' of all simulators - inertia - anything that can move needs to be light and rigid. I initially had steel main shafts in my DIY actuators (see my thread) but decided to start again and went to allot of trouble minimizing the weight of everything on my sim , even rebuilt my frame and pedals out of carbon fibre and it really does pay off. Just like a race car - the lighter it is - the quicker it is.

You're right @SeatTime, I didn't thought about the inertia in the center shaft, and the one I got for the prototype is quite heavy.

I printed off the adapter for the swivel to mount in the end of my carbon tube. I added a nut at the top to stop the swivel from migrating into the carbon tube, I may add a washer to give a little more surface area. I'm making the assumption that the actuator doesn't need a lot of pull force, but I will test it to see what its capable of.







I'm not sure how much help the shaft clamp is providing being that the carbon tube is so strong. The adapter is a snug fit so I would think a little epoxy may be all that is needed or I might cut a slot in the carbon tube to allow the clamp to provide some squeeze on the adapter.

Why do you think a softpot is not up to the job. I've looked at these and think they would work well with a spring loaded wheel roller (maybe the same as used for sliding window screens) attached to the ballscrew. The softpot would be taped to one of the side supports. A rolling wheel would minimize abrasion and wear on the softpot -- the company rates them for 1 million contacts. Its a bit more involved than an encoder but about the same as a gear reduced potentiometer or hall sensor.

It should provide position easily. The only unknown is how fast or latent is the position info as well as accurate -- i would think it instantaneous but it may take a while for the value to settle. I plan to try test this out in a few weeks using an arduino.

My concern was not with the reaction speed as this should be almost instantaneous, it is after all just a variable resistor. My worry is that it will wear quickly due to the speed of the actuator and because many of the motions will be small back and forth movements mimicking road noise and bump strips. I can imagine that this will be very harsh on the linear soft-pot. But I'm still designing my actuator (trying to find standard parts that require little or no machining as I only have basic tools to work with), and so I've not ruled out the use of a linear soft-pot just yet.

It should work, however I have seen in other threads where calibration wouldn't complete when using the sabertooth/kangaroo combo and a potentiometer. It will be interesting to see how it works out for you and if it ends up being a viable solution at a lower cost then I'm sure everyone will appreciate the effort.

I had a little time over the 4th of july week to keep playing with my actuator design. Most importantyl I had some time to work on the 3d printed bearings and learned a lot of hard lessons. I started off trying to print the J260 material (which IGUS lists as the most dificult to process) on my monoprice maker ultimate, absolutely no go. I couldnt get the material to flow very well and it wouldnt stick no matter what I tried for a build surface (build Tak, blue Tape, Glass etc.). I switched over to my prusa MK2S and struggled until I pushed the temperatures to the absolute maximum of the prusa (295 Extruder and 115 Bed) then things started working a little better. I was able to print all three bearings, two 12mm and one 25mm using the three sample packs of material IGUS provided.



The parts came out pretty rough and required a ton of clean up, also I had to print a lot of them to get the fit just right in the bearing holder and on the carbon fiber shafts. Also I would print two identical parts off the same print file and the fit would be completely different between the two. the parts just dont seem to hold the same precision as the pla parts I have made in the past. All in all I think I will give up on the idea of 3d printed bearings, because most printers cant handle the temperatures required to print this stuff and that defeats the purpose of the project for me.

The parts that I did make will work and look pretty good in the end, but I wouldnt want to got hrought this process six times over for a whole rig build.


To get the 25mm bearing printed I actually had to print it in two sections becasue without an enclosed printer the part would cool down and eventualy seperate from the build surface.



So I spoke with IGUS and disclosed my frustration with the material and of course they recomended using their flange bearings (the same ones most guys are getting free samples of) but I want to keep the project cost down and also not have to rely on free samples. If someone wants to repeat my buid and can't get free samples that would be several hundred dollars extra. In the end IGUS is sending me standard poly bearings in 12mm (RJM-01-12) and 25mm (RJM-01-25) to test out. I will be redesigning my system to eliminate the chinese flange bearings completely and print my parts to house the IGUS poly parts directly.



This is a bit of a setback as I had all of my parts for that secion of the actuator completed and ready for assembly.

I finished my 3d model utilizing the IGUS polymer bushings instead of the 3d printed ones. I will start printing everything today, but it will probably take a week to get all the parts made.
Here is the link to the 3d model on GRABCAD
https://grabcad.com/library/hoddems-diy-linear-actuator-r2-1

here are a couple of shots of the new design.

A couple of pictures of only the 3d printed components.

I got a nice little surprise at work today.



They came in a huge box which seemed like a big waste until I found this inside.



Time to get printing the new housings

There are a number of references like that that are gold mines for information. The Omega Engineering ones are great for all sorts of measurement, heating, etc. (If anyone has a pellet heater or grill and having a hard time finding reliable cartridge heaters to light the pellets, best place around!) Small Parts used to be the same for motion control until Amazon bought them out.

@Hoddem, those are just very beautiful designs!

Bom, então eu entendo que você usará o motor DC, o motor stepper não será melhor e mais preciso?

I once got a set of books from SDP-SI, if you are ever looking for small components like gears and pulleys they have an absolutely amazing selection and I have used them for many projects in the past. I also received the highly coveted McMaster catalog this year, they are highly sought after and often go for a bout $50 on ebay. I usually sell my old ones whenever I get a new one.

It's fun to get catalogs like this, but with the online resources getting better I rarely if ever pick up a paper catalog anymore.

translated in google
Well, so I understand that you will use the DC motor, stepper motor will not be better and more accurate?

In this case I need the speed and simple operation of the DC motor, by adding the encoder I essentially have a low cost servo motor. It should operate at a much higher speed then a stepper motor and with the encoder be fairly precise. I had the same thought initially though, I have a lot more experience with stepper motors.

I will be using a sabertooth motor driver along with the kangaroo controller
https://www.dimensionengineering.com/products/sabertooth2x60
https://www.dimensionengineering.com/products/kangaroo

First part off the printer (since the design change).



The bearings are designed to be pressed in and are pretty loose fitting until they are pressed into the correct housing size. I was a little worried that the 3d printed part would give and the bearing wouldn't be pre-loaded correctly, to my surprise the fit on the 12mm shaft was about perfect. I just used my bench vice to press these into place and they went in pretty easy.

I'm printing the housing for the 25mm linear bearing now, if it works out this is a much easier method and cost will be about the same because I don't need to buy the Chinese bearings or the 3d print filament.

These bearings are only about $6-7 each.

Sorry I completely forgot it was in English, kkkkkkkkkkkkk.

A friend made a stepper motor, it was very good the final result is very precise, but a high cost.

I'll wait for the result to be yours, I really liked your idea and maybe it will cost you a lot.