F1 like 3DOF Simulator with heave

I have started my own project for a F1 like 3DOF Simulator and hope to succeed with you guys helping me finding the white spots.

So far I have started buying the Fanatec Equipment and bought a ready to rumble rig from SimXPro:
Fanatec CSL DD 8nm, McLaren GT3 V2 Wheel and CSL LC Pedals.
Rig:
https://www.simxpro.shop/de/SimXPro®-FR-DD-LusoMotors-2S-Bundle-p361045733

It's all built up. The next step is to add motion.

Following the DIY 3DOF Tutorial of @Sebastian2 @https://www.xsimulator.net/community/threads/my-3-dof-with-heave.18449/unread I'm trying to figure out what worm Motors I need.

Therefore I have the following calculation and hope that someone might have a look at it and does a quality check.

As a first example I took the motors of @Sebastian2 Post and recalculated his steps:
The rig including me has a weight of 170kg (All values are rounded up and the result contains some buffer).

The chosen motors have 24V 500W 50:1 with 60rpm.
https://german.alibaba.com/product-...&ckvia=share_a315b8bf3b884ec4acc78e95e2b8b61e

Torque_per_Motor = 9550 * 0,5kw / 60rpm = 79,583nm

I calculate according to @Sebastian2 the CTC with 100mm.

Static_Load = Torque_per_Motor * Number_of_Motors / CTC
Static Load = 79,583nm * 3 / 0,100 = 2387,49 N or 243,52398kg

Conclution: The three motors should be able to move enough weight.

Acceleration = effective_Force / Mass = (Force_from_Motors - Force_from_gravity) / mass
= (Force_from_Motors - Mass * 9,81 m/s^2) / mass
= (2387,49-170*9,81)/170 = 4,17 m/s^2 or 0.425222g

heave_distance = CTC * sin(angle) = 0,100 * sin(40) = 0,064 m or 64 mm

Linear_speed = 2 * Pi * CTC * RPM/60 =
2 * 3,14 * 0,100 * 60 / 60 = 0,628 m/s or 628 mm/s

So what are your thoughts about this calculation and the selected motors?

Hi codac, I feel honored to have done something that you feel is worth taking inspiration from.

I think the torque you calculated is too optimistic. The 500 W in the product description are rated electrical power, which is higher than the mechanical power at the output of the gearbox. Those worm drive gearboxes are quite inefficient.

You provided a link to the motor you think about buying. If you scroll down on that page you can see a spec sheet saying the rated torque at the gearbox output is 56 Nm (for the 500W 24V 50:1 variant).

It seems like you are using the same motors like me. I am very satisfied with the power my rig has, even though I currently only use about 330 W per motor due to issues with regenerative current.

On the other hand, the moving part of your rig (incl. yourself) is 50 kg heavier that mine. I guess if you find a better way to deal with regenerative currents than I did, you will do just fine.

I'm curious on how what you are planning to do with the SimXPro rumble rig in order to get it moving. Will this whole thing be moving around? So you are just planning on building the base with the motors under it?

To be honest, look at you posts (regarding the 3D prints for example) Your posts are extremely structured, detailed and easy to follow. Thank you indeed for that, as it helps novices like me tremendously.

I haven‘t bought a motor yet, as I would like to know what I really need to be safe.
What torque do you think would I need in my case? It‘s really hard to find good motors, as Simukit is offline since a while and it‘s not clear wehen they come back.
I‘m planning to use 2 x sabertooth 2x32 with appropriate resistors to use the regenerative currents.

Yes, the plan so far is to build a plate underneath it with the motors attached and metal rods between the motors and the rig.

Thank you for your kind words!

The required torque depends on the design of the rig (often called "inherend design leverage" on this forum) and your requirements regarding acceleration and travel.

Lets start with the later. I've built my rig with focus on flight sims. Thus I wanted huge angular displacement. Your rig seems to be focused on racing. I assume you require less travel (I might be mistaken here!). But I guess you do want to be able to faithfully reproduce imperfections of the road and hitting the curb. So high angular acceleration might be more important to you than high angular travel.

So let's come back to the design aspect. You are planning to use 3 motors, which will likely form a isosceles triangle, possibly a equilateral triangle. You will have to decide on the length of the sides of this triangle. Shorter sides make for larger maximum angular travel of the seat, but require more torque for a given angular acceleration. The stiffness around the yaw axis is also decreased.

