Marco's 2DOF motion simulator based on Thanos' design

Hi.

I'm Marco from Germany.
Since I'd like to obtain as much input and feedback from you guys as possible I'm going to keep this in english rather than my mother tongue.

At first I'd like to thank the community for this incredible board. I've been into racing-games for quite a bit, remembering an old F1 game, Lotus Turbo Challenge and Test Drive III as the first games I really enjoyed playing whilst still being a little chap. Time flew by and the PS with the incredible Gran Turismo caught my attention with its fifth coming just months away. And this game really got me hooked. I built my first Cockpit for GT3 with a racing seat and the Logitech DFP. Years later, with the PS3 an GT5P I now have the G25 and a refined Racing-Rig.

[img600px]http://img594.imageshack.us/img594/2341/rig1.jpg[/img600px]

During my interships (becoming a Mech.-Eng. very soon) I welded a few pipes together and included a few details like the changable inclination for both the wheel and the pedals as well as blue lights (just for effect).

Just recently, whilst googling for new ideas on how to improve such a rig, I came across this website, or I found Thanos' youtube video of his simulator, and I thought: you absolutely have to do this, this is what you always wanted.

Further research on this site brought this thread (http://www.x-simulator.de/forum/sony-may-change-peripheral-policy-of-ps3-to-open-t1805.html#p17349) to my attention, accompanied with some disappointment. Yet I'm monitoring it in constant hope that someday (very soon ) Sony will change it's policies and give us the info we need to make GT drive the motion simulators' actuartors. I've played a bit of GTR2 and LFS as well, but none of those have that X-factor for me, so I'm really hoping here. Yet the possibility of building and later experiencing a 2DOF motion simulator is something I really want to do. And if the PS3's peripheral policy doesn't get changed I still would have a really nice simulator with GTR and LFS to enjoy.

My plan is building a 2DOF simulator which is based on Thanos's modified Joyrider build. To do so I'm working my way through:
http://www.x-simulator.de/forum/step-by-step-approaching-to-thanos-work-t1218.html,
http://www.x-simulator.de/forum/mario-s-tronic-joy-based-sim-t2315.html and
http://www.x-simulator.de/forum/viewtopic.php?f=25&t=559.

I still have some reading to do, but I'm doing it thoroughly, building an excel file filled with the partlists needed for building it. Thanks to my experiences with metal, my plans are to build a lightweight metal construction based on the principles of Thanos' simulator. I'm planing on using square-pipe, joined together in stiff welded frames. I already looked up square-pipe dimensions with weight/length ratios in the regions of the PVC-pipes used by Thanos, Mario and Knob. A general design-fault I could already make out is the problem with stepping in and out of the simulator. That's why I'm planning on building a structure that simplifies entering and exiting the simulator. Another problem for me would be making the PCB-board. But the solution I found for this problem is a website that makes these for 20-30€.

So, what are my next steps:
-reading everything in the aforementioned weblinks
-building a CAD-model that gives me the measures for my pipes
-get a light bucket racing seat
-build the frame as a start

When this is on the way I'll parallely start with the electronics part, ordering the stuff needed for the controlling, incudling the PCB-plates and the wiper motors. I hope my smoldering-skills haven't disappeard altogether... :
I'm thinging of using a drive belt for the transmission of the forces like Thanos recommended to reduce maintanance.

So, that's it for the moment.
Great forum guys, keep up the good work and I hope to be able to contribute with an own simulator in the not to distant future :hi:

Greetings,
Marco

Re: 2DOF motion simulator based on Thanos' design

Welcome Marco!

Good luck on your simulator and keep us informed!




PS. I think I saw someone here doing a quick screen capture hack of a (sony?) console racing game to read the acceleration data using some simple photoresistors... I'll try to find it again and give a link here... :lol:


Thanos

Re: 2DOF motion simulator based on Thanos' design

Hi Thanos

Nice to see the man who inspired all of my plans answer so fast to my post.
I actually read that thread already (http://www.x-simulator.de/forum/2do...-sato-t2405-30.html?hilit=gran turismo#p24099), but I don't think it's what I'm looking for.
I want to get the actual acceleration data, not a combination of a screencap for roll and accelerator/brake pedal input for pitch data.

btw:
Do you have any idea on the weight of your pitch-enseble (the pitch pvc-cage) without the seat, wheel, monitor and person?
So I can at least roughly calculate if I'm in the same regions?
Because, like I wrote, I'm planing on doing quite a bit differently, but without getting to heavy (dynamics, but whom am I telling ^^).

