Peacemakers 6dof Unicorn Build

Updated Design - July 2022
---------------------------------
** ACTUATOR V3 FILES**

https://www.xsimulator.net/communit...-6dof-unicorn-build.12523/page-13#post-229856

Here are the files for the V3 of my actuators. Motor mount and bearing mount are the same as original
To suit 80ST motors

My measurements are as follows:

6 x 580mm PVC 80mm storm water pipe (82mm OD x 76mm ID)
12 x 8mm x 580mm rods or tubes tapped at each end with m5 thread
18 x 12mm x 550mm linear rods or tubes with m5 tapped at each end
18 x RJMP-01-12 (LMUU12 Igus linear bearings - china clones from aliexpress)
6 x 600mm 1610 ballscrews
6 x 608 bearings


Giving you an actuator stroke of ~500mm at a speed of 500mm/s

Enjoy

Build videos see here:






Updated Design - July 2020 (older)
---------------------------------

- Thanos AMC-AASD15A controller - Tindie
- 80ST-M02430 with AASD15A servo driver - Aliexpress
- 83mm OD, 78mm ID PVC DWV stormwater pipe - Bunnings
- 8mm OD, 4mm ID Aluminium tube - Ebay
- M6 Threaded rod - Bunnings
- 3/8 Stainless Intrument tubing - free from work
- WD40 Dry PTFE spray - Bunnings
- WD40 Lithium grease - Bunnings
- 1605 500mm Ballscrew
- FK12 Bearing - Aliexpress
- Carbon rod 25mm OD, 21mm ID - Aliexpress
- Tie rod end (Rose joints) etc
- 3d printed parts



Concept:
Basically moving away from Limit switches and potentiometers and having the ease of plug and play is how this came about. The top hat is made of 2 x 3d printed end plates sandwiching against the PVC pipe. The top end plate houses the linear bearing and the base end plate houses the FK12 ballscrew bearing. The 2 x aluminium rods act as 2 things, the clamping force and the linear guides to stop the ballscrew twisting. They are tapped with an M5 thread and are a few mm shorter than the pvc pipe to ensure they pull up tight. Both end plates have locating holes for easier alignment.

The Slider is pretty straight forward.. guides the ball screw up and down the 2 aluminium tubes to avoid it spinning. Oring placed underneath like the SFX100 design. Softer calibration and avoid binding. Probably not that necessary but i had ordered the oring kit anyway so why not lol. Sprayed the rods with Dry PTFE spray also.

Mounting the motor isnt as simple as using threaded rod and nuts as there isnt enough room against the motor to get a spanner. 250-300mm socket head bolts are expensive so i ended up using some solid instrument tubing and turned them to 160mm each on the lathe. The bottom end plates on the top hat have counter sunk nuts on one end and the rod goes through the motor, motor spacer and screws in through the base plate. Then the instro tube goes on and a bolt sandwiches all these together. Then the unijoint plate goes over top and the whole thing is solid as a rock. The coupler is then tightened.

I used the 80ST motors for a couple reasons.. price being one really big one.. theres a big difference actually, almost $100AUD a motor! (I think the SFX100 users have driven the 90ST prices up to be honest).. But also because my old actuators used a 10mmx17mm motor coupler so i could lathe out 2mm to the required 10x19mm size and not buy 6 more couplers.

Another big bonus of this redesign is a slimmer, neater, better looking design which is dust resistant and probably a tad quieter as well.

So Far everything seems to be going smoothly (famous last words lol). Stay tuned

Update 1/4/20 - Here are the 3D Printed files in case anyone wants to use or alter the design to suit different size bits and bobs. Included are Inventor Files, STL and STEP files. Just don't go trying to make money off of them - https://drive.google.com/open?id=1cYwBu5xCy5mEMSg9mgW_1AeAleIZNCaj

Update 13/5/20 - Heres a build video and general walkthrough of my current actuator design i'm using:


Original Post
---------------------------------
Hi all, I'm Chris, i signed up over a year ago and got inspired to do a 6dof setup of my own, i ordered a crap load of parts based off silentchill/wannabeflyer/seattime build and then went and did a silly thing and got myself a girlfriend... haha. So not much time for my projects but i finally knuckled down this last week and made a start on my actuator.

