Christopher, Thanks again. Watching this again. I built my 3DOF platform and posted a few videos on it but I am really chomping at the bit for something like you built. Great Job! I just need to take some time from "having a great time driving my SIM ' to understand the AC component as I am using DC. Either way, we are both having too much fun! You, more than me.. DMAX
Thanks!, it is still entertaining me daily, I feel that using the AC Motors are actually easier than it is with DC, as the AC motors handle the PID algorithm controlling for you, all my circuit has to do is ask the AC controller where it should try to go. Either way, the added DOFs are absolutely worth while for more immersion. if you are looking to get into it, fist step is to purchase 1 AC motor and play with it, get it moving, then build up from there.
@@Christopherknauf So here I am again. Watching.. I lightened up my 3 dof rig by 52 pounds and holy crap! Great move.. Now thinking of CF frame for the 6 DOF rig. I guess I gotta just jump in and start making it.. I really like the range of motion your rig has.. I gotta get that .. Thanks again for posting this.. DMAX motion sim
What would you think about mounting this on a super sturdy 360° Platform? The Platform would provide general yaw for corners and such and the 6dof rig would provide the finer details. I am thinking about building one for myself and this is what I would add. To solve the rotational interface issue you could mount the pc etc. on the moving platform. Driven by a hefty 2kw motor of course.
Are you referring to the Current(A) consumption? if so there are 6 - 750w Ac Servos. I have a ammeter but have yet to monitor the actual current consumption. if you are talking about speed, I targeted 1200mm per second with the selection of motor/gear but have yet to actually test validate, mainly due to having to detune it so much from the motors capability, due to the occupants safety.
@@Christopherknauf thanks for the targeted speed, helpful too. I was asking about Wattage (or amp if you tell me if you are on 110v or 230v). Yes, the servos are 750w but doesnt necesarily mean that your platform consumption will be 4,5kW. That would be insane. If you have an Amperemeter (or multimeter clamp) you can measure consumption by multiplying Amp X Voltage (= watts). I really would apreciate.
@@Christopherknauf super, ty. I'm in Spain and its very important to know how much power will I need at home. Yes I'm planning to build something similar. Thanks for your answers
zero/imperceivable backlash is key,... when the motors quickly reverse direction,. you do not want to feel a clunk..... these motors/gear i specd out have very little,.. and is smooth as butter. the motors/gears had been specd out to be able to lift me + a chassis ~1 foot, against gravity in less than a half a second. the swing arm length is key as well, the longer the swing arms,.. the greater movements it an produce,.. such as all the angular degrees and other sway/surge/heaves power and backlash would be the first I look at, then swing arm length.
@@itzikbh301 I wear many engineering hats for Goodyear, mainly in racing designing RFID and Tpms systems, but starting motion simulation w/digital tire models.. this project was my resume for that
ive not touched my sim, it sits as a trophy to my life. life is good, never stopped building, and i try to respond to anyone who still contacts me about the project for interviews and advise on their own builds. Ive got big plans though, i just bought a racecar (factory five roadster), I will likely start some videos documenting the process... of... me ... making bad financial discissions......... i expect to get it in the next few weeks.. so stay tuned my dude!
Seems you got it dialed in pretty good. Did you reduced the collision forces to reduce injury to yourself? I also don't see much side movement for cornering. Unit must take a toll on the body if dialed too high I guess?
I have a filter on most of the inputs, this reduces the high frequency movements on the axis. Therefore protecting my neck most of the time. And I did leave out the side to side and forward back axis on the tune.... But only because I forgot to...and yes... It absolutely takes a toll on the body... I've had to stay of for weeks at a time due to it repeatedly straining the body.
@@Christopherknauf I can totally see a helmet and hans being useful here!! This is nuts, I keep dreaming, might be able to do a 2 or 3dof rig finally at least
@@henryjk9419 i have not ridden in over a year,... destroyed my body.... tons of fun though.. no regret. good luck on your build, most fun i have had on a project,. ask if ya have any questions!
@@Christopherknauf that's insane!! I'm getting back into sim racing while building my rally car for some training, I'm slightly worried a sim will end the cars if I can go anywhere while at home, saw a pretty cool 3dof +traction loss I might have to take some inspiration from, looks like might fit reasonably inside without being bolted down, perfect practice for a roll cage!
First off, Great Job! I know it's a lotta work. What kind of controllers are you using? I have 50:1 DC motors and just ordered some JRK's to upgrade from my ITB2's, (DC batt powered).. *simtools etc. Not the same as yours obviously if you are using AC motors, but What are you using to control the motors.
I designed my own controller PCB using a ESP32 as the main microcontroller doing the logics, as an engineering challenge to myself, i spent the last 2 years understanding and going through iterations of the design until it was reliable. I use AC servo motors w/50:1 planetary gears, the project is open sourced there should be a link in the description. if you wanted to look and tinker.
Thanks Christoper, I was looking on the xsim forum for about 6 months before giving it a shot. After 3 months *weekends * I'm up and running even though i have not posted my project (actually thinking about making a youtube vid series on it ) either way, I have exp designing circuit boards etc so i will take a look at the ESP32 bit chip. I (and many others most likely) appreciate you posting .. It looks GREAT!! D
Christopher, Thanks. I'm looking over the PCB GBR files - Not sure who you had make them (anyone in the USA? you could recommend?) .. Yeah my Dad (Space engineer with the MARS rover) always told me, whatever you think will work, probably wont, but i always figure it will for some reason and not been wrong yet.. Anyway, good talking to you. Thanks for sending the files link. D
Wow, that's an intense and insane build 🤝 It's been 2-3 years since you built this motion simulator, right? So, how many times did those motors and their drivers fail and get replaced?
thanks!, original motors, replaced 1 driver shortly after using a mig welder on one of the arms while it was all plugged in, may or may not have been related,.... my body(back and bracing leg) though failed from the forces and have been not using it while I recover....hence no video updates with it...
