DIY Motion Cabinet: Is it even possible?

[framed_niner]

There are several arcade cabinets that had deluxe motion cabinets. After Burner, Cruisin USA, Daytona USA 2, and modern games like Fast and Furious Arcade and Typhoon. These games had cabinets that rotated, tilted, shook, and were overall some of the most immersive games ever made. Unfortunately, the full motion cabinets in the arcade are not able to be replicated at home. Or are they? There must be a way to recreate these in DIY cabinets. Here are the main obstacles that I can think of

Mechanism: How will the cabinet physically rotate?

Encoding: How could the system be controlled by a computer?

Software: How can the emulated games interface with the motion cabinet? Maybe custom games are the only solution?

And of course...

Safety: How to prevent major accidents that could cause damage to people and property?

That last point is pretty important. Anyways, maybe we could use this thread to brainstorm how this could be done? Or maybe I am just flat out insane for thinking this is possible.

[PL1]

Related thread:
https://forum.arcadecontrols.com/index.php/topic,164999.0.html


Scott

[lilshawn]

i don't know that i'd go with arcade games specifically... ideally you want to start with a game that has some kind of access to it's internal telematics so your interface can take that data, parse it, and turn it into a signal stream you can use with a controller (Arduino or other programmable microcontroller)  for motion (motor, light, valve, solenoid, etc) most arcade games are proprietary and you'd have to manually load configurations into your controller to get it to work with more than one.

for PC games, simhub is a good start. it basically wedges itself into games and extracts data out for use with your interface. it's racing focused so if you are looking for something specific it might help to start there and see how people are doing motion and integration on other games first... there is tons of videos out there on simhub integrations.

[RandyT]

I might have to check out Simhub.  I see that it supports my controllers now, and it seems that Thanos calls it the "best" motion software.  The joystick hook functionality also seems to be part of the base functionality of the software.

Yeah, the arduino approach is the cheapest, but certainly not the easiest.  I think it can only do a couple of motors and seems aimed at the windshield-wiper-motor seat rockers.  If I was to try to build something like that, I would make sure I used some very high-quality pots (it doesn't seem to be set up for encoders) and build it in such a way that they are easy to replace.  They will be getting at least the same amount of wear, if not considerably more, than what happens to them in an arcade controller.  Jitter on-screen is one thing, but it could be catastrophic for a motion platform.

But one thing I will say is that nothing is absolutely safe, especially with a roll-your-own solution made from Chinese Arduinos and used wiper motors.  Safety is designed in at the beginning, by looking at every possible fault situation, accurately predicting the result and then designing in appropriate mitigation to minimize possible harm in that eventuality.  Just that aspect alone pretty much nullifies the idea of doing something like this on-the-cheap and being able to be confident that no-one will be injured.  Those who do take these and similar routes, including myself, understand that there are some risks and have chosen to take them in order to be able to enjoy the tech within their respective budgets. 

I think something many don't fully realize when looking at the motion arcade machines is that they are expensive, heavy behemoths which likely had their movement hobbled in some way just so they could ensure safety for the player.  When they suffer a fault, the worst case scenario is an attendant might need to help them out of the seat.  Replicating that at home, within a budget, might be possible with some high creativity and building skills, but it's not likely going to be possible for the average person.

[pbj]

Randy, stop being a wet blanket and go into detail about how far your took your project.  I’ve got $6 in hand for video of you being thrown out of it.

I sold a fully functioning Miata for $2,000.  I don’t think any arcade experience is going to top that, but if you’re differently abled then do whatever you have to do… I’m a waddling barrel and I could still maneuver it.

[RandyT]

Randy, stop being a wet blanket and go into detail about how far your took your project.  I’ve got $6 in hand for video of you being thrown out of it.

I sold a fully functioning Miata for $2,000.  I don’t think any arcade experience is going to top that, but if you’re differently abled then do whatever you have to do… I’m a waddling barrel and I could still maneuver it.

My post shows pretty much the extent of where my project sits currently.  Honestly, I'm lucky I had the time to get it that far.  It's functional, but I wouldn't call it "safe".

Trust me, the $6 is nowhere near enough for me to show you the pain that thing is capable of inflicting.  Whiplash is a real thing, and the forces are enough for it to destroy itself. The right malfunction or a misconfigured application is all it takes to put a non-trivial hurt on the machine or me.  This is from experience, whether you believe it or not.  Each of the actuators I designed are theoretically capable of flinging ~250kg (~500lbs) close to 10"/second, based on similar designs, and there are four of them on the machine.  In other words, with a good platform, there's more than a small chance that the 4 units could lift your Miata 10" off the ground in not much more than a second.