If you make the connection between the top platform and the motors at the outer corners of the SimXPro you have bought, you will have quite long triangle sides. Or as they often say on this forum: you have much inherent design leverage. This means you will not need much torque for a given angular acceleration, but your angular travel won't be too big either.

Have you found other 3DOF (with heave) rigs on this forum that also feature a a huge top platform like you are planning on? What motor power, ratios and CTCs are they using?

Also: I was under the impression that most 3DOF for racing have traction loss instead of heave. Am I mistaken? Or is there a reason you want heave instead of traction loss?

First of all @Sebastian2 thank you so much for taking the time and explaining everything, I really appreciate that.

You are absolutely right, its about the acceleration, street imperfections, curbs, etc.

Another project that is quite the goal I would like tonget close to is the following.
He has 28 degrees of roll, 22 degrees of pitch and uses roll, pitch and heave.
His reason for the heave is:
„I have found adding true heave to be one of the best 3dof upgrades and really boost the immersive feeling of actually being in the car.“

His motors are 440W, 1:60, 3000rpms, CTC I dont know. But he has a by far more light weight rig, he lowered the seat so that the lowest point is below the frame and extended the ledt and right side of the frame.

https://www.xsimulator.net/community/threads/f1-style-2-dof-mantis-simulators-fs2.16071/

https://www.xsimulator.net/community/threads/f1-motion-sim-upgrading-2dof-to-3dof.18398/

The roll and pitch angles you mentioned are total values, right? So, referring to a leveled seat, you want to be able to roll up to 14 degrees to the left and up to 14 degrees to the right, so that makes a total of 28 degrees, right? I prefer to call that +/-14 degrees as I find this leaves less room for (miss)interpretation and is still quite short way to say it.

Are you able to design your rig in CAD? It really helps to determine where you need to place the motors and whar CTC you need to achieve a certain angular travel. Virtual trial and error tends to be a lot quicker than real-world tests.

This rig you linked to looks nice. Also, I noticed that the top part seems to be way lighter than the SimXPro stuff you've bought. Just look how he used mostly 40x40 profiles, while the SimXPro uses a lot of 40x80 and 40x120. I'm not sure if so much stiffness is really needed. It sure adds a lot of weight. You could ask early_m if he is satisfied with the stiffness of the top part of his rig. It might turn out that it was not wise buying the SimXPro seat.

On the other hand, you can compensate your higher weight by using stronger motors. It also seems like early_m uses gas struts. I just took a quick glance at his build threads, but I saw 200 N gas struts being mentioned. 3 of them would "neutralize" 3x200N/9,81m/s^2=61 kg of static weight (though they have no effect on angular acceleration). So you could go with 500W motors and 300N gas struts for example.

As you are considering the same type of motor as I bought, let me give you a piece of information I wish I had before I ordered: make sure your gearbox is of the NMRV-Type, like early_m did. This is a standardized type where you can get repleacements relativley easy. The gearbox I ordered (and you are considering) is of a non-standard-type which is hard to find a replacement part for. Why would I need a replacement? Because I'm not satifsfied by the ammount of backlash my gearboxes developed after some hours of use. It's not really bad, it's just not as good as i want it to be. But I'm stuck with a non-standard gearbox, a motor with non-standard gearbox mounting holes and a non-standard motor-shaft. There's no way to put a lower backlash gearbox on my motors.
NMRV-Gearboxes have a little more friction (thus slighlty lower output torque - 50 instead of 56 Nm for the 50:1 ratio, see attached picture) but they are supposed to have a lot less backlash. All while costing the same.

You can open the alibaba link you posted above, get in contact with the company sales representative and ask for a low backlash NMRV-variant of the motor you are currently looking at. The representative I've talked to was really helpful.

Yeah, you're right, its total Pitch and Roll.
Well basicly I guess I could CAD give a try, I have very basic skills in it.
I've seen another base plate approach at the same post of @early_m of @Michail with his CAD draft. I would go for the way you described with the triangular fitting of the motors.


I'm just asking myself which program to use that is suitable for beginners also. And how would I derive from a CAD model which CTC I would need to use, therefore I would need to simulate the movement, right?
I have contacted @early_m but as he is currently selling his solution professionally, he already stated that he has no time in helping people from the community with their individual needs, which is quite understandable.