@@akshayjose6555 i am doing better than i was, though still leery of getting in the seat again,.... the PCB/algorithm is for only a 6DOF configuration. though the interface is the same with even larger motors.
those are the Gerber files, you send the entire zip file to the PCB manufacturer and they make the PCB to the spec detailed in the files. you can use a "Gerber file viewer" to look at it in 3d,.. there are several free on the interweb
WOW! Man this thing is completely off the chain! A job well done. How did you finalize the balancing? I just bought the DOFReality P6. I am going to use it as a springboard into this amazing hobby. (FD) I own a Fab shop. I really like your set up!
Thanks !!, It has been a career changing project for myself, and I enjoyed every moment of the build.. Balance was done more or less by eye at first, I knew I was going to need to be leaning back in order to offset the weight of the front pedal area/monitor, as well I needed to make sure the front did not get in the way of the actuations. I fabbed up the chassis in my basement over a year time, all the while using it as a static simulator to make sure all the mounts and pieces where in the right place and that I was comfortable while driving. ... if something did not look right.. chop it off and reweld it back at a slightly different angle .. it was a fun time!
I have not, I would estimate it ~300 lbs all in. the base disk is 100lbs alone ... super fun carrying that down to the basement. another 100 in motors and gears. the space frame, is thin wall tubing @ 1lb /foot. and I have ~25 feet of tubing all in. another 25lbs for a seat, 10 for monitor,10for wheel,... and you start running out of things at this point,.. speakers?... super light,...pedals??... ~3 lbs? . so yeah,.. ~300lbs. => ~136kg
Hi Christopher, first - huge congratulations for this build :-) I am truly impressed :-) I am going to build a rig and I am (was) decided to linear actuators based on the same motors and drivers. But... In your rig the responsiveness and light-looking movement are very convincing ;-) So maybe that would be better way... Can you tell me what are your parametres for pitch/roll against all other in precentage in sim tools? In almost all builds with linear actuators I see that there is so little pitch/roll movement when accelerating and braking and on long turns, so almost zero G-force simulation in these scenarios. In your build it looks as it should, I think. I am not sure if it is matter of settings or geometry and actuators. Could you, please, tell something about that?
hey dude, thanks for the words of encouragement. I long debated the linear as well. here are my current dirt 2 simtools settings - imgur.com/a/eaKeuDU I mix forces and use pitch and roll w/gravity as an added constant acceleration to reproduce surge and sway.I think is what you mean. the platform can do ~22 degree in each rotation axis direction from home. and has a full heave travel of 11.25 inches, in .6 seconds.. it may be able to go faster,... but this is limited to not destroy the chassis,... at that speed its also uncomfortable for the rider,.. I did the test with a heavy punching bag.... the fact that I have the monitor on the same plane as the rider, means I can swing them around more as well... as the perspective is fixed to the chassis.... keep up the questions,. There are many active builds goin on, I try to help where I can. but there is enough on my giyhub project site to get a DIY'r their own, I took 2 years,.. but one guy took 3 months to get his done.
By no means is this a step by step, but i have open sourced the project, and I hope to fill this with enough of a guide to get someone close. github.com/knaufinator/6DOF-Rotary-Stewart-Motion-Simulator the heave full movement is ~14 inches, and can be seen here during a movement test ua-cam.com/video/65dXTsMQKn8/v-deo.html I dont generally make it do the full movement with me in it,. as.. it can be a bit much and hurt depending on what axis was extended and how fast... the goal is to hurt .. just a little..
@@GamerPro-nw5lf i dont believe there is,...not saying its not possible, but I have yet to see the combo together, I have a ps +psvr, so I had thought of it,... the game would need to send the stream of the telemetry over UDP packet to a PC in order to decode the data to be converted into something the simulator understands... unless you create a converter directly on the sim driving circuit,.. (certainly doable)... I believe the "Dirt" series racing games have this support,.. where they send UDP packets... not sure if it supports VR though... I dont think it does?.. I did get VR working on this rig with a PC,.. it was a pain to get working where it did not hurt ur brain,.. but it did work out,. and was next level surreal experience ..
It does, telemetry of the vehicle such as acceleration vectors of the vehicle.. is sent out of the game to interpretation software.(simtools)..then sent to my microcontroller/circuitry that pushes the arms around at 1000hz
Great setup! Have you measured the actual latency for the movement (e.g. when you touch the curb how many ms later the platform moves)? Assuming this video is accurately syncronized to game video it seems like the latency were under 10 ms which seems much better than most commercial platforms can achieve. Once you get your rally driving skills as good as your building skills, you will be going really fast.
Thank you, I took pride in its design and construction. I am sending data to the rig @1ms interval. its up to my settings in the driver as well in the app on the PC to filter out some particularly large impulses.... There is no noticeable delay when driving, and I could even make it faster and more slammy,,. to where it would prolly kill you...I could make it swing around and disintegrate itself.. so I take my time to make sure ..Its tuned well,.... it was my intention to get it faster and stronger and cheaper,... than most/all commercial platforms can hit. though downside,...... my back has been tenderized,... hence I have several ways to stop/kill power for these reasons
I have my most detailed build log/info/picture album here -> www.xsimulator.net/community/threads/compact-ac-servo-6dof-rotary-stewart-platform.14769/ .. its not super detailed, but many people have built their own based on this design and gotten a working sim as a result,
Thanks Dave!, ... no issues with heat after several hours of abuse. I kept an eye on it early on with a IR heat gun, and the heat rise was inconsequential for both the motors as well the motor controllers.
