Hoddem's DIY Linear Actuator

1Nm with a 1605 ballscrew gives more than 100kgf but there is resitance in the actuator, it,s a safe number.
Dont look to much at peak torque, the motor draws so much current it will burn out with in a minute.
The Sabertooth 2x32 got soft current limitting you can set wished peakforce in controller and keep heat under control.
More than 3G can become harsh
regards Ad

Hi @adgun nice design in the making, and i take it that the anti rotation / Carbon tube holder is Profiled with just a small clearance to suit the internal profile of the extrusion . ( nice touch )

Will you be using the AMT103/n Rotary encoders i think u have mounted to the rear end of the motor shaft ... only mention this because this new design is very clean in its make up and assembly as well as the added safety of no finger traps, as all the moving elements are enclosed ... Very nice concept , admire your work sir cheers for posting

Well not a ton of progress on the actuator project this summer, I have been slowly ordering components over the last few months and I am happy to say that I will have the last few parts withing a week or two. Here is my makeshift assembly station.





I have already done a full test assembly of one actuator, so once the last few parts are in it wont take long for final assembly.

In the mean time I am slowly tearing down my existing rig and transferring everything to a new VR only setup. Here is the steering wheel assembly before and after, lots of weight reduction now that I am exclusively running in VR.









Finally here is the concept for the new rig (minus a seat), as you can see I have left the seat mover attached so if all goes well I will have a 3dof chassis with a 2dof seat mover on top.



I should have a full build guide for the new actuator up soon.

2dof motor placement, Imo you might want to raise them to above or Inboard the main frame rails. I see you have the SFX kinda mounted high, You ideally want the COG pivot as Low as possible and unfortunately you'll loose the room for the 2dof Motors.

Im just now finalizing my 2dof + SFX setup. I look forward to seeing how yours comes together.

Ya, like you said if I go much higher I will crash the seat mover motors. The seat mover design is carried over from before so its a proven working design that would be easy to use for this build. I still have a lot of decisions to make though and most of them will come through testing. The really great thing about the extruded aluminum is that I can just move stuff around as needed. If you notice the frame is pretty long (72 inches), compare that to about 55 inches for a simlab P1-X. I am purposely leaving the frame long and planning on condensing it down after I decide on final placement of all the components.

Are you giving up traction loss on your sfx100 build?

Ya i went a tick longer with my base rails as well, Copying over my Previous dimensions from my 2x4 Rig, I also might be moving my motor's from current position to behind as im not sure im diggin the motor levers flopping up when no power is applied, I did opt for 50:1 But that they do No Lock out like many assumed when No Power is applied, Yes they are much stronger vs 25:1 in that regards but can easily still be moved with the forces at play on the End of the lever.

Traction loss i decided i Had to get the base SFX and 2dof up and running and will look into TL next. I have No intentions on giving up the TL as much as many say they dont miss it, I drift and so do a few buddys who frequent the rig often and it would be a shame not to have that tail wagging feeling. Im not sure the 50:1 Gear motors up to swinging the BEAST that is this new rig, but if needed i'll look into other motor/ballscrew Actuator style for the TL motion.

Pictures from the first actuator build with a little bit about each component.

3d printer settings
Material - eSUN 3d 1.75mm PETG black
Printer - Prusa MK2.5S with MMU2.0
Layer Height - 0.25mm
Perimeters - 8
Solid Layers - 12 top and bottom
All other settings are default from prusa for generic PETG filament




Overview of all of the parts used.



End cap for the linear bearing installed
1x Extrusion part # AFS-100100-10-250, Misumiusa.com
1x Linear Bearing block 3d Printed - PETG - 13:37 - 70 meters of filament
4x M10 x 40mm SHCS
4x 10mm washer




Linear bearing installed
1x LMK30uu Bearing
4x M6 x 25mm SHCS



Slider attached to the ball screw
1x Slider 3d printed - PETG - 15:17 - 73 meters of filament
1x SFU1605 L250mm ball screw
6x M5 x 30mm BHCS



Fixed end bearing attached to the ball screw
1x FK12 ball screw end support



Ball Screw end support attached to fixed bearing block
1x Fixed bearing block 3d printed - PETG - 15:27 - 75 meters of filament
4x M4 x 25mm SHCS



Linear shaft inserted into slider
1x 30mm Hollow Linear Shaft 270mm Long, 20mm ID