Marco

Re: 2DOF motion simulator based on Thanos' design

The weight of only the PVC structure is minimal. I can calculate its around 15-20Kg... but with all extra motor mountings , supports, screen and seat, reaches up to 60kg... I can still drag it around the room if I need to move it!

I wanted to make it from metal frame when started but the lack of workshop and welding machine made it impossible for me. I had to settle with the easier to work with, PVC pipes...

Thanos

EDIT: Just noticed... ONLY the inside cockpit bare PVC???? not more than 2kg...!!! LOL

EDIT2: Maybe you could rename this thread with a more relevant name since it about YOUR simulator: Marco's 2DOF motion simulator based on Thanos' design

Re: 2DOF motion simulator based on Thanos' design

What?
Really?
This light?
I don't think I can do this with steel. the weight/length ratios i found for PVC tubes of the sizes you recommended were higher.
and not just the bare PVC, that's missleading.

Let me rephrase that.
What I really need to know it the weight of everything that needs to be moved/accelerated including every mountingpoint, but without the seat, wheel, monitor and person.
And including everything.
Then I can see where I have to save weight to be able to build my stiff construction.

Re: 2DOF motion simulator based on Thanos' design

Welcome Marco and good luck!

From your post to me it is obvious you are more of a racer than a flyer. In that case I would recommend you to build a seatmover style simulator. It will give you more realistic driving experience, a lot more responsive for road surface and smaller bumps, while stimulating the right muscles to give you the g-forces. The joyrider style frames offer greater actual g-forces, but the very large movement range and the relatively slow wiper motors cannot reproduce the road and surface very good.

The AMC has both of these options built in (JoyRider or SimForceGT). There are seatmover projects based on wiper motors and AMC on the forum, check them out!

Regarding the materials and weight of a joyrider frame I posted in Marios thread:

As a mechanical engineer I'm sure you understand what I mean.


ps: IMO this is the best approach so far, I hope it will be completed: articsimulator-joyrider-made-of-wood-t2063.html

Hi Frakk.

Thanks for the info.
What fascinates me about the Joyrider Style are the great g-forces you mentioned. And I'm actually quite fond of a Flight Sim, yet I don't find it as fascinating as a good racing game. With Thanos' design as a base I could easily do both and have strong feedback. For things like small bumps etc. I'm concidering a buttkicker or something similar.
What I've seen of the seatmover-fraction isn't as appealing to me. The smaller movements create smaller g-forces, yet better responsiveness. What I really like about them is, that they seem far more compact in comparison. A joyrider alone needs a vast amount of space already...

I think a seatmover would be a lot like this: http://www.youtube.com/watch?v=hjXjCGgoQEg, which I could experience on the 24h vom Nürburgring (the 24h race) in 2008. The experience was more shakey than feeling the ac/deceleration and cornering forces. My reference in sporty driving is go-carting and whilst doing this I actually dislike the bumpy ride, yet I enjoy the forces in the corners.

About the materials: I'm aware of the different materials' densities. Yet for rigidity a good amount of steel always does the trick

I'm going to inspect the articsimulator now.
Another idea I have is contacting that guy from Frankfurt (not too far away) I saw somewhere on the forum. I think it would be nice to see/experience his rig, because I think his simulator is a seatmover.

Edit:
I just came across this thread http://www.x-simulator.de/forum/oscarh-s-homebrewed-2-dof-car-simulator-t1359.html?hilit=oscar with the original design by tornado-blu altered by OscarH.
I think this could actually be the best solution for my problems, because it combines easy access, high g-forces due to large angles and a compact overall design!
Another point is: it's already a steel construction, thus easy to realise for me.
The only question left concering the angles is this: how large are your maximum angles, Thanos?
OscarH/tornado-blu's rig as 30° in each direction.

Seat movers work on totally different principles to simulate g-forces. They move you from the stationary pedals and wheel, producing muscle tension the same way as the forces in a real car. If you want a smoother ride, there are lots of ways to tune it down.

I'm not trying to persuade you either way, just giving my opinion. If you like flying and want a general purpose sim, the joyrider is the way to go. OscarH's simulator is very similar to the articsimulator in construction. If I was to build a 2DOF seat, it would be very much like the artic sim.

