My 3 DOF with heave

I made progress.

Remember when I wrote, that inverting the axis assignments for roll and pitch didn't have any effect on MSFS? Today it worked. Roll and pitch work as intended now. I must have screwed something up when I tried it last time. There were some... external stressors that might have distracted me.

I did some flights with the little cessna over New York City and so far the rig seems to be able to carry my weight without getting deformed. Seems like I've stiffened up he right spots.

The mechanics on one of the motors makes a clacking sound every now and then. Feels like it happens on reversals, which might indicate slack. Its hard to tell whats happening down there when you are sitting up there.
Also, the motion doesnt feel very smooth. It feels like one or multiple motors move in small increments rather than continously. I need to look into that. 50% Smoothing on all axis didn't help, which should rule out un-smooth input from the game.

I'll try to look into recording motion and then doing playback. Thus i don't need to sit in the chair but can kneel next to it. This should help locate the source of the clacking sounds and maybe the reason for the choppy motion.

Irun Webcam free software to convert phone to webcam for recording, run over WIFI automaticaly

I am currently in the process of optimizing the SimTools axis assignments, filters and tuning center settings.

I started out flying a small single engine Cessna in Microsoft Flight Simulator 2020 and optimized everything for casual civilian flying (roll and pitch <20...30 °). I found settings that gave results that are satisfying for me.

But I also want to do wild stuff with military jets like the F/A18 in MSFS 2020. So i went and changed the settings i found for civilian flying. Roll and Pitch are currently based on roll angle and pitch angle (not roll rate and pitch rate), so i need to use the anti-roll-over filter option.

However, it doesn't work as expected by me:
https://drive.google.com/file/d/1PlwZu5xYFxhQnm-yjrZSsoj0w-Gp4mdK/view?usp=sharing

There is some delay between screen and seat motion, because of the smoothing I applied in order to reduce the violent nature of the motion that happens when i get inverted..

As you can see, the roll angle of the seat increases while my plane rolls from 0° bank angle to 90° bank angle.
When I increase the bank angle of the plane further (90 ° to 180 °), the roll angle of the seat starts to decrease.
As the plane approaches 180° bank angle (almost flies inverted), the seat is almost level again.
All this makes sense to me.

But then, when I exceed 180° bank angle, i get thrown around violently. This happens no matter if i do a right or a left hand barrel roll.

This doesn't seem right to me. I was expecting the rig to smoothly start to lean to the other side when i pass 180 ° bank angle.

Do you guys have any recommendation on how to fix this?

Unfortunately in the configuration of SimToools I am not able to help, in my 2DOF motion I use FlyPTmover, but I have in turn a question if you could explain the electronic part of your project. It looks very elegant and safe to use, I am currently using 250W motors but if I wanted to change them to larger ones I would need a hint on the electrical system. Do you perhaps have a wiring diagram ( even a working one).

You can still try the FlyPTmover program , it turns out that it supports 3DOF configuration like yours. I am posting a link to a video of a short flight of an F/A-18 in DCS and when performing a barrel the platform in the preview behaves correctly. I also post my profile for FLyPTmover, of course if you would like to try my profile remove the arms from the frame, the movement limits introduced may not be appropriate.

Noorbeast's advice was helpful. I deactivated all DOFs but roll and i found that anti roll over was working as intended. Then I slowly added the other DOFs one by one, always testing in between and was able to overcome all troubles. I think I can get SimTools to work the way I want it to, so I'll stick with it for now.

I'm still in the process of tuning, though. The F/A18 in MSFS does 140 degree per second roll rate. So after 0,64 seconds i am at 90 degrees bank angle and after 1,28s I am inverted.
With "Roll" as input and anti roll over activated, throwing the stick to the side will make the rig move to maximum roll angle within said 0,64 s and then back to center within the next 0,64 s. Thats quite a ride. Also, it doesn't feel right when the seat returns back to horizontal while I keep rolling the plane over. I guess anti roll over is not a good option for aerobatics and fighter jet simulation.

I am currently experimenting with roll rate instead of roll as input. Unlike roll+anti-roll-over, roll-rate doesn't cause a reversal of seat motion while I keep rolling.
But still it's quite a ride if I throw the stick to the right or left. Everything is cool as long as I do slow input changes, but if I move the stick abruptly, we'll, so will be the movement of the seat. Unfortunately, if the seat moves abruptly, this also leads to abrupt inputs from my arm to the Joystick. I had the situation that I wasn't able to stop the joystick from shaking because the seat (and thusly my arm) was shaking so violently. I had to press emergency shutoff...

Smoothing, Joystick deadzone, decreased axis range - all those might be solutions to this problem, but they all have serious draw backs.

@Attyla.pl
I'll try to come up with a wiring diagram in the next couple of days. Do you have a specific question?

@Attyla.pl
Attached you find the wiring diagram you asked for

Feel free to ask any questions.