God help you if something goes wrong and you aren't able to shut it down, as was once the case when I stupidly tried configuring an untested roller coaster whilst in the seat of the thing.  It was throwing me around so bad, it took a good 10 seconds for me to hit the escape key to stop it...on the keyboard that was on my lap!  That 10 seconds felt like 10 minutes.  The machine had thrown itself off it's floor pads because it was violent enough to only have 2 of the 4 actuators in contact with them at certain points during the ordeal, and it felt like it was close to flipping on it's side.  My neck ached for a week afterward.  That reminds me, I really need to get that industrial slam kill-switch installed. 

I'm not trying to dissuade anyone.  But the hard truth is that (almost) anyone can build an arcade cabinet that won't fall over and hurt you.  But not everyone can design something with enough power to move in near real-time with a person attached, and do it safely.  Driving a Miata is not the same as designing one, and if one average person engineered it, you would be foolish to drive it.

If you aren't an experienced engineer/builder (or at the very minimum, so inclined), and have a limited budget, just go here.  It's about as cheap as these things get and you still have to assemble it, but at least the hard work is done and you will be unlikely to hurt yourself or loved ones.

[pbj]

This one time, I got suckered into participating in a "do you like video games" study.  Basically they put you on a platform in a basement that could move in any direction, you sat a foot away from a screen that completely filled your field of vision, and they'd have you move around in some video game that kinda looked like Battlezone.  The joystick and platform input would conflict with each other and they wanted to see what it did to your brain and eye activity.  So far, so good.  Well, they custom molded a mask that immobilized my head, and I was too fat for the seat belts so they duck taped me to the chair.  At one point, some of the equipment fell off and hit me in the face and bounced into the pit below the platform.  When I told the researcher afterwards what had happened, he started yelling at me in Greek and wanted to know why I didn't say anything.  I mean, you've got me in a basement (in Texas), all lights off, sound proof walls, duct taped to a goddammed chair.  Throwing stuff at me would have been the least crazy part of it.

They also wouldn't let me me play Mario Kart on it after the session was over.  Jackasses.  But that was the most precise and immersive motion thing I've ever experienced.

Here's my contribution to science:

https://elifesciences.org/articles/63405

[lilshawn]

[RandyT]

youtube...

Now just imagine what would happen if one of those upper ball joints was out-of-spec, the engineer did not fully consider the forces they would encounter, or all of the possible angles combined with those forces.  If one of them snapped, that design could allow the two remaining supports, combined with gravity, to angle his body to face directly at the broken actuator arm, which would still be functional and possibly drive it into all of his important squishy parts.  Things like this are the ultimate confidence test of one's own engineering prowess and is probably representative of the first stage of becoming a Darwin Award nominee.

...But that was the most precise and immersive motion thing I've ever experienced.

Here's my contribution to science:

https://elifesciences.org/articles/63405

What they had you on was a commercial-style simulator platform of the type used in aeronautics/NASA/etc.  There are massive ones using this design, which have complete cockpits and nosecones sitting on the platform.  They are usually referred to as "6DOF" simulators, as they can move in all directions in 3D space.  Due to the fact that there needs to be 6 actuators to pull this off, they are the most expensive type, and it's probably one of the safest types, given the number of supports and their distribution.  The home version from the same company I linked to earlier can be found here, however, their design is of the glorified wiper-motor approach, which limits the range of motion to some extent.  This makes it inherently safer, but probably not as immersive as something which could literally put you near 90 degrees relative to the ground plane in any direction, if the actuators have enough travel.  Some have rolled their own versions of this type from common materials to pretty good effect, but using them is akin to stepping aboard a thrill-ride at the jankiest run-down carnival you can imagine and expecting nothing to go wrong.

The sad part about your contribution is that the Aperture Science guys probably threw out your results due to the testing anomaly, which is why they were upset.  They didn't offer you cake to get you to participate, did they?  Spoiler: There is no cake.   

[pbj]

Yeah, and he also yelled at me for not going exactly over the targets.  There was inertia in the video game, so I kept trying to come to a sliding stop.  Apparently I was supposed to drive all the way to the target and release the joystick.  “Stop taking so long!”  Has nobody played Crazy Taxi?