@@Christopherknauf Is your whole rig suspended by the motors alone? It looks like you don't have any gas springs similar to DOF Reality P6 to help the motors.
The gears yes, they hold all the weight, at all time. I spent along time sourcing strong enough as well just as important,... Low Backlash, the gear is very precise, and is almost imperceivably the switch direction. many other systems you will feel a sudden slam... from the backlash of the gear,... these you cant tell, and it takes what the data i send it asks without that slam backlash.
maybe I buy the DOF reality H6, fabricate a tubular cage, extend the stock servo arms and make any other necessary adjustments to the geometry to achieve +/- 35 degrees pitch, roll, and yaw. and then lose some damn weight.... or suspend the cage from the ceiling with some springs, or a 300 lbs counter-weight on a 2:1 pulley system to increase the counter weights response time to 19.62m/s/s so I don't overload the little servos of the H6 with my weight. better yet - Perhaps one long travel dual rate coil over spring could be added below the cage right in the center.... I would have to get one for a very very light vehicle, or I suppose I could fabricate one to fit the right coil spring... or a gas spring made from a pneumatic ram plumbed into an off the shelf air compressor so you can set the preload using the regulator... and the volume of the tank would take up the travel without varying the force too much. Then, without over stressing the servos, I could achieve 30 cm of vertical travel which would theoretically allow me to achieve nearly 0.5s of zero g.. orthopedic spine surgeons feel free to chime in here or drop a business card.. hahaha. because that would require infinite acceleration. But I might be able to achieve 0.4s of nearly zero g.. however the software is not designed for 30cm travel... making each pebble on the on the road feel like driving over a brick... a crude solution to this problem might be a linearity controller for the gain of each channel. Does your setup include an accelerometer or is trimming completely dependent upon the geometry?? what sort of software changes would be required when changing the geometry so much. I know that the response gradient from 0 to 180 degrees would be a little out of whack with 20cm servos. with a 5cm servo it provides a convenient rounding of the effects as the servo approaches 0 or 180 degrees. I don't know the usual operating range of these servos in this application. probably not 180 degrees... (oh I see now that it's only about 20 degrees on these settings... so your setup can probably do 30 cm of travel! haha. I can fabricate, and I have some experience with arduino, but I don't have all the skills necessary for a software undertaking of this magnitude. yet. I'll have to see what software and controller is included with the DOF Reality H6. I figure let those guys do the heavy lifting for me, and I'll make some modifications.
Wow,.. that was long winded :) ... There is an accelerometer on my chassis but it is just for monitoring and I had a Kill cam that it triggered for awhile @ any event over 2G,... slow mo red tint of the carnage... but that hurt to much as I was egging it on.... and this platform needs respect or it will hurt you. My PCB/Controller can be easily modified for dimensional changes as long as the motors are kept within the realm of the normal 6dof rotary positions. there are variables that can be adjusted in the code that define the certain lengths and position of arms and rotation of each motor platform height so on so forth. there is an excel file that i based my algorithm on in the Github project under technical\ that allows a user to input there expected dimensions to see if it will work on the platform, its also got a animated visual representation of the rotary platform as you traverse through axis changes. The motor/gear combo this uses is far better than the H6 platform, the gear has VERY little backlash, faster and more powerful movements. my goal was to keep the monitor attached as that will help with immersion and opens you to tune the platform for more movement. you mention the 0-180 movement,.. and you are correct in that there is a non -linear response, the code handles where the arm should be at a given time and accounts for the non linearity by speeding up as needed when more movement is needed to achieve a linear feel to the rider. as long as there was enough speed overhead for the motor to turn. yes, my full heave is around 11.5 inches/ ~30cm,... its a interesting feeling during full speed heave,..it can move faster than gravity would have. i dont often do this test... but I have,.. for science. The software is easy... its done... it uses Arduino IDE ... but you point it to a ESP32 microcontroller (much faster processor) the PCB is also done, you just send the gerber files to a PCB shop (jlcpcb.com ~$25 USD for 5 boards) and a few easy to assemble components,... buy a motor and gear to test... minimal investment so far,.. spin it up using the software "SimTools" connected to your favorite car racing game, if the motor spins,.. you are 50% there. buy more motors,.. and fab a chassis. one gentleman built his platform based off of mine in 3 months. I have been helping several others in their builds, its a growing community and we are trying to improve things continuously. I try to make myself available for questions... by all means,.. question. :)
@@Christopherknauf Thanks for doing this project open!! I did buy a DOF reality P6. I tried it with longer arms, (had to add air springs, dampers,) the overrun on down functions was uncontrollable, and not enough power to lift even with help from some pneumatic cylinders for adjustable air spring - so I had to shorten them a little. yesterday I finally broke one off the frame at the mounting, (damaging that side of the geabox, now I have to mount it on end) now I think that the gearing in the Worm Reducers included in the DOF Reality P6 is too high for the included motors. They're way too fast, and not enough torque. I believe they are 1500RPM motors about 345Watts, and a 25:1 gearbox. so now, having broken a gearbox yesterday, I've been learning about worm reducers. I think I can replace the 6 gearboxes to double the torque, adjust some things in sim tools for the 1/2 speed, and fit longer arms for more travel. Might have to switch to planetary gearboxes when I push the limits on those and they break.