Exactly, his rig is very low profile and of course I noticed that his profiles are slimmer and therefore lightweight. And you are right, buying SimXPro might not have been a smart idea. But now I would like to make the best out of it.

I have weighted the rig just now and it has 56kg all together (Wheel, DD, WindSim, Flaglights, etc) and me having 95kg. So its 151kg currently. If I rememeber correctly @early_m weights 20kgs less, so thats quite a difference.
As I'm not going to add more weight to the rig, I would require a buffer for my own weight that differs +-10kg from time to time.

Thank you for the explanation of the gas struts so that seems to be a feasable solution.
Could you explain what difference of weight you are referring to, when you suggest 3x300N gas struts = 91,74kg?
And another question: You say I should go with 500W motors, but before you said that the rated torque at the gearbox output is more relevant?

And thank you even more for your hint with the NMRV-Type.
God you really are pushing things forward for me. Thanks again @Sebastian2

I have an offer for the following motors:
NMRV24V-90ZYT 500W-input speed 3000rpm, output speed 50rp/min, 1:60
Guess that sounds quite good, no?

You are welcome. I've read a lot of stuff on here before I started building, so now its time for me to give something back to the community.

About CAD software:

Simulating motion is not necessarily a requirement. This would imply that moment of inertia as well as oscillations etc. are simulated. This would go way beyond the usual scope of a DIY motion simulator.
I think most, if not all, CAD software today works with something called relations. Example: the motor with its cylindrical output shaft can be one part, and the motor lever with its cylindrical hole can be another part. Both can be joined by using 3 relations.
First, the motor-saft and the hole in the lever shall be concentric.
Second, the end of the shaft and the outer side of the lever shall be in the same plane (so the lever can not move in the axial direction.
Third, the angle between the motor lever's centerline and the motor shall be a number determined by the user. Then you can type in different numbers for the motor lever angle and the CAD software will move the lever accordingly. If you also connect the top platform with similar relations, it will also move when you change the values for the motor angles.
So it's all based on geometry alone. There is no actual force, torque, velocities, mass or moment of inertia. But it let's you see how much roll and pitch angle a combination of motor position, shaft angle and CTC will give you. You can also check if there are collisions between parts.

During my study years, I came into contact with different CAD software, namely SolidWorks, Pro/Engineer (renamed to Creo nowadays) and CATIA. To me, Solidworks was, by far, the one with the most modern, intuitive and user friendly interface. That's what I use for my hobby projects today (you might want to get creative acquiring a license though, or get ready to spend a lot of cash). For my work, I sometimes have to do basic stuff in Creo and I am in awe how the interface designers managed to make it so convoluted.
There might be other, even free CAD software that is better suited for your needs, though. Never looked into that topic.

Learning CAD for building a motion sim might be overkill. But I guess if someone decided to DIY a sim, he has a history of DIY projects and will take on new projects in the future. CAD might be a nice tool for those, too. Also: having basic CAD skills "unlocks the 3D printing skill tree" which can turn out to be very, very valuable for a lot of projects - including the motion sim (think mounts for electronics and accessories).

About the gas struts:

I suggested to use 300N gas struts instead of 200N simply because your top platform seems heavier than the one early_m built. I don't know how much heavier, though, so 300N was just a guestimation. Three gas struts with 100N more each would result in 3x100N/9,81m/s^2= 30 kg more weight compensation (regarding static situations). This seemed like the right ball park for the weight difference to me.

You might not want to compensate all of your weight with gas struts, though. I wouldn't suggest using much stronger gas struts. Static weight can be your friend when it comes to suppress effects of backlash. In the end, gas struts can be quite cheap (20 Euros each in germany) so looking at the equipment you've already bought, I assume you can afford some trail and error with gas struts.

About Motors:

In was suggesting to use a higher powered motor assuming you would keep the ratio and speed constant. So higher power means higher torque. And you will need that to achieve as similar performance as early_m, because your top platform is heavier.

The specs of the motors you've come up with are looking good. Is there also a 1:50 ratio option? It probably won't make much of a difference, but early_m also used 1:50 if I remember correctly. If your ratio goes too high, you will have more torque than you need while not being as fast as you want to be. It's just very hard to tell beforehand where the threshold is. In theory, you can compensate a high ratio thats too high by using longer motor levers. Longer leavers also mean more maximum heave travel, but some say backlash might get worse.