The time perception study may have been worse, that guy had a show on PBS.  Hours spent in a sensory deprivation closet saying whether number sequence one or two counted down faster.  Made $40 on that one.  That scientist got famous for throwing graduate students off a cliff and testing if time dilation during a traumatic experience is real.  (I’ve experienced it, it’s real)

Neuroscientists are all nuts, and I’m married to one so I’m allowed to say it. 

[RandyT]

That scientist got famous for throwing graduate students off a cliff and testing if time dilation during a traumatic experience is real.  (I’ve experienced it, it’s real)

I don't know why they call it "time dilation".  Time doesn't get larger/longer, it's just your perception of it because the processor clock for your brain goes into overdrive due to the desire for survival.  Hard to think your way out of a bad problem if you only have seconds to consider your options.  It's probably really bad for your brain and body, or I would think they would be synthesizing the hormone so you can do it on demand.  They probably already did and found out that for every second your brain was running that fast, it took a month off your lifespan.  Candle burning twice as bright and all that.

Neuroscientists are all nuts, and I’m married to one so I’m allowed to say it.

Don't be too hard on them.  They do what they must because they can (is it getting old yet? )

[Xiaou2]

One of the problems with certain motion sims, seems to be a lack of speed in their responses... likely also depending on their range of motion travel.

 The sitdown version of Outrun, suffered from this.   It was driven using a worm gear system...  and the motor speed, in combination with the
worm gear ratios... resulted in a slight delay, from what was happening on-screen.

 I had this idea to solve this... and not sure if its valid or not... but here goes...

 Rather than using a single Actuator, which will have a hard limit to its "inches per second"...   Imagine using Two actuators together, on the same axis leg.

Basically, if you had a 2 axis motion sim... which would typically use only 2 linear actuators..   Now it will use 4 actuators.

 I believe you should end up getting Double the Distance per Time Unit... as you are expanding 2 drivers on the same axis, at the same time.

[======]---0

vs

[======]---00---[======]


One axis would be mounted to the floor-base..  and the other, mounted to the platform... and both actuators connected together in the middle.

[RandyT]

One of the problems with certain motion sims, seems to be a lack of speed in their responses... likely also depending on their range of motion travel.

 The sitdown version of Outrun, suffered from this.   It was driven using a worm gear system...  and the motor speed, in combination with the
worm gear ratios... resulted in a slight delay, from what was happening on-screen.

This is what I was referring to when I stated that arcade motion platforms are hobbled for safety.  They don't really attempt to provide a full-blown simulation, rather they attempt to give the player a "taste" of movement associated with what is happening on the display.

The Outrun machine is a really poor example of a sim.  So much so, that I hesitate to give it that title.  But there are no issues with it needing to be fixed, as it does exactly what it was designed to do.  The only thing one could do to improve it (i.e. if building from scratch) would be to completely discard the design and build something which behaves the way it should for use in an environment where you aren't worried about a lawsuit .

TLDR:  Motion machines don't "suffer" from anything other than the design intent of the machine's respective creators.  Be it for safety or cost, they do what they wanted them to.

Side note: I see that there is now a YAW3, which comes after seeing complaints from at least one YAW2 user indicating that they have had more down time, due to mechanical issues, than usable time with the unit.  I also see that it is very important that the load be balanced over the center of the machine, requiring some substantial counterweight to the back when accessories are added to the front (there is literally an iron lifting weight box behind the seat.)  The max load on the unit is only 287lbs, which is a little better than half of ONE of my FOUR actuators.  I suspect anything added to the machine, including counterweights, subtracts from this number.

I have to give props to those folks for their perseverance if nothing else.  Designing a marketable system with the smallish footprint, range of motion and price point that this has (even at $5k) is no simple matter.  But at the end of the day, it's still going to be something best suited for flight-sims.  While it seems that heave is something which is being considered, I have to wonder if the approach they are taking to the rest of the machine would continue to make sense if that feature becomes important to their consumers.  That stated, I suspect that the range of motion it offers (in the full 3-axis system..I still can't tell if the yaw is included in the base "Yaw3" configuration ) would lend itself well to any and all of the old arcade games, so long as proper motorsports sims weren't a major concern.  But at ~$5000+ said and done, potential purchasers really need to make sure they understand what is included, identify their use case and make sure it does it well.