@@nashole23 there will be some tweaking in simtools for sure, watch out for backlash in the gear, my gears are something like 6arcmin?... and it wiggles about a mm or so @7 inches. no noticeable backlash from the riders point of view.. unlike my first gear box choice @ ~12arcmin backlash
If you die in the Matrix, you die in real life..... lol... maybe not... but crashing can hurt/injure you if the settings/gains are to strong.... I have not ridden in the sim for several months due to back/leg strain from this .... for instance... currently rehabilitating ....
Ive been asked that a few times, I am not in a position to make more, nor do I have intention of selling this unit as I put two years of my blood/sweat/tears into building it... Though I am giving away the code and design to the world!... And helping many people build their own! I recently started making semi instructional videos of the process.
@@bmarco1602 So, that's something I debated even,.. starting up a business selling these things,.. The design is not fool proof and can easily hurt the rider,.. there is allot of power on tap (~5 Miata's worth of torque), I have been injured several times from making mistakes. I am concerned with insurance/legal issues in dealing with safety/death/dismemberment... . One of the side effects I was looking for was when I open sourced this platform was to get other people involved so they could start making changes and evolving the platform (safety),.. there are a handful of folks building their own based on the design. But as far as me building a custom bespoke one to spec,.. I would have to be heavily persuaded,... its allot of time for me to invest. I am open to brainstorming and shooting ideas around, though I am only one man... and only a hobbyist with the welder :)
No,...though i did see it a few months ago for the first time, I wrote my own algorithm for handling the platform and baked it into the micro controller.
@@NischGTM I do use simtools, it's been real good to me. I use much of the suite, got a fan on dash output, transducer on vibe and I've gotten real good at mixing the outputs to my sim
I definitely have used it in FH4, and it is absolutely fun.. you need to dial the forces back quite a bit as its arcady .. .. and you spend most of your time at high speeds... and stopping instantly is not out of the question,.... and that .. hurts. though I dont think it supports VR....
It's amazing, for the moment you can try with oculus rift CV1, is quite cheap (150-200$) if you search it used, and if you put the sensor on the frame you'll have a self compensation of the movement
Agree, though I have since built a room for it that minimises the cluttered view... I am currently saving my dollars for a new project ..... I want to build a Shelby cobra .... I will accept donations of VR gear though in the mean time... :)
not for sale, but the project is open source, with a link in the description of the video. there are several people all over the world using the project to create their own interpretation of this simulator. i am actively working on improvements, including safely features, and Android app over Bluetooth for configuration.
11/10 bro, you nailed it. Force Dynamics wow'd me a decade ago but building this at home is an incredible achievement. Thanks for sharing!
You are a true inspiration for me as an engineering student. Thank you sir!
Christopher, Thanks again. Watching this again. I built my 3DOF platform and posted a few videos on it but I am really chomping at the bit for something like you built. Great Job! I just need to take some time from "having a great time driving my SIM ' to understand the AC component as I am using DC. Either way, we are both having too much fun! You, more than me.. DMAX
Thanks!, it is still entertaining me daily, I feel that using the AC Motors are actually easier than it is with DC, as the AC motors handle the PID algorithm controlling for you, all my circuit has to do is ask the AC controller where it should try to go. Either way, the added DOFs are absolutely worth while for more immersion. if you are looking to get into it, fist step is to purchase 1 AC motor and play with it, get it moving, then build up from there.
@@Christopherknauf So here I am again. Watching.. I lightened up my 3 dof rig by 52 pounds and holy crap! Great move.. Now thinking of CF frame for the 6 DOF rig. I guess I gotta just jump in and start making it.. I really like the range of motion your rig has.. I gotta get that .. Thanks again for posting this.. DMAX motion sim
This is the coolest 6 dof machine on the interwebz i think
I believe it
Geezus that looks awesome.
Feels it as well!
Man this is amazing! Great project!
Would love to see someone drifting in a set of Corsa with this set up
This is amazing, I want one!
What would you think about mounting this on a super sturdy 360° Platform? The Platform would provide general yaw for corners and such and the 6dof rig would provide the finer details. I am thinking about building one for myself and this is what I would add. To solve the rotational interface issue you could mount the pc etc. on the moving platform. Driven by a hefty 2kw motor of course.
U nailed it bro sooooo good. Love ur vid and ur serious face while playing hahahhaah
It's intense, ;)
YES
That thing is badass. Probably costs more than my house
It was only 5k usd ...ish .. it's all in the GitHub link in the description,.. there are many folks who have done their own with the design
@@Christopherknauf well I'll be damed. Ty friend.
This is impressive.
most impressive thing about this whole clip is that fact that he has a rest pedal when he's not on the clutch 😂
so cool - i want it 8)
If I had such a simulator, I wouldn't get married at all.
Working well
great! reaction time and reaction speed seems to be fast! ¿Do you know the consumption average/peak ? Thanks!
Are you referring to the Current(A) consumption? if so there are 6 - 750w Ac Servos. I have a ammeter but have yet to monitor the actual current consumption. if you are talking about speed, I targeted 1200mm per second with the selection of motor/gear but have yet to actually test validate, mainly due to having to detune it so much from the motors capability, due to the occupants safety.
@@Christopherknauf thanks for the targeted speed, helpful too. I was asking about Wattage (or amp if you tell me if you are on 110v or 230v).
Yes, the servos are 750w but doesnt necesarily mean that your platform consumption will be 4,5kW. That would be insane. If you have an Amperemeter (or multimeter clamp) you can measure consumption by multiplying Amp X Voltage (= watts).
I really would apreciate.
@@BallestaRacing Yeah, the servos are 230v, I have a clamp style ammeter, I will use it tonight when I get off of work.
@@Christopherknauf super, ty.