I want to make it easier to get in and out of the seat. I found that there is a way to make the rig go to it's lowest position when outside of a game, by using the "Startup - Output" and "Shutdown - Output" options in the Interface Settings of SimTools.

For both I entered [A<001>][B<001>][C<001>] - see the attached picture, but it doesn't work. Am I missing something?

I still can't get the "Startup Output" and "Shutdown Output" to work. I noticed that the last screenshot I posted had no delay time specified, so i tried 10 and 5000 ms (see attached screenshot), but no luck. I guess not too many people are reading my thread, so I might try a new posting in the "DIY Motion Simulator Building Q&A" forum in a couple of days.

Meanwhile I wanted to share the fan mod I did for my PSUs. Each of my 3 Meanwell PSUs had 2 fans, each of the 40mm type, that where runnig at 5000 rpm (or something like that) when the rig was idle and even faster when the motors were moving. They made a loud whining noise.
I removed these 6 tiny fans, installed a 3D printed "adapter" and one single 140 mm 2000 rpm Noctua fan. It runs on 24V coming directly from one of the PSUs. See attached pics. It's much quieter now and the fan noise is much less whiny.
The airflow of the 140 mm fan is about 3 times higher than the combined airflow of the 6 tiny fans, so i should be safe regarding the thermals. The PSUs also have a over temperature protection.

I also added "workbench wheels". I need to move my ring before and after each use, so some kind of wheels was neccessary. Before, I used "normal" wheels that were mounted below the rig. But these gave me the problem that when the rig was moving a lot, the wheels added a lot of flex and bouncing back and forth to the entire structure. The new workbench wheels can be folded away, so the rig stands firmly on the ground when in use. This reduced the overall "wobbliness" of the rig.

very cool, but i see the motors has only 90Watts ? is this correct ? i think its a bit to weak for the leverage and the weight ? at the beginning h had 120Watt motors on my 6DOF rig and this was definitly to weak .... i upgrade to 6x 350Watt Motors

I agree, 90 W would be way to little. My motors Motors have 500 W, although due to issues with regenerative braking i had to reduce PWMmax so i only have about 370 W usable.

On Feb 03 I gave detailed info on my Motors:

About the motors (which have not arrived yet): I have ordered what seems to me like the 500 W version of the PGSAW crab pull motors that are often used by members of this forum. I found it on Alibaba:
https://x.alibaba.com/AvP1w3?ck=pdp

It doesn't cost much more than the 200 W version (that's commonly used here), so I went with it. Also, you can customize the shaft length before ordering, which I found very useful during the mechanical design phase. You can even order a shaft with D-cut, so no need for huge clamping forces on the lever arms.

Specifications:
Voltage: 24 V
Gear reduction Ratio: 50:1
No Load speed on output shaft: 72 rpm
Rated speed on output shaft: 60 rpm
Rated Torque: 56 Nm
Rated current: 25A

ok then i got something wrong ... i just see some "90" value on the sticker of your motors ;D

are these wormgearboxes on the linked motors reliable ? it seems that they are not oil filled just greased ?

Oh thats just the diameter of the motor housing.

Little stiffness update

I put in a lot of effort to increase the stiffness of the rig and reduce the overall wobbling around after sharp and sudden movements.

In the design phase in CAD I noticed, that I need to get the seat up quite high, to avoid touching ground with the pedals when the seat pitches foreward.
Dummy me came up with the idea to do that by using long connectors between the motor levers and the seat platform. In the end, this caused a lot of lateral flex and wobbling back and forth and left and right when the seat made sudden movements.

Now i use connectors that are as short as possible (to decrease lateral flex) and as long as necessary (to allow the desired range of motion of the motor).

I also placed the bottom platform on stilts, to get enough ground clearance.

I attached a before and after picture.

Another problem was rotational flex around the vertical (yaw) axis. This was caused by my decision to put the 3 motors (and their connection points to the seat platform) as close together as i could. I did this to get maximum angular displacement of the seat. But the price was way too much flex around the vertical axis. Everytime I did a sudden roll maneuver, the whole seat kept wobbling around the yaw axis - my rig isn't even supposed to do yaw!

Now I placed the motors and their connections farther away from each other (they form a bigger triangle now). Thus I get about 25 instead of 30 degrees of either pitch or roll (if I would allocate 100% axis respectively), which is still enough for my needs and the rotational flex around the vertical axis is now negligible.

I consider the stiffness of the rig quite good now. I did a videos analyis of me using the rig while doing some wild maneuvers in a F/A 18 and the only thing wobbling around now is the squishy biomass in the seat (aka "me"). The top platform itself basically sits still as soon as the motors stop moving.

I also got annoyed by the whining sound my 3rd motor made everytime it moved. This was caused by the Arduino only being able to control the 3rd motor using 4 kHz PWM frequncy, which is quite audible. The only other available frequency for motor 3 is 31 kHz, which my Cytron drivers can't handle.
Motors 1 and 2 can be controlled by a variety of frequencies, from which I chose 25 kHz, which is the highest my Cytron drivers can handle.