Fixed bearing block attached to extrusion
4x M10 x 40mm SHCS
4x 10mm washer





Shaft coupler installed on ball screw
1x shaft coupler 10mm x 12.7mm, D30 L42



Motor mounting plate attached to dc motor
1x Ampflow E30-400 dc motor
1x Motor mounting plate 3d printed - PETG - 3:54 - 20.5 meters of filament
4x 8-32 x 0.625 SHCS
4x #8 washers
Note: Motor shaft has been trimmed down to match coupler length



Motor Assembly attached to extrusion assembly
1x Spacer 3d printed - PETG - 7:04 - 35.4 meters of filament
4x 1/4-20 x 3.00 BHCS
4x 1/4 washers
Note: The coupler was aligned with the sight hole in the spacer and then snugged. The motor was then removed (coupler halves separated) and coupler was tightened properly before re-assembly.



Encoder installed to motor
1x encoder, E2-50-375-IE-D-G
2x 4-40 x 0.375 BHCS
1x encoder cable
Note: encoder was already installed from previous actuator build, it is best to install first.



End cap added
1x End cap 3d printed - PETG - 2:39 - 13.4 meters of filament
4x 8-32 x 1.0 SHCS



Finished Assembly!

So I was playing with my new actuator using a drill to run it in and out and this happened.




My first instinct was to blame the printer or the material, but if you look at the parts they did not break along the layer lines. What happened was I ran the slide to the maximum travel of the actuator at which point it makes contact with the linear bearing. The bearing is held in place with 4x M6x25mm screws, its hard to see but the print failed below the 4 mounting screws. ultimately I think this part will fail every time because no matter how beefed up I make it, it will still only be 4x M6 screws holding the force of the actuator.

Here is the fix


Now when the slide crashes into the new spacer it will put even pressure on the entire linear bearing block and on the other side are 4x M10 bolts threaded into the aluminum extrusion to help support the force. Initially I am going to print this spacer as a separate part because I already have most of my parts printed, but it could easily be printed as one piece.

Another option would be to flip the linear bearing around so the mounting screws are inside the actuator, that would give a similar effect and spread the force out even more. The bearing housing would need to be slightly modified with a grinder or dremel though so I am going to go with a spacer.




Now this should never happen once the kangaroo and sabertooth 2x60 are controlling the actuator so it may not actually be a problem, but I feel I should at-least throw a spacer in for the little effort it will take.

SFX100 actuator fitted with DC motor! Cool!

Just take care in case the motor run off... you don't want to hit the end-stops fast!

I was typing while you were posting this... I guess you already found out what may happen

Too late! 3d printed parts are no match for a Dewalt 20v drill at max speed and a 1605 ball screw. Thankfully the Kangaroo motion controller is much better at motion then I am.

has a dewalt drill got enough torque and rpm to run a ballscrew actuator ?

how well does it change direction ?

Yes a Dewalt drill runs it great, too well obviously as I smashed mine pretty good.
The drill was just for testing though, I wont be using it for my actual rig.

Hey @Hoddem excellent write uo and assembly Process Pictures , i just converted my ^dof Liner actuator to AASD15A servo motor drive and Now Have Both SFX Variants ... i too looked at going the SFX route but with minor tweaks to the design ... Controller all here now and just waiting for few mare actuator parts as i switched to 100 x 100 Kinetik Extrusion .. Great work wil post the Pics On my Wannabes 4DOF actuator post..

@wannabeaflyer2 that is awesome man I will be sure to follow your thread. I haven't been on xsimulator much the last year or so, lots of catching up to do.

Small update, I finished assembling all 4 actuators. not a lot of pictures for this update.

I didn't buy the encoder cables from US digital because they are expensive and I'm not sure how long I will need them overall. I ended up picking up a 50ft shielded cat 6 Ethernet cable off of amazon. I split that into 4 equal length cables and then terminated each end to make the encoder cables. The Ethernet cable comes with a drain wire so I connected that to ground on one end and left it floating on the other end. I have the ground and drain end connected to the kangaroo motor controller.

Once I had my encoder cables made up I started setting up the sabertooth/kangaroo combo. I wanted a way to test the system without hooking up power supplies so I connected 3x dewalt 20v 5AH batteries in parallel as the power supply for the sabertooth 2x60s. Then I assembled all 4 actuators into a square and added my racing chair.