I'm in Spain and its very important to know how much power will I need at home. Yes I'm planning to build something similar. Thanks for your answers
@@BallestaRacing hit around 21A @ 220v ua-cam.com/video/HSEPRs3rGTw/v-deo.html
This looks absolutely INSANE!! #icanonlydream lol
Thanks!, and it can be even more insane!! its not at its full potential due to me not wanting to be hurt!!
Christopher Knauf keep it up man, can’t wait to see more 😁👍
I know the best is 6 DOF instead of 4/3/2… but what is the most aspect important when buying this item? Angular degrees? Engine power?
zero/imperceivable backlash is key,... when the motors quickly reverse direction,. you do not want to feel a clunk..... these motors/gear i specd out have very little,.. and is smooth as butter. the motors/gears had been specd out to be able to lift me + a chassis ~1 foot, against gravity in less than a half a second. the swing arm length is key as well, the longer the swing arms,.. the greater movements it an produce,.. such as all the angular degrees and other sway/surge/heaves power and backlash would be the first I look at, then swing arm length.
Super!
Now that's a nice toy my friend 👍👍
Wonder if need for speed works on that system??? Tuff😎😎😎
Amazing work man. NO WORDS.
Who designed your PCB?
I did, I am an engineer and built it up in my spare time.welded the chassis and designed and coded the PCB/microcontroller
Bravo!!! As an electrical engineer myself i salute you, really no words man, where you work?
@@Christopherknauf one man company.
@@itzikbh301 I wear many engineering hats for Goodyear, mainly in racing designing RFID and Tpms systems, but starting motion simulation w/digital tire models.. this project was my resume for that
@@Christopherknauf Good job man, Keep it going.
No more updates? I stopped by to see if there's anything new. Hope to see more content. This video shows some great camera views.
ive not touched my sim, it sits as a trophy to my life. life is good, never stopped building, and i try to respond to anyone who still contacts me about the project for interviews and advise on their own builds. Ive got big plans though, i just bought a racecar (factory five roadster), I will likely start some videos documenting the process... of... me ... making bad financial discissions......... i expect to get it in the next few weeks.. so stay tuned my dude!
@@Christopherknauf Looking forward to it!
Seems you got it dialed in pretty good. Did you reduced the collision forces to reduce injury to yourself? I also don't see much side movement for cornering. Unit must take a toll on the body if dialed too high I guess?
I have a filter on most of the inputs, this reduces the high frequency movements on the axis. Therefore protecting my neck most of the time. And I did leave out the side to side and forward back axis on the tune.... But only because I forgot to...and yes... It absolutely takes a toll on the body... I've had to stay of for weeks at a time due to it repeatedly straining the body.
@@Christopherknauf 4:50 looks like those impacts still pack a punch though. Such a cool set up
@@Christopherknauf I can totally see a helmet and hans being useful here!! This is nuts, I keep dreaming, might be able to do a 2 or 3dof rig finally at least
@@henryjk9419 i have not ridden in over a year,... destroyed my body.... tons of fun though.. no regret. good luck on your build, most fun i have had on a project,. ask if ya have any questions!
@@Christopherknauf that's insane!! I'm getting back into sim racing while building my rally car for some training, I'm slightly worried a sim will end the cars if I can go anywhere while at home, saw a pretty cool 3dof +traction loss I might have to take some inspiration from, looks like might fit reasonably inside without being bolted down, perfect practice for a roll cage!
First off, Great Job! I know it's a lotta work. What kind of controllers are you using? I have 50:1 DC motors and just ordered some JRK's to upgrade from my ITB2's, (DC batt powered).. *simtools etc. Not the same as yours obviously if you are using AC motors, but What are you using to control the motors.
I designed my own controller PCB using a ESP32 as the main microcontroller doing the logics, as an engineering challenge to myself, i spent the last 2 years understanding and going through iterations of the design until it was reliable. I use AC servo motors w/50:1 planetary gears, the project is open sourced there should be a link in the description. if you wanted to look and tinker.
Thanks Christoper, I was looking on the xsim forum for about 6 months before giving it a shot. After 3 months *weekends * I'm up and running even though i have not posted my project (actually thinking about making a youtube vid series on it ) either way, I have exp designing circuit boards etc so i will take a look at the ESP32 bit chip. I (and many others most likely) appreciate you posting .. It looks GREAT!! D
apparently the link was not in this vid - github.com/knaufinator/6DOF-Rotary-Stewart-Motion-Simulator
Christopher, Thanks. I'm looking over the PCB GBR files - Not sure who you had make them (anyone in the USA? you could recommend?) .. Yeah my Dad (Space engineer with the MARS rover) always told me, whatever you think will work, probably wont, but i always figure it will for some reason and not been wrong yet.. Anyway, good talking to you. Thanks for sending the files link. D
@@davecorr4421 I use jlcPCB, outa china.
Wow, that's an intense and insane build 🤝
It's been 2-3 years since you built this motion simulator, right? So, how many times did those motors and their drivers fail and get replaced?
thanks!, original motors, replaced 1 driver shortly after using a mig welder on one of the arms while it was all plugged in, may or may not have been related,.... my body(back and bracing leg) though failed from the forces and have been not using it while I recover....hence no video updates with it...
@@Christopherknauf Hope you're okay now. Can I follow this project to make a 3 DoF platform? (with 3 more powerful motors)
@@akshayjose6555 i am doing better than i was, though still leery of getting in the seat again,.... the PCB/algorithm is for only a 6DOF configuration. though the interface is the same with even larger motors.
what are the gko/gbl for - what do you use to open them?
those are the Gerber files, you send the entire zip file to the PCB manufacturer and they make the PCB to the spec detailed in the files. you can use a "Gerber file viewer" to look at it in 3d,.. there are several free on the interweb
WOW! Man this thing is completely off the chain! A job well done. How did you finalize the balancing? I just bought the DOFReality P6. I am going to use it as a springboard into this amazing hobby. (FD) I own a Fab shop. I really like your set up!