So I added a second Arduino for controlling motor 3, also using 25 kHz. Works like a charm now.

About the Startup/Shutdown Output:
I started a dedicated thread about this and got the answer that these options are basically broken in SimTools v2. They are supposedly working in SimTools v3 but I'm currently too occupied with other stuff to make the switch. Meanwhile I added some kind of steps to the rig, to make getting in and out easier. Hardware modification to counter software limitation.

so you do only flightsimulation, right? ok, very obvious because theres no mount for a steering wheel
very nice work!
is this the normal position in a plane for the pedals? looks like sitting in a normal chair with legs hanging downwards. for example, on my rig the pedals and stuff goes straight forward and not downwards, so theres no problem touching the ground while tilting forwards
are you satisfied with the aluminum build and the screwed joints ? you mentioned that is stiff enough, i can just talk about my expirience with mine, its a hybrid steel and aluminum rig, its a welded steel frame for the lower and the seat part and everything in front is made from aluminum profiles and they had a bit of flex because of screwed joints and because its a weaker material. the steel part is not bendable at all ;D

Thank you for your kind words and your interest.

Currently, I'm only interested in flying games, especially military flying games.

As far as I know, military as well as most civilian pilots sit like you described, with pedals way below the seat, just like on a normal chair. See attached pic. This is the way i saw it in all the sims I've played.
Sail planes are an exception (there might be others), where the pedals are at about the same level as the seat.

I am satisfied with the 4040 aluminium profiles and the screwed joints. You just need a lot of them
Especially on the lower platform I needed additional struts to increase the torsional stiffness of the rods the motors are attached to. I also needed to combine inside corner brackets and external connection plates on most connections. Please take a look at the attached picture.
The result is as stiff as it needs to be.

I have to admit though, that I have overestimated the stiffness of the aluminum. While doing the design in CAD I didn't plan for those anti-torsion struts. But the alumium extrusion profiles made it easy to rearrange components or make struts shorter where needed. So I'd say aluminium is more flexible than steel in two ways:
1) it bends more under stress
2) you can execute design changes easily, often without creating any dirt or noise at all (no need to go into a workshop and whip out a welding device or an angle grinder)

But I also have to admit that this adaptability of aluminium would not be needed as much, if it was as stiff as steel in the first place.

Still, i don't regret my choice of working with aluminium profiles. I like that I can change stuff so easily. It's almost as easy to work with as wood, while being much stiffer. Steel would be another leap in stiffness, but you'd need to have skills, equipment and location for welding. And its way harder to change something about a welded steel frame.

I'd definitely advise against 3030 (or smaller) aluminium profiles. 4040 is the way to go, maybe even 4545. But only go 4545 if you get all the connecting hardware you need in that size. I'm under the impression that 4040 hardware is much more common.

Also, stay away from those light and extra light variants of 4040 profiles. You need all the stiffness you can get.
Maybe for the seat platform, where weight matters, you could get away with reduced stiffness but I wouldnt risk it. The aluminium extrusion components of my seat platform weight about 15 kg while the whole platform (including me) weights about 120 kg. Saving 20% weight on the 15 kg seat frame would mean a reduction of 3 kg, so the whole seat would be down to 117 kg, which is 2,5% less than what we had. I think this is negligible, while the reduction in stiffness, at about 10%, is not.

thanks for your explanation, thats an amount of bolts you have ;D hehe .... yes thats why i choose aluminum for my front rig, because of the flexibility to change things as desired!!!! and i did change a lot since the original setup there ;D the rest ...stays fixed , no need to do something there so i welded this stuff. for me it was enough to do the front part with the "light" 4040 Profiles do reduce weight ... it is already very heady this way so i put a gas strut in front to compensate that, and it works fine.

i do mostly racing sims so i have the pedals in one height with the seat ... but not like formula cars thats a way to uncompfortable and i think my setup is a good compromise also for flight sims ;D i just do msfs and if star wars squadrons count .... i have also some other dogfight flight games but not tested them yet

Yeah, when I was talking about avoiding light profile variants, I should have added "for structural parts".
All the stuff thats used for holding peripherals can be done in 4040 light, ultra-light, maybe even 3030 (although it will be difficult to join 4040 and 3030).

However, the more restrictive the light variants are used, the smaller the overall weight reduction will be.
But you still lose flexibility, because a ultra light strut, that was meant to be used for the seat platform, but became obsolete by a change of plans, can not be repurposed to increase stiffness in the lower platform, should the need arise (well, of course it can be used that way, but it will give less stiffness than the full weight variant).

I think you went for an absolutely sensible compromise when you've put your pedals at about the same level as the seat. Some issues in mechanical design get solved this way and i think it can be done without sacrificing comfort. To be honest, i simply never thought about going that route. The thought just didn't come up.

Check out this thread https://forum.dcs.world/topic/117783-f-18-cockpit-geometry/