Hoddem's DIY Linear Actuator

The insert for the rod end is PLA, every 3d printed part is PLA right now.
Not sure about the strength yet, have to test it.

I went with carbon fiber because It was readily available and relatively cheap.

OK, So it was a little after midnight when I finished bolting everything together and taking a few pictures so I didn't have time to add comments.

Assembly was easier then I thought it would be, but It definitely takes some coordination to get everything put together. Assembly jigs would be very helpful, especially to hold the actuator vertically at different phases of assembly.

I actually assembled the whole thing and then had to disassemble it because I couldn't get it to rotate. Turns out that the 25mm shaft clamp is capable of compressing the carbon fiber tube and the 20mm brass bushing to the point that it actually pinches the ball screw. The reason I added the bushing was to give some extra strength in this area, I may have to see if I can find a different bushing or even use an additional piece of carbon fiber tubing. I don't know that it will even be a problem, but I had to be careful not to over-tighten the clamp. This wont be a problem when the actuator is pushing, only when it is pulling.




I tried to run the actuator with my bench power supply which is good for about 5 amps and it was a no go. I looked around for other sources and found my kids powerwheels truck battery which is 12v. So I grabbed the leads on the motor and jammed them into the battery contacts and got a hell of a spark and the actuator jumped and took off. I did that a few more times alternating directions before I called it a night. It was hard to hold the leads and focus on how the actuator was running in the limited amount of testing I did so I cant really comment on how it works yet.

I am hoping to get my test stand put together tonight and get the actuator mounted along with wiring the Sabertooth and Kangaroo boards, but I don't have a good power source other then my car battery. I will probably order more of the DPS-600PB power supplies (I have 3 in my racing rig) for the test stand. I'm thinking about getting 4 of them so I can do up to 24V @ 1200W. Reason being, I'm not sure If i will stick with the 12v motor or change to the 24 volt motor. I also may just run the 12v motor at 24v and see how it handles it. a lot of unknowns right now, but I know I need a lot of power to run this thing.

http://www.ebay.com/itm/HP-DPS-600P...296081&hash=item33d479fd5e:g:VlIAAOSwom1Zi5Zw

I also got a chance to check the kangaroo board and the grounds are all shared so I can actually save two wires and use the RJ45 connector for everything.

My new wiring Plan

IEC320 #1
1 - 12V M+

IEC320 #2
2 - Ground M-

RJ45
1 - 5V - Encoder
2 - Ground - Encoder, Limit 1, Limit 2
3 - A - Encoder
4 - B - Encoder
5 - Limit 1
6 - Limit 2
7 - 12V Fans
8 - Ground Fans

If I use 14 AWG power wires it will be slightly more wire area then a 10 AWG wire which is what is on the motor.
10 AWG = 5.26mm^2
14 AWG = 2.08mm^2 (x3 = 6.24mm^2)

something like this

I'm sure you'll confirm anyway, but all the Kangaroo ground pins are common.

Limit switches, encoder, etc.

Yup I had a chance to check it out last night, Its good new because it saves me a few wires.

I built up the test stand tonight, It went together pretty quick. I made a couple of videos of some testing I did, no loads yet as I'm still jamming wires into my kids powerwheel battery. If you listen closely you can actually hear the spark when I make contact.





I ordered my 4 power supplies today and the parts to finish up the wiring. I wont have the power supplies until next week, but hopefully next week I can have the sabertooth controller running and start doing some load and speed tests. I'm hoping to write some arduino code that will put the actuator into a loop and just let it run under load to see what happens, if anyone has experience with that id gladly take the help.

It is louder then I expected it would be, but I suppose that is what you get with home built. I am also hoping things will wear in as I run it, I could feel some tight spots along the length of the ball-screw as I ran it up and down by hand before assembly. I believe that is variation in the OD of the ball-screw or wobble in the ball-screw hitting the bushing inside the 25mm carbon fiber shaft, the motor didn't even seem to notice though.

this is very laud indeat, much lauder than my current setup. i think the noise comes from the gear to the encoder.
i cannot imagine this comes from the ball screw. is it possible that the motor itself make this noise ?

With the rift on your ears it will be fine !

i don't think the noise is from the encoder gear, I think it's coming from the ball screw.

headphones would eliminate the problem, but with 6 going at once I fear I would only get one late night session in before I'm kicked out of the house for good.

there isn't that many moving parts so it should be fairly easy to find the source.

It could be the bushing riding on the ball screw, so I may need to find a different solution the that problem.

Cant you remove this damn bushing ID16 ?

could you try without the PVC? I don't know if'it' s create some sort of resonance inside maybe

great progress with your project.looks like a commercial actuator. if you ever change filament give PETG a try.easy to print and its bit harder than PLA .
this is a useful plugin for some testing without writing any code https://www.xsimulator.net/community/threads/output-a-sinusoid.10283/

I could remove the bushing and I may try that if I can find a good alternative. I don't know that I even need it, but I now know that the carbon fiber tube compresses a fair amount in the shaft clamps so I feel like it is a good idea to have something where the bushing is. I might try to drill it out, but that might be more effort then its worth.

I can't really test it without the pvc because the pvc is holding everything together. I suppose I could add a bunch of nuts at each section and just use the threaded rods, but it wouldn't have much strength.

before I tear anything apart though I want to get my electronics finished up and see how it acts at different speeds.

Awesome, that looks like exactly what I would want. Thank you for the help

I was also thinking that for initial testing without setting up a computer I could use the RC input mode on the sabertooth and the actuator will be just like an oversized servo motor. Then I can control it with my RC remote which would be pretty cool and very simple to setup.

If it's the PVC-tube that acts like a resonator, the noise should be eliminated/reduced by simple pushing the pipe together with your hands. This would add mass, reduce the vibration and hence the emitting noise.
Or your use your ears for identification. If it’s the ballscrew this also should be noticeable, because it is moving. But maybe you’ll nee a second person for that

as i really like this actuator i modified it for my usecase.
I used different type of tube materials with different diameters. also changed the mounting plate to an aluminum type.
The Main Shaft is made of polished steanless steel, the position sensor is an magnetic position sensor instead of an encoder type sensor.

Nice, I originally wanted to do a belt drive motor like that, but I figured I would try to keep it simple for the first go around. Seeing how tall this thing is I think that will be the way to go if I can keep it cost effective.

Amazon and eBay left a few packages today. It should be everything I need to get the power system all setup.



I don't think I will have time this weekend, but maybe I will pull a few late nights this week.

Hey @Hoddem - is that a shunt in that photo? If so, you measuring total current or on an actuator by actuator basis?

Looking damn fine, btw!

@Zed, you have a good eye. I didn't know if anyone would look close enough to see that. It's a shunt for a 100a current meter. It's a digital system so I'm not sure how fast it can react, hopefully fast enough to see the peaks. It also tracks total power so I can see on average what my power output needs to be. I'm planning on using it for the one actuator to get an idea of how much power I need and verify if my wiring is ok to handle the current.

Found this on Craigslist, it's got me thinking