My build is as follows:
Based on designs from SeatTime, Silent Chill, wannabeflyer

Motor: MY9764 450w 4500rpm 24v
Drivers: Sabertooth 2x32
Ballscrew: 1605 L500 with bk/bf12 bearings
Controller: Thanos AMC1280USB
Potentiometer: Vishay 50k 10 turn (not sure if this ok.. hopefully)
Linear Bearing: Aliexpress Igus bearing special
Power Supply: 600w LED 24v PSU
Joints: Rose Joint on top, Universal Joint on base
Shaft: Carbon fibre - 25 x 21 x 500
Adhesive used: 3M Scotch-weld NP 460 NS
Coupler: D44 L50
Guide: PTFE

Here's my progress so far. Had to stop because i couldn't find my belt and pulleys i ordered over a year ago.. might turn the house completely upside down tonight.. i was hoping to do some load testing tonight.
I designed a horseshoe style locator that mounts on the underside of the ball screw, one side is a guide that runs between 2 pieces of Teflon to minimise backlash on the shaft, universal joints and carbon rod coupler, it puts all the strain on the ballscrew housing instead. The other side will drive the belt/pully system for the potentiometer and limit switches.

The igus bearing was sourced from Alibaba for about $240 Aud for 6. Igus Australia would not return my emails for a quote so screw them! haha. These bearings came and they were actually double length with 2 sleeves in them so i cut them in half to be more like the ball bearing linear bearing size. It also maximises travel distance.

The mounting plates are cut from 10mm steel. I now realise that was overkill.. but this is more to test out my designs and make sure everything lines up etc. might look into 10mm Aluminium or 6-8mm steel instead when i cut out the rest.

The sole reason for going 24v is to basically avoid using a battery in my setup. I see now the motor is quite a bit bigger than the 12v variants but ive managed to keep the top hat to same size as teh 12v builds.

I'll stop rambling on now.. If you have any hints or ideas please shoot them this way, i'm all ears

P.S. please ignore the teflon plate, it's just a proof of concept, it was cut with a jigsaw by hand. im waiting on a bigger sheet to arrive, it will be cut nicely with a band saw next time lol.

It's Alive!!!
Ok so here is the first test with simtools. I had to get the misses to drive so i could film and stand really close to the E-Stop, haha. It's her first time on any driving game but the point is more to see how the actuator reacts.
Few things first. While testing using axis(a) it used the full range of the actuator which is far more than i need as im only using half of my 10 turn pot. when testing using 6dof output it seems to control to the limits set in simUtil and operates a good 10% either side of the limit switches. When i fired up Project Cars even though the actuator is centered it drives to 100% and wants to treat that as the new center, Dirt did the same thing but went to the bottom limit, say 0%. I went and changed axis limit to 50% and then it wouldn't do that anymore but it still close to the limits, so i changed to 35% but didn't notice a great deal of change from 50%.. anyway i was confident, somewhat, so i crank up the pwmmax to 80 and nervously give that a go. A few times it seemed to get close to the limits but all was well for now. Fpid is at 5 but dont see much difference between 1 and 5, my motor still twitches a bit at idle regardless.. maybe because im running 50k ohm pots.. not sure

I am still waiting for my metal pulleys and gears and am using 3d printed gear and pulley so dont want to go cranking up the speed too much for now in case it slips and all hell breaks loose. Don't really know what settings to use for my axis so found Vulbas did a post with his 6dof settings and copied his Axis 1 settings for now.
As for load tests, i am waiting for my power resisters for the 2x32 so don't want to go loading it up too much without them. Once i get my ball screws in the mail i should be good to build the other 5 actuators. Might just yolo it and build the thing and cross my fingers it holds up.

I am currently modelling a anti backlash carrier like silentchills to run v-wheels instead of the teflon slide i got currenty and to include belt clamp and limit switch ramp. Also want to make a spiral spacer for the ball screw rod and carbon tube to help keep it centered and minimise flex .All in all i'm happy with the initial tests.

I was working with AMC1280 to begin with but couldn't get motor signal out so swapped over to SMC3 on the Uno, so far so good! Thanos has informed me that because i was trying to operate actuator from 0-45% with a cneter set at 22% it wouldn't work with uni-direction sabertooth. But now im running the SMC3 with the pot at 50% i might try the 1280 again soon with 50% center and see if i get any action. but for now SMC3 is just working for me.