You see, everything is connected to everything, there is no objectivley optimal solution. If you optimize too much for one thing, another thing will get worse. It might be a good idea to stay close to what's proven to work unless you have additional time, money and nerve to spend on experimenting. That's why I suggest keeping a similar ratio (and thus speed) as early_m, and only compensate your higher mass with more power (and thus torque)

Alright, I'll have alook through different CAD solutions and try to model my case.
3D-Printing btw. is a skill I have already. That is no problem at all.

Your explanation about the gas struts makes now sense to me.

Regarding the motor: Yes there is a 1:50 version available but @early_m has 1:60 motors.
The sales representative first offered me a NMRV40 motor (the one I've mentioned above) with a torque (500W) of 2.65N.m, the output torque is related to the speed ratio, and the final output torque is about 150N.m
He then offered a NMRV50 motor (500W as well) saying that it would be much more stable at its rotation as the NMRV40 has a smaller gearbox and a shorter service life. Both versions can reach 150N.m. but NMRV50 gearbox carries 150N.m more smoothly.
The specs of NMRV50 is attached.

Is there anything I could do, to further make sure if the motors are a good fit for my project, before I order?

Ok, if early_m has 1:60, I would recommend you follow that lead. Your rigs are quite similar after all.

I agree, you should use the NMRV50 gearbox. Bigger gearbox means lower specific load on the teeth, so lower wear and lower backlash after long term use.

Do the motors come with a shaft or do you need to acquire them separately?

If they come with a shaft you should make sure they are long enough for your purpose (usually this can be customized without additional cost). Also, think about how you will fix the motor levers to the shaft, as this affects if you need cylindrical shafts or a D-Cut.

BTW, the 150 Nm are most likely stall torque, so they are more of a marketing figure. After all, they are 3 times as high as the 56 Nm rated torque of the motor you looked at before.

The motors come with a double shaft. The technical drawing says it has a M10x24 screw hole on the outer end. For the length it says 50 each, I guess thats 5cm for each shaft.
Do you have a source for good levers?
I love the ones of simukit, but as I said they are not available.
The ones that I found are using teeth to grab the shaft, but I don't think that this fitting is strong enough?!

https://amzn.eu/d/09q2Rvls

https://amzn.eu/d/03sJDnu9

At @early_m solution it seems that the screw goes through a screw hole right through the shaft, right?

About motor levers:

The shafts you've linked on Amazon are about 30 mm long, so CTC will be less that that. I think you might have overlooked that, as I can not see how one could achieve any substantial angular travel with a CTC so low.

Sourcing suitable motor levers is not easy. From what I've read on this forum, lots of people have them custom made. They either have access to a CNC mill (or know somebody who does) or get them manufactured in China (via Alibaba I guess).

I've built my levers myself, based on some aluminium blocks that I ordered from Ebay. For tools, I only used a common drilling machine, drill stand, drills and a tap. I guess that's quite commonly found hardware found even in basic workshops. Alternatively, all of it can be bought by like €/$ 100...150 in good used condition.

I described the process to make the levers on the first page of my build thread. Just look for "drill" and you'll find it. I'm too stupid to provide a dircet link.
I recommend not using the measurements I presented in the mentioned post in my build thread, though. The lenght of the tap was too short and couldn't stand the torque I needed to apply on the screw. For my second, working, iteration, I used the measurements found the picture I attached to this post.

The motor lever concept I used works best (requires?) a motor shaft with D-cut.

I am under the impression that levers with clamping concept and cylindrical shafts are more common on this forum. early_m also went that route. One could also make them in a basic workshops out of a simple aluminium blocks. You would need an electric hand saw and blades suitable for cutting aluminum.

About motor shafts:

I agree, it looks like early_m has a hole in the motor shaft in a radial direction. I think that is a good solution. Question is if you can get your hands on such a shaft. Drilling it yourself with basic tools like I mentioned above might turn out to be a challenge. Turing lots of aluminium into chips is easy, doing the same with steel can turn difficult fast.

You should be sure about the lenght of the shaft that comes with the motor. Ask for a drawing if in doubt. They should have no trouble providing one.