Thanks !!, It has been a career changing project for myself, and I enjoyed every moment of the build.. Balance was done more or less by eye at first, I knew I was going to need to be leaning back in order to offset the weight of the front pedal area/monitor, as well I needed to make sure the front did not get in the way of the actuations. I fabbed up the chassis in my basement over a year time, all the while using it as a static simulator to make sure all the mounts and pieces where in the right place and that I was comfortable while driving. ... if something did not look right.. chop it off and reweld it back at a slightly different angle .. it was a fun time!
@@Christopherknauf So inspiring!! I posted this video to my DOF group. You really knocked this out of the park. I wish you all the best!
@@Christopherknauf Have you measured how much your rig (the part suspended in air) weights without you in it?
I have not, I would estimate it ~300 lbs all in. the base disk is 100lbs alone ... super fun carrying that down to the basement. another 100 in motors and gears. the space frame, is thin wall tubing @ 1lb /foot. and I have ~25 feet of tubing all in. another 25lbs for a seat, 10 for monitor,10for wheel,... and you start running out of things at this point,.. speakers?... super light,...pedals??... ~3 lbs? . so yeah,.. ~300lbs. => ~136kg
@@Christopherknauf Wow! Those motors make it appear like a lightweight rig.
wow. do you sell these things?
Nope! But I do help people build their own, ... everything is open source!
Hi Christopher, first - huge congratulations for this build :-) I am truly impressed :-) I am going to build a rig and I am (was) decided to linear actuators based on the same motors and drivers. But... In your rig the responsiveness and light-looking movement are very convincing ;-) So maybe that would be better way... Can you tell me what are your parametres for pitch/roll against all other in precentage in sim tools? In almost all builds with linear actuators I see that there is so little pitch/roll movement when accelerating and braking and on long turns, so almost zero G-force simulation in these scenarios. In your build it looks as it should, I think. I am not sure if it is matter of settings or geometry and actuators. Could you, please, tell something about that?
hey dude, thanks for the words of encouragement. I long debated the linear as well. here are my current dirt 2 simtools settings - imgur.com/a/eaKeuDU
I mix forces and use pitch and roll w/gravity as an added constant acceleration to reproduce surge and sway.I think is what you mean. the platform can do ~22 degree in each rotation axis direction from home. and has a full heave travel of 11.25 inches, in .6 seconds.. it may be able to go faster,... but this is limited to not destroy the chassis,... at that speed its also uncomfortable for the rider,.. I did the test with a heavy punching bag.... the fact that I have the monitor on the same plane as the rider, means I can swing them around more as well... as the perspective is fixed to the chassis.... keep up the questions,. There are many active builds goin on, I try to help where I can. but there is enough on my giyhub project site to get a DIY'r their own, I took 2 years,.. but one guy took 3 months to get his done.
in this video my yaw/traction loss algorithm was messed up in my microcode,.. so it was limited to ~30% of its potential... I later fixed it...
Now that....but with VR
Donations accepted!!! ;)
Is that a max setting for heave movement? You dont have step by step for this project? Thanks
By no means is this a step by step, but i have open sourced the project, and I hope to fill this with enough of a guide to get someone close. github.com/knaufinator/6DOF-Rotary-Stewart-Motion-Simulator the heave full movement is ~14 inches, and can be seen here during a movement test ua-cam.com/video/65dXTsMQKn8/v-deo.html
I dont generally make it do the full movement with me in it,. as.. it can be a bit much and hurt depending on what axis was extended and how fast... the goal is to hurt .. just a little..
This is possible used in PS4? Gran turismo Sport, Driveclub? In psvr?
Pc only, ps games do not "generally" have the telemetry output required to feed the sim
@@Christopherknauf is there any 6dof to play racing games on ps4 with psvr?
@@GamerPro-nw5lf i dont believe there is,...not saying its not possible, but I have yet to see the combo together, I have a ps +psvr, so I had thought of it,... the game would need to send the stream of the telemetry over UDP packet to a PC in order to decode the data to be converted into something the simulator understands... unless you create a converter directly on the sim driving circuit,.. (certainly doable)... I believe the "Dirt" series racing games have this support,.. where they send UDP packets... not sure if it supports VR though... I dont think it does?.. I did get VR working on this rig with a PC,.. it was a pain to get working where it did not hurt ur brain,.. but it did work out,. and was next level surreal experience ..
I’m wondering does it actually track the motion that is seen in the game? If so how
It does, telemetry of the vehicle such as acceleration vectors of the vehicle.. is sent out of the game to interpretation software.(simtools)..then sent to my microcontroller/circuitry that pushes the arms around at 1000hz
Great setup! Have you measured the actual latency for the movement (e.g. when you touch the curb how many ms later the platform moves)? Assuming this video is accurately syncronized to game video it seems like the latency were under 10 ms which seems much better than most commercial platforms can achieve.
Once you get your rally driving skills as good as your building skills, you will be going really fast.