Quick little update..
So i modelled up an anti backlash block with limit switch contact and belt clamp (based on @SilentChill s setup). Did a dummy fit and all looks great! still waiting on my metal belt clamps so modelled one up to use for testing purposes, it'd probably go alright actually.. But yeah, the wheels i stole off my old 3d printer and are 2mm smaller than the ones i got coming but the gap looks to be spot on 1-2 mm so i'm confident my measurements are all good for my wheels i got coming! bye bye teflon slide! Also note i have limit switches without the roller and extended arms on their way too, in case you are flipping out..

My mates workshop is pretty busy so my metal plates aren't getting cut anytime soon.. i have PTFE filament due here tomorrow so might just start printing them and see how strong they come out and start assembling the rest of the actuators. Pick-up steel in a couple days so can start fabricating the base and frame this weekend hopefully. Room has been cleared out ready for her! Good (and stressful i'm sure) times are coming.

Ok, another progress update, its been a week of breakdowns and spending money. First my welder wire feed broker and its obsolete and just cheaper to buy a new Mig, so i did, then my 3d printer started playing up and again its not much dearer to buy a new one.. so i did, at least i got a nice big 330x330x400 build platform now But... then it struggled to heat the big platform and reach 250+ on the hotend to print PETG at a decent speed, so i banged a 24v power supply in it as everything is rated.. everything except the heater cartridge of course. garghh. so now waiting a week for another one of them to arrive.. so actuators aren't really happening at the moment until i can print my bits. But i will say 24v kicks arse! got to 260 deg on the hotend in like 15-20 seconds, and the bed wasn't far off haha.

I managed to knock up the top and bottom frame so here is some pictures of something at least. It has 3 levels of adjustment, foot pedals, Gear shifter and seat on its rail so it should be able to accommodate most body types. I might 3d print the pedal and shifter mount eventually to save a tad of weight but we'll see.

Still need to figure out where im going to mount the wheel as i want it to be maybe interchangable/removable.. but will more or less just be a bar across.

Busy week for me! Finally got actuators together and did a test fit. Got the bulk of the wiring sorted ready for bench testing and then reassembling inside where it will belong. Funny how the closer you get to D-Day that you've been building up to for months the more nervous and scared you become to run it through it's paces haha. Just don't wanna go destroying shit (or body parts ). The frame was a bit wide at the front and close to the front actuators so made the necessary mods to help there. But all in all pretty happy with my actuator angles etc.

So a few questions if anyone is able.to help.
Is there an axis orientation people like to use.. Like Axis 1 = front right actuator then 2,3,4 etc go clockwise around? That That way it's easier to perhaps copy the 6dof settings and just dial them down for initial run and do my own tweaking from there?

Is it worth me getting hexapod? What exactly are the benefits?

Any general tips on testing/firing up for the first time would be great.. I was just going to set my motors to low PWMmax values and use the 6dof sliders and test the yaw,pitch etc sliders.. are there any good articles on setting up the axises with what axis's should be doing what when doing certain DOFs, if that makes sense. Maybe I'm just blind or searching the wrong shit haha. Any tips would be greatly appreciated

That's what I use.
I also searched the forums, and I think most are using the same. But I might be wrong.

I thought the same.
But you will never get a good combination while simtools has no hexapod calculation.
The ideal is to apply sway to the sway in your rig. For that to be exact you need to calculate the length of your actuators, not guess what each one will be affected.

Hexapod calculation, that's what it does.
I don't have it, and I'm not sure, but I think it's only compatible with TronicGR AMC1280USB controller.
I have my own application to do this, but it's made to work with the interface I use (two ESP32).
Maybe I could add your Arduino interface. What commands do you send to the Arduinos?
My code receives the current position from the interface to calculate speed. We want all actuators to reach destination at the same time.

Just don't seat in the chair

Hey Flypt, thanks for the reply!! Sorry for the late response as i've been bed ridden for the last 3 days. Bloody gastro! garghh! Now got a few days off to try and get something happening with this rig.

I have been following your thread, it's such a good read man, Amazing work. You sir are very talented!