Also, be sure the shaft is as long as you need it to be. On my rig, the minimum shaft length was dictated by the connector thingy (the one that connects the motor arms with the top platform). More precisely, it was dictated by the corner brackets I used. I needed a shaft that's quite long in order to not have the bracket collide with the motor, when the motor lever moves down. Check the attached picture.

Its really easy to check for such collisions if you have a movable CAD representation of the seat you want to build.

Sourcing or fabricating the lever arms can be tricky. On my first seat I was going to make them just like the ones shown in the previous post out of 12mm thick aluminum. I had planned for the screw to set down in the keyway slot on the gearbox shaft. Somewhere I changed my mind and instead bought some 300x50x6mm pieces of aluminum off of Amazon and fabricated some that didn't use a set screw. I used a few basic shop tools, drill press, band saw, hand files, and a belt sander. You could use simpler tools, but it would be much harder to keep things aligned.
The first set I made still had some alignment issues, so I had to make a second set. I used them in pairs on each side of the gearbox.

I used some hand files to make the key slot, which was a lot of effort. Getting that keyway straight from one lever to the next was the really hard part. I 3d printed a sleeve and a cap to go on the shafts to hold the levers in the correct place.

Overall, I would say the ones shown in the previous post would work great and be far easier to make if your gearbox shaft has a flat or a keyway in it.

I‘m impressed that a 3D print could hold the lever… Did you use 100% infill PLA?
The work you have done is tremendous.. wow.

I found the following link that was quite useful in addition to your infos.
Having a look at the technical drawing, I could use the solution of @noorbeast or @Nick Moxley.
Both would require some welding.
The Shaft has a M10x24 screw hole at the end of the shaft and a key slot.

The 3d print is just a spacer sleeve that fits on the shaft and then an end cap to keep the levers from moving in or out on the shaft. They really don't have any load on them, the torque is transferred thru the key on the shaft.
Here's a picture of the pieces after I removed them from one of the Gearboxes.

Hope that makes it clearer what I was talking about. I don't think you could 3d print anything reasonable to take the torque that is on the lever arms or shaft.

@Gefahren Oh yeah, that makes it more clear.
@Sebastian2 How much did you pay for the delivery of your motors?
I think I have a reasonable offer for motors + individual customized couplers. But the delivery costs are just totally insane.

Sorry for taking so long to answer

I paid 80 USD for railway shipping to germany. It should have taken 35 days but it was more, like 50 days or so. The train was held up in belarus, so this might have been due to the war.

For 160 USD shipping cost, there was an option to get the motors via air mail in 6...8 days but i decided against that.

I had all the mechanical and electrical work to do which even took a little longer than the motors took to come via rail.

Long time no update, so finally I have some good and some bad news.
Good news first:
I've ordered 3 x NMVR50 Motors with 3000rpm, 500W, 1:60, 24V and double shaft each at Alibaba and the merchant :Xuzhou Yinxin Electromechanical Equipment Supply Station

Costs:
NMRV50-60:1 90ZYT-24V-500W-3000rpm 3 x 125USD = 375USD
Double Shafts customized 3 x 9USD = 27 USD
Couplers (fo attaching the lever) customized 6 x 9USD = 54 USD

Good thing is they arrived safe and sound.
First bad thing: They only accept UPS Express Saver which is extremely expensive with 270USD (keep in mind that the weight was ~30kg); delivery was within 3 days.

Now, one even worse thing happened.
I've wired up the first motor following this tutorial only thing I didn't do is attaching capacitors to the sabertooth 2x32.

I've followed each step from setting up the baud rate for the sabertooth to 115200 up to step 17 with increasing PWMmax. With every 1-step increasing PWMmax I could hear a silent "click" sound. At around 18 the motor started so rotate very slowly. When I clicked on 19 the motor stopped suddenly. Since then I can not achieve any more to get the motor running. Increasing PWMmax has no effect, I hear no clicking sound. Restarting the Arduino didn't help, restarting SMC didn't help, swithing the motor off and back on didn't help.

The PU Input measures 24V on the sabertooth, the Arduino measures 3,3V at the sabertooth but no matter how high I go with PWMmax (max was 50) the connection to the motors remain at 0V.

Any suggestion how I can get the motor back running? Any tests I can execute to find the root cause?