Thank you, I took pride in its design and construction. I am sending data to the rig @1ms interval. its up to my settings in the driver as well in the app on the PC to filter out some particularly large impulses.... There is no noticeable delay when driving, and I could even make it faster and more slammy,,. to where it would prolly kill you...I could make it swing around and disintegrate itself.. so I take my time to make sure ..Its tuned well,.... it was my intention to get it faster and stronger and cheaper,... than most/all commercial platforms can hit. though downside,...... my back has been tenderized,... hence I have several ways to stop/kill power for these reasons
Is 15 amp enough for servo motor 6 dof?
ua-cam.com/video/HSEPRs3rGTw/v-deo.html , Hit 21A @220v full tilt. 15 would certainly pop the breaker
I'm the kind of person who likes detailed instructions. Is there a section in your files / web site that has a detailed explanation of the build?
I have my most detailed build log/info/picture album here -> www.xsimulator.net/community/threads/compact-ac-servo-6dof-rotary-stewart-platform.14769/ .. its not super detailed, but many people have built their own based on this design and gotten a working sim as a result,
photos.app.goo.gl/6UpUdS5Q1otJAuhEA is my personal album, it has more photos than on the forum
Very nicely done sir! Do you have any issues with motor cooling?
Thanks Dave!, ... no issues with heat after several hours of abuse. I kept an eye on it early on with a IR heat gun, and the heat rise was inconsequential for both the motors as well the motor controllers.
@@Christopherknauf Cool thanks for the reply! I will try using or building servos, for my next rig rebuild
@@Christopherknauf Is your whole rig suspended by the motors alone? It looks like you don't have any gas springs similar to DOF Reality P6 to help the motors.
The gears yes, they hold all the weight, at all time. I spent along time sourcing strong enough as well just as important,... Low Backlash, the gear is very precise, and is almost imperceivably the switch direction. many other systems you will feel a sudden slam... from the backlash of the gear,... these you cant tell, and it takes what the data i send it asks without that slam backlash.
@@Christopherknauf Is that off-the-shelf gearing or have you tuned it somehow to reduce backslash?
how much does the motion rig cost? without wheel and pedals just the rig
~$4-5k USD and many man hours.
Looks like car is doing wheelies and nollies.
mental note,... need to build a skateboard platform for the sim..
Hello Christofer, do you sell this maskin?
No sir, but I have helped many people build their own based on the design..
Oki so how much for the staff..?
Do you live i germany. ?
@@JohnRambo.k ~5,000USD, I live in USA my dude
@@JohnRambo.k www.xsimulator.net/community/threads/compact-ac-servo-6dof-rotary-stewart-platform.14769/
How much does this whole setup Costs ?
~$5k USD, I spread it out over the course of 2 years, wife barely noticed me spending all our extra monies :)
maybe I buy the DOF reality H6, fabricate a tubular cage, extend the stock servo arms and make any other necessary adjustments to the geometry to achieve +/- 35 degrees pitch, roll, and yaw. and then lose some damn weight.... or suspend the cage from the ceiling with some springs, or a 300 lbs counter-weight on a 2:1 pulley system to increase the counter weights response time to 19.62m/s/s so I don't overload the little servos of the H6 with my weight. better yet - Perhaps one long travel dual rate coil over spring could be added below the cage right in the center.... I would have to get one for a very very light vehicle, or I suppose I could fabricate one to fit the right coil spring... or a gas spring made from a pneumatic ram plumbed into an off the shelf air compressor so you can set the preload using the regulator... and the volume of the tank would take up the travel without varying the force too much. Then, without over stressing the servos, I could achieve 30 cm of vertical travel which would theoretically allow me to achieve nearly 0.5s of zero g.. orthopedic spine surgeons feel free to chime in here or drop a business card.. hahaha. because that would require infinite acceleration. But I might be able to achieve 0.4s of nearly zero g.. however the software is not designed for 30cm travel... making each pebble on the on the road feel like driving over a brick... a crude solution to this problem might be a linearity controller for the gain of each channel.
Does your setup include an accelerometer or is trimming completely dependent upon the geometry??
what sort of software changes would be required when changing the geometry so much. I know that the response gradient from 0 to 180 degrees would be a little out of whack with 20cm servos. with a 5cm servo it provides a convenient rounding of the effects as the servo approaches 0 or 180 degrees. I don't know the usual operating range of these servos in this application. probably not 180 degrees... (oh I see now that it's only about 20 degrees on these settings... so your setup can probably do 30 cm of travel! haha.
I can fabricate, and I have some experience with arduino, but I don't have all the skills necessary for a software undertaking of this magnitude. yet. I'll have to see what software and controller is included with the DOF Reality H6. I figure let those guys do the heavy lifting for me, and I'll make some modifications.
Wow,.. that was long winded :) ... There is an accelerometer on my chassis but it is just for monitoring and I had a Kill cam that it triggered for awhile @ any event over 2G,... slow mo red tint of the carnage... but that hurt to much as I was egging it on.... and this platform needs respect or it will hurt you.
My PCB/Controller can be easily modified for dimensional changes as long as the motors are kept within the realm of the normal 6dof rotary positions. there are variables that can be adjusted in the code that define the certain lengths and position of arms and rotation of each motor platform height so on so forth. there is an excel file that i based my algorithm on in the Github project under technical\ that allows a user to input there expected dimensions to see if it will work on the platform, its also got a animated visual representation of the rotary platform as you traverse through axis changes.
The motor/gear combo this uses is far better than the H6 platform, the gear has VERY little backlash, faster and more powerful movements. my goal was to keep the monitor attached as that will help with immersion and opens you to tune the platform for more movement.
you mention the 0-180 movement,.. and you are correct in that there is a non -linear response, the code handles where the arm should be at a given time and accounts for the non linearity by speeding up as needed when more movement is needed to achieve a linear feel to the rider. as long as there was enough speed overhead for the motor to turn.
yes, my full heave is around 11.5 inches/ ~30cm,... its a interesting feeling during full speed heave,..it can move faster than gravity would have. i dont often do this test... but I have,.. for science.