Oh so you've created a program to input your rig dimensions and then test your rigs sway etc? does it then give you what figures to put into simtools? or does it run as an overlay/plugin to simtools and when a game asks for sway it then performs the calculations and sends them to the rig? I'm sorry i am a bit confused haha.
I'm not good at coding at all so have no idea what commands are sent to the arduino. How do i find out? I'm using SMC3-SPS Sabertooth packet serial PID sketch if that helps (https://www.xsimulator.net/communit...ket-serial-pid-motor-driver-w-softstart.9277/). I would be keen to be your arduino test dummy if you'd like.. haha.

Currently i'm just trying to tune my motors as they are twitching a fair bit from the floating of the pots. I'm using 50k pots, not sure if that matters.. everyone says 10k pots, would 50k make much difference?

Or... i could dust off the AMC1280 and give that a crack with what limited information is available.. i see SeatTime got his working nicely with hexpod.. thats my other option i guess.

As for not sitting in the chair, oh hell no! but in saying that my usb cables aren't that long and i sit there on the carpet inputting commands and i'm scared it might wigg out and then sway or roll into my head and knock me out haha. Gotta treat these things with respect, for sure!

Well here we are, full sim setup in my room trying to tune the mofo and its just being a right pain in the butt!! but lets take a moment to appreciate how cool it looks while setup in the room it will be living the rest of its life.. haha!
I'm trying things like heave and it's like 2 actuators out of the 6 just lag behind the rest.. causing the whole rig to shake itself to shit. not sure whats going on to be honest.. maybe there is a mechanical restriction on those actuators that cause them to be like half a second behind the rest.... the annoying thing is that they aren't actually on the same driver either, they are 1 output on 2 different sabertooths.. garghh! I'm runnin out of ideas fast! anyway here is how the setup looks currently. I am a little drunk right now so i am going to let you all in on a few little things i've learnt so far this build haha. I didn't realise how far guys have come with 3d printed couplings to the ball screws so my method is so primitive in the sense that i was naieve enough to think that if/ (when) shit goes wrong i'll just buy a new rod and ballscrew.. but now i think it's way beneficial to have a serviceable ballscrew.. when i was test driving one of my actuators i accidentally drove it too high and it hit the spacer at the end of its limits and popped the seal and a few ball bearings fell out.. now i have to replace the whole ball screw and rod and what not.. i'm a little worried that i've commited to this sealed unit and now any failure will result in a $100 per actuator fix. oh well. i was still working off 1 year old tech haha. Now my inventor skills have increased 10 fold just in the last couple months i would have done shit so much differently! theres so many different threads and post to constantly follow. it becomes overwhelming at times.. every post you revisit you rethink your whole approach over and over again and all that goes through your head and all the questions raised about the little things like "these ballscrews are only rated for this and that" and are you sure thats going to take the weight" etc but the funny thing is that @SilentChill is a perfect example of that being null and void! countless hours of racing videos and his rig just takes an absolute beating! i feel like we are trying over complicate shit, looking for ways to improve a formula that doesn't really need that much improving! in the process we just create more problems haha. but Dont get me wrong i love where this forum is going and the stella upgrades people are coming up with but sometimes the old, if it aint broke, dont fix it.. rule should apply haha. Anyway i have sunk countless hours into my rig lately and my misses is ready to feed me to to the sharks haha.

I have managed to salvage an old clutch pedal for an analogue handbrake! i have a thingiverse file for the 3d printed part i created if you're interested, give it a search. I just wanna rip that bad boy while playing dirt on my rig already!! haha.

You'll have to excuse my drunken rant last night, i was ready to torch the thing! haha

Yeah thats what i was thinking.. hence why i thought it might be more of a mechanical issue. I will grab some screen shots tomorrow and get a video of it and upload it when i get a chance.

It looks awesome matey well done !!