I know steel is good for rigidity, but the whole platform will suffer from the extra weight and inertia. The electronics and motors will run much hotter, speed and response will be much slower. You won't get the full motion for faster changing forces because the platform won't have enough time to reach it's position.

hi marco,
i am agree with frakk,a raceseat make many more fun as a joyrider for carracing.
i have both if you want you can test it by me.were is your town?

Thank you for the kind offer.
I'm from Darmstadt, just a few km off Frankfurt.
So to Bad Zwischenhahn it's quite a bit of travelling...

With the Wochenendticket (weekendticket) I'd be on the Bahn (railways) for a whole day just getting to you and back.
Reading up on other threads I spotted a mod from Frankfurt.
I'll ask him, maybe he'll agree for me to visit him.

And I'm all undecided now.
Build a universal simulator (flying and driving) or build a race-sim that totally rules in reactiontime and every single detail but lacks a bit in g-force-quantity...

So many choices.
But, as far as I understood all of this, I can do both with Thanos' AMC V1.5.
So I'm gonna make that partlist now anyways and start with the soldering and get the wipermotors.
Using wipermotors is set in stone for me, because I want to keep it cheap for my first one. If everything goes well I still may take the second step and try a scn-configuration on another project.

Using Thanos' AMC boards you should build a Joyrider.

The reason for my suggestion is that it is a Proportional controller only. If you understand PID control loops you know that P-only control is the slowest and least reliable (compared to PI or PID control). You can make it go faster, but the platform will overshoot (doesn't stop where it should) and start shaking to correct the position.

Based on this fact, the AMC will perform better with a slower platform with more travel, and not so good with a faster one with much less movement. If the AMC had full PID control and the dSMHB was optimized for it's purpose, performance for both platform would be 10x better.

I don't know if the rumors are true about the ULTIMATE MOTION CONTROLLER (UMC), but it would be great if Thanos improved on his old design.

Hmm, I'm getting more and more confused.

So which kind of controller would you suggest in the case of a seatmover?
And is it equally easy to build?
My electonics-skills are limited to soldering. That's why I like Thanos' AMC, I just have to get the PCB and all the parts (sounds easier than it is btw ^^) and solder them on the PCB.
Once I get it all connected and programmed (hope I can get some help from a friend at the university who won the robocup soccer worldcup in his class this year) the easy part, building the rig itself, can start and I'm back on my hometurf...

Right now there is no competition for the SCN5 actuator based seat movers. You can pick up only the actuators, a powersupply and 2 USB to RS485 converters to make them move.
I would recommend the Starter Kit from SimXperience: http://simxperience.com/SimXperiencePro ... fault.aspx
It has everything you need, all in one box, you just plug it in a USB plug, install the Commander/X-Sim, punch in some numbers and you are ready to go.

These actuators and electronics will give you the best performance you can get at the moment for a seatmover. I know it is a bit pricey for your budget, but in reality they are very cheap for what you get, and the support from SimXperience and the forum is something you won't find anywhere else. There is a lot of good info and dimensions on the frame, building it will be very easy for you.

Building the Thanos' solution you can calculate the price of:

-PCB's needed for the 2 boards + LCD if you need that
-Price of all the components + blown/damaged components for safety
-Price of the wiper motors
-Cables, connectors, AVR programmer
-Your time to build it and make everything work + the frustration and time trying to find errors if something doesn't work.
-The lack of warranty if anything breaks

You will come up with the cost of materials only, and you can decide on how much your time is worth to put it all together. If you are lucky and have an electronics friend help you out, the build time isn't that bad.

In the end, you still cannot even compare the motion quality with an SCN5 seat. Even low quality 30fps youtube videos show the difference in speed and accuracy, then you can imagine the difference in the feel!

There is no doubt, that the SCN5 outperform any wipermotor based setup, but even with the AMC and some wiper motors you can realise quite good and fast effects, depending on kp settings and XSim configuration.
http://www.youtube.com/watch?v=WVqUM4XR ... 8BY-LaPfM8

2 wiper-motors ~20Euro + AMC + h-bridge pcb and parts ~100Euro. That is a total of round about 120-150Euro, compared to a SCN5 setup for ~800Euro.

I agree that the build of the AMC and the pot mounting can be a pain in the arse, but you told us that you are familiar with soldering. Now you have to decide for yourself, if your priority is rather to save costs, or plug-n-play high-end actuators.