After setting up the kangaroo/sabertooth combo in describe and then running through the simtools game engine setup I was able to control all 4 actuators. I have a lot of tweaking to do, but my kids absolutely love this thing. I just play with the sliders and try to knock them out of the chair. I was able to get about 15 minutes of testing before my batteries were dead.




I wasn't able to get a good "tune" on the kangaroo's and the motion seems pretty jumpy. I'm hoping its a combination of using under-powered batteries and not getting the tune correct, but I wont know until I get the setup moved to the garage for more testing. (I now have 4 massive dents in the carpet, Wife Not Impressed).

I also started mocking up the new rig, again I left some of the rails long so I can move and test things before I commit to trimming it down.



This brings me to my next challenge, and from reading the forums there seems to be some strong opinions on this topic.

4 actuators vs 3 actuators
As you can see I have already started mocking up the rig for 3 actuators, but I do have 4 built and it seems a shame to let one go to waste. (future traction control maybe)

4 Actuators in the corners




4 actuators spaced evenly around the seat



3 actuators, 1 in front


3 actuators 1 in rear, I can move the front actuators further forward to give more stability as needed.


I can test all 4 configurations easily enough, but In my mind I want the center of rotation to be as close to centered on my seat as possible.

I also had a few spare minutes to work on my power supplies, I got the final wiring done on two of them.



I ended up pulling the connector of the power supply, soldering the pin headers so that it powers up as soon as it is plugged in and then soldering wires in. The supply labeled floating has all of the case grounding removed so that it can be connected in series with the other supply to get 24v output.



A couple of observations from my limited testing.
1.5" x 3" extrusion is not stiff enough, I am getting a ton of flex. I may look into 1.5"x4.5". I know the simlab chassis is using 40x160 roughly 1.5 x 6". I have enough extrusion that I could make my own 1.5x6 bars.

sim tools does not play nicely with the kangaroo and homing on startup. During the homing sequence the kangaroo ignores all new commands. during that time simtools sends positional commands and then once the kangaroo is homed, BAM the rig jumps to center at full speed. This would be avoided if the rig is already homed and ready when simtools is started. Another option would be to put a pause in the output to allow the kangaroo to finish homing.

@yobuddy is there anyway to invoke a pause/delay in sim tools when sending the startup commands? I imagine not, but that would be very helpful here. I see that if I set the startup to 5000ms it will send all the commands and then wait for 5 seconds before sending positional commands, but that still leads to a big jump to after 5 seconds.

ideally I could do something like this
1, start
1, home
simtools delay 5 seconds
1, p600 s200 - move to center at a slow speed

Regarding your last point about using the startup commands to home and delay.
I have a similar thing where I need to go through a homing routine but I found it not suitable to do this via the startup option in Simtools.
I found it was getting tedious going through the homing sequence on every startup.
Especially in the early stages while fiddling around with tuning and fault finding other issues, you start and stop many times.

I used Foxterm as a separate and standalone app where the commands and serial connection parameters are defined in separate macro files and xml files.
Then you setup one or more foxterm.exe shortcuts with parameters to run those startup macros.
I then pinned the shortcuts to the windows start panel (another option is the quicklinks in Simtools)
With it that way, its a quick task to hit windows key, tap the applicable shortcut to go through homing and move to the middle position,
The foxterm app will launch, connect to the port, run the macro, then closes the ports and exits the app so nothing else to do or click to move on.
I only do this once for the evening and then following that its all Simtolls or Mover driving it therafter.

I have a few macros setup in my case due to a long ballscrew.
I have a homing routine that finishes in the bottom position, so I can climb in.
When I'm sitting in the rig, I tap the shortcut to raise to the middle position (becasue climbing out of the rig at middle position is a bit high and akward and Im gonna trip over one day.)
I also have a macro to goto highest position so I can oil the ballscrews.

I also did this with the powerdown. My rig backdrives on powerdown and it kept dropping everytime it disconnected from Simtools if using powerdown on shutdown.

With powerdown separate, then everytime you disconnect from Simtools or Mover, it holds the last position until you powerdown separately when your ready.

This is a good opportunity to raise a feature request with @yobuddy for the next version.
It could be an additional tool button where its a one shot startup routine that runs a single command line using the current interface parameters.

@JAD, that is a pretty good solution. I have to admit I was thinking about something similar, but it would have been a really basic script. This seems like it will be way cleaner and your right, honing every time I start simtools may not be ideal.
I will bookmark FoxTerm and play around with it when I get time.

thanks for the info.