The software is easy... its done... it uses Arduino IDE ... but you point it to a ESP32 microcontroller (much faster processor) the PCB is also done, you just send the gerber files to a PCB shop (jlcpcb.com ~$25 USD for 5 boards) and a few easy to assemble components,... buy a motor and gear to test... minimal investment so far,.. spin it up using the software "SimTools" connected to your favorite car racing game, if the motor spins,.. you are 50% there. buy more motors,.. and fab a chassis. one gentleman built his platform based off of mine in 3 months. I have been helping several others in their builds, its a growing community and we are trying to improve things continuously. I try to make myself available for questions... by all means,.. question. :)
@@Christopherknauf Thanks for doing this project open!! I did buy a DOF reality P6. I tried it with longer arms, (had to add air springs, dampers,) the overrun on down functions was uncontrollable, and not enough power to lift even with help from some pneumatic cylinders for adjustable air spring - so I had to shorten them a little.
yesterday I finally broke one off the frame at the mounting, (damaging that side of the geabox, now I have to mount it on end)
now I think that the gearing in the Worm Reducers included in the DOF Reality P6 is too high for the included motors. They're way too fast, and not enough torque. I believe they are 1500RPM motors about 345Watts, and a 25:1 gearbox.
so now, having broken a gearbox yesterday, I've been learning about worm reducers. I think I can replace the 6 gearboxes to double the torque, adjust some things in sim tools for the 1/2 speed, and fit longer arms for more travel. Might have to switch to planetary gearboxes when I push the limits on those and they break.
@@nashole23 there will be some tweaking in simtools for sure, watch out for backlash in the gear, my gears are something like 6arcmin?... and it wiggles about a mm or so @7 inches. no noticeable backlash from the riders point of view.. unlike my first gear box choice @ ~12arcmin backlash
What happens when you roll the car?
If you die in the Matrix, you die in real life..... lol... maybe not... but crashing can hurt/injure you if the settings/gains are to strong.... I have not ridden in the sim for several months due to back/leg strain from this .... for instance... currently rehabilitating ....
hows the poke feeling?
Hi. Will you sell the assembled platform ? Thanks
Ive been asked that a few times, I am not in a position to make more, nor do I have intention of selling this unit as I put two years of my blood/sweat/tears into building it... Though I am giving away the code and design to the world!... And helping many people build their own! I recently started making semi instructional videos of the process.
@@Christopherknauf but couldn't I pay you to make one for me?
@@bmarco1602 So, that's something I debated even,.. starting up a business selling these things,.. The design is not fool proof and can easily hurt the rider,.. there is allot of power on tap (~5 Miata's worth of torque), I have been injured several times from making mistakes. I am concerned with insurance/legal issues in dealing with safety/death/dismemberment... . One of the side effects I was looking for was when I open sourced this platform was to get other people involved so they could start making changes and evolving the platform (safety),.. there are a handful of folks building their own based on the design.
But as far as me building a custom bespoke one to spec,.. I would have to be heavily persuaded,... its allot of time for me to invest. I am open to brainstorming and shooting ideas around, though I am only one man... and only a hobbyist with the welder :)
Looks pretty impressive. Have you tried PT Mover yet?
No,...though i did see it a few months ago for the first time, I wrote my own algorithm for handling the platform and baked it into the micro controller.
You still use Simtools though, correct? Or did you write your own software to interpret the telemetry data as well?
@@NischGTM I do use simtools, it's been real good to me. I use much of the suite, got a fan on dash output, transducer on vibe and I've gotten real good at mixing the outputs to my sim
I would like to have that particular simulator for Forza Horizon 4!
Imagine FH4 with VR and full motion simulator!?
I definitely have used it in FH4, and it is absolutely fun.. you need to dial the forces back quite a bit as its arcady .. .. and you spend most of your time at high speeds... and stopping instantly is not out of the question,.... and that .. hurts. though I dont think it supports VR....
You should play with HP Reverb vr headset.
One day, this thing has been sucking my wallet dry for 2 years... But it's definitely in the future
It's amazing, for the moment you can try with oculus rift CV1, is quite cheap (150-200$) if you search it used, and if you put the sensor on the frame you'll have a self compensation of the movement
@@lucasozza2596 I'll look into it thanks
@@Christopherknauf i got 2 dof but i really want that Heave movement. Boucing up and down must be really fun sensation
@@ji3200 It iS!, this platform was a 2DOF prior to this, and heave is the most immersive improvement, you feel the bumps... its amazing!
Would be better with VR. Then you can't see the room so the efects would seem more realistic.
Agree, though I have since built a room for it that minimises the cluttered view... I am currently saving my dollars for a new project ..... I want to build a Shelby cobra .... I will accept donations of VR gear though in the mean time... :)
Damn,what is the size of your shoes:)?
Hey, no lookin at my feet (12)
Have you ever tried it with VR?
Not yet!, I want to .. but I also want to build a race car now that I am all trained up. I will accept donations of VR gear though at this time :)
@@Christopherknauf 😁
@@Christopherknauf for sale
Sim for sale.. Please
Talk to me whatsapp 0096550105716
not for sale, but the project is open source, with a link in the description of the video. there are several people all over the world using the project to create their own interpretation of this simulator. i am actively working on improvements, including safely features, and Android app over Bluetooth for configuration.
@@Christopherknauf Thank you for the clarification. 🌹 🌹 🌹 🌹 Want to connect to WhatsApp 0096599313976
@@Christopherknauf You should definitely patent it before some asshole makes money off your designs.