Regards

Edit:

But the feeling of success, once the setup is working, is really great, on the other hand ;D

Hehe, made me grin...
I actually really like the building-phase, if it doesn't get too frustrating.
And since I'm still a student and this is my first motion simulator I'm going with Thanos' AMC and an arcticxsim - OscarH-style Rig. My current living-situation and budget both tend to this.

I still can build a new one with the experience from the first one as soon as I'm in a job, hence earning some cash...

I think so far all the possible problems which can occur during the building process have been sorted out, so you can profit from the experience of the other members, who built an AMC already.

Btw, if you are going to solder Thanos' DIY h-bridge, you rather should use the logic type Mos-Fets Frakk suggested, than the ones in the original parts list, to prevent overheating.

Viel Erfolg!

You are in luck because others in your country have already built the solution and can help you with shopping lists for your parts. In some countries, I think it is very difficult to find all of the parts.

Please keep us informed. I have always been curious how long the average DIY'er spends assembling the PCB. I suspect 2-4 months from research / part shopping till completion, but I am not sure about this since in some countries, a shopping list with is already available and sometimes, unpopulated PCB's as well.

In any case, welcome to the forums!

:

No doubt! But it is not for everyone. The more mistakes and errors there are, the more we learn and the end is even happier. Given there is enough will and sanity to carry through.

@Bernard: It is hard to say how long it takes to build the AMC+Hbridge. As you said it really depends on country, parts availability and also the individual's skill level/electronics background. Not to mention the amount of time spent on it each day. I would say 2-4 months is realistic for someone who never done anything like this before. Someone with experience and parts ready to go could do it in a day.
The assembly and soldering only takes about 2hrs, it is the troubleshooting and tracking down some problems that can take days or more. I think a lot of people underestimate the complexity and difficulty that is involved with electronics assembly, even with some very simple boards like the AMC.

Anyhow, Marco I think you are doing the right thing. You will learn a lot by building everything, and also save money. : :thbup: :clap:

Thanks guys, for all the moral support so far.

I think I'm as lucky as Thanos concerning the electronics side...I found all major parts in a local store's online shop (Zimmermann Darmstadt, http://shop.zeg-shop.de/shop/index.asp saying it's all available in the store in at least 5 days
Yet I didn't find some stuff, like the sip headers or the ALPS switch, but I think that's only because I don't know the german terms. I think it would be easiest to contact a german fellow x-simer (if I may already call me that ^^) and ask him/her for his/her german partlist. It would make things a lot easier for me.

About the wipermotors:
How do I know if these suffice?
http://cgi.ebay.de/ws/eBayISAPI.dll?ViewItem&item=140437918287&ssPageName=STRK:MEWAX:IT
They're from the backwindow of a T5 VW van. And I could get 2 motors which are alike.
Do I even have to buy 2 motors that are the same? Or could I buy to different ones (e.g. a Skoda and a Fiat)? Do they have to be from the front or doesn't it matter?
What concerns me is, that powering and controlling those if different could produce different intensities in movement...or am I talking rubbish here due to my less-than-beginner electronics knowledge?

Almost all of the components can be substituted to something else. You don't have to use the exact same connectors or ALPS button. These buttons go on many different names and button heights, just look for momentary tactile switch.
I personally prefer the breakable socket and header rows like this:http://media.digikey.com/photos/3M%20Photos/929400-01-36.jpg
The only downside is that they don't protect you from plugging it in backwards. Their socket counterpart on boards is great because you can use small wires to make connections, without any pins.

As for the motors, there has been a lot of the same question asked and discussions here.
The only advice you will get is: depends on the motor and your geometry/gearing on the sim. Do you need more speed for higher gear/pulley ratios? Do you need more torque with no external reduction?

General guidelines:
Always try to match your motors for setups where they move simultaneously. Ex. would be seatmover setups, 3-6DOF platforms, where the position of each servo is tied to the rest of the servo positions. Different motors can cause asymmetry of motion and/or wrong movement in 2-3DOF setups, or it could even damage the frame above that.

If you build a joyrider frame where each motor controls each axis, you can use different ones. It might be better to use different ones since the range and balance is already different for each axis, therefore the motors need to match these differences for optimal performance. I would only experiment with this in the later stages though, for now just get the same 2 motors if you can.

I would stay away from window lift motors like the one in your link, or those sliding door opener motors. They will be way too slow or very weak.
Don't worry too much about the specs as they won't tell you anything and you have very good options to reduce/increase gearing on a joyrider frame. Try to find a larger WIPER motor from a larger car or truck with big windows. (isn't it logical?)