WOW how did I miss this post , Great work and Yet another actuator variant hits the ground running LOL , loved the Controller layout and the attention to detail . One thing LOL and im always harping on about this But you really need to make sure the Limit switches are place so that they stay enabled when trigger By the actuator belt carries ....I have Again recently seen my actuators travel past the switch and retriggering .. Do that a few times up against a hard stop and soon the ballscrew nuts/carrier will be kaput,, these units are very very quick and get dangerous in the blink of an eye .. My switches have long lever arms to ensure the switch stays off when activated but even then sometime gravity will drive the Actuator past the switch ..... (BANG)

now thanks to some advice found here on this forum, I have fitted latched relay to control limit switch operation ... so far they do the job great,if the actuator over travels due to speed /Gravity bat least it can longer be accidentally re-energised so its a slow movement to mechanical stop rather than full speed scary bang stop

As always guys not a critique just some lessons learnt being passed on ....if you already knew this Then as they used to say in mission impossible " this message will self distruct in 5 minutes "

wow.. your build looks grreat!!

would it be asking too much for the dimensions of your top frame and thickness of the steel tubes that you used? @Peacemaker105

still simming?

Hi everyone, sorry i have been MIA for best part of 12 months.. life got in the way and all that. But coming back and checking on the scene i see there has been great strides made! Holy cow, i cant believe it. Well done to all involved!! In a way i am a little relieved as it will make things a whole lot easier going forward now.

I had the rig up and moving but had reliability issues with potentiometer mounts etc braking and causing me to just plain be too scared to trust it haha. The past 6 weeks or so i have been designing a new version of actuator for my rig using @Thanos new controller and 80ST-M02430 motors. This seems a no brainer to me. Gets rid of all the bulk like pots and limit switches making for a cleaner slimmer design and something more full proof and reliable. One downside is probably the 400mm max travel using a 1605 ballscrew by going this route but honestly my old design only gave me that anyway after losing some stroke to limit switches and pot holder etc (poor design by me really haha). is wont be much different.. and perhaps there will be a way to run more travel in a firmware update, who knows?
Also this system will be PLUG AND PLAY!! how good is that.. all that wiring gone.. just 6 Servo drivers plugged into the AMC-AASD15A with a db25 plug and plugged into each motor with the pre wired plugs on the motors.. amazing! Hopefully the 250mm/s travel speed is fast enough..

Anyway, here is the First concept actuator built and ready for bench testing in the next few days when i get time. Currently printed in PLA for R&D but will be printed in PETG once i'm happy. Cylider is 83mm Storm water pipe (not that fragile PVC stuff, this stuff is strong as hell) Internal Slider will maybe printed in Nylon or CNC'd from teflon.. but might even give PETG a go to be honest.. they use it in the SFX100 with next to no wear at all.

Hopefully post a video of it working in the coming days.
Happy Australia Day!

Hey mate, sorry for the late reply.. found out all my notifications go to junk... i'll try be more active now though. The "top hat" as i like to call it is 60mm x 100mm. the aluminium extrusion is 20mmx60mm which will be 2060 aluminium extrusion, if you're searching for it. hope that helps haha

Funny, i didn't even see this post mate! but you were right.. over shooting was a thing and also just sheering them off altogether was another.. hence the redesign haha. fingers crossed things go a little smoother this time round

Hiya @Peacemaker105 so far i have had only good results and impressive performance using the New AC servo dives and my system has taken a pounding ( read that as CRAP driver during my attempts to learn to drive fast and stay on the track , i have really stressed these Servos and actuator and cant say i have noticed them missing a beat so far ... You'll be pleasantly surprised at how easy the Controller wiring was (IMPORTANT NOTE it is mains so some knowledge of safe practise when dealing with AC) but for the most part i found it straightforward and easy .

Those servos and controller added to your actuator will perform great . you wont regret the initial Bill LOL once you being chucked around in the sim ( Limit switches ----What limit switches ) OK my 4DOF sim actuators are capable of 150 mm linear movement, but not had any system crashes other than the Idiot Driver not lost calibration or position that i can say i noticed and not tried to kill the operator either So all's good

Its still early Days for the 4DOF sim but gradully getting friends of to stress test the platform and will see how it goes further down the road as it were sorry Hog your Post but felt i needed to pass my enthusiasm on

I believe the AMC-AASD15A can currently do up to 2500mm stroke, so you shouldn't have any trouble with a max travel above 400mm.

Great work.

The max stroke for 1605 leadscrew is 640mm on the AMC-AASD15A. Look at the readme page on github.

https://github.com/tronicgr/AMC-AASD15A-Firmware


Thanks
Thanos