Flight simulator motion platform for home use.

G

Thread Starter

George Fitch

What type of equipment is needed to duplicate commercially available motion platform designs using off the shelf equipment? Platform would be controlled by a PC running Microsoft Flight Simulator 2004 with commercially available flight data extraction software (FSUIPC at http://www.schiratti.com/dowson.html), with the motion cue and control software written by me in Visual Basic or Borland Delphi.

I've been interested in constructing my own 2 or 3 axis motion platform for some time now. Scouring the internet, I have come across some interesting designs; some good, some bad, and some really ugly-expensive!

Two sites have almost the same design for 2 and 3 axis (or DOF, Degrees Of Freedom) all-electric actuated motion platforms:

www.inmotionsimulation.com (check out the videos with Flight Simulator 2004!)
www.servos.com

Both use what looks like 230v, .5 to 1.0 HP, AC gearmotors connected to some type of controller / drive / inverter box, and a sensor to determine the shaft position. All of this is wired to a PC via a controller(s) to control the motor positions and ultimately control the platform's pitch, roll and heave (up/down) movements.

I know these guys are not using proprietary hardware, because both use the same controller / drive on some of their models (and even the same gearmotors, at times):

http://inmotionsimulation.com/f-18.html
http://inmotionsimulation.com/videos/boatsim5.wmv
http://inmotionsimulation.com/videos/boatsim7.wmv

http://servos.com/lowcostload.htm
http://servos.com/3axisprod.htm

Both want $15,000+ just for the platform, and I don't want to spend more than $10,000 (long term) on the entire project including platform, simulated fuselage, controls, computers, displays, etc. The guy at inmotionsimulation.com looks like he builds them out of his garage! $15,000? Yeah, right!

I'm looking at constructing a 2000 pound capacity platform, 2 axis to begin with (pitch/roll), probably 3 axis in the future (heave; up/down).

Can anyone please point me in the right direction about the following:

- What gearmotor speed and torque ratings should I be looking for? Anything else I should know about gearmotor specs?

- What do I need to control an AC gearmotor's exact shaft position, reversing, etc.?

- What is that controller / drive box hooked up to each motor? What function does it provide?

- What PC-based controllers are available that can be programmed via Visual Basic or Borland Delphi (or their own language, if it is easy enough) to control all of this?

- What else would I need to hook all of this together that I may be overlooking?

- Any other comments? Am I nuts? Should I even attempt this? Am I going to kill myself taking a simulated Cessna for a little spin?

I have many years of electrical (just not industrial motors and controllers) and computer programming experience, so assembly and testing shouldn't be a problem. Programming the proper motion cues should be fun!. :) I just don't want to purchase a $400 motor or $600 controller and find out it doesn't do what I need it to and end up with a paper weight. I'm also looking for some first-hand experience about what I should and shouldn't purchase (brands and models) and where I can get it for the lowest price.

I would appreciate any help that is offered.

This is a great site, thank you in advance...
George
 
C

Curt Wuollet

I'd start with something faster than VB and more universal than Delphi. Like C for instance.

Regards

cww
 
Hi George,

Some of these tips may be old news to you. This forum is great for your questions:
http://forums.frugalsworld.com/vbb/forumdisplay.php?f=48

Also http://www.AVsim.com is the premier site for your needs.

While we don't offer servo drivers at your level of need some of our products might be of help. They are all easily used in VB. Our page is here:
http://www.phidgetsusa.com/cockpit_simulators.asp

Details the relavant products for your Sim. It may be possible to use some of our interfaces to control your larger motors and/or serovos by adding in some of your own custom electronics. This would significantly bring the cost down.

We are also working with a gentleman who is building a software interface to Peter's software which will make interfacing the Phidgets with MSFS plug and play. I might be able to have him get you a copy ahead of time if you are interested.

I hope this helps you out!

Matt
(you can email me back through the site)
www.phidgetsusa.com
 
E
I've always wanted to do this myself.

The control boxes look like VFDs. I'm guessing they have found one that will hook up to a serial port to take speed commands. Otherwise you would have to produce +/-10Vdc. Those look a little like Mitsubishi VFDs, I would check on ebay and the web to do a direct match. The features available for those drives may tell you how the control scheme works.

> Both want $15,000+ just for the platform, and I don't want to spend more than $10,000 (long term) on the entire project including platform, simulated fuselage, controls, computers, displays, etc. The guy at inmotionsimulation.com looks like he builds them out of his garage! $15,000? Yeah, right! <

I must say that $15000 isn't bad for a properly engineered motion platform -- even from a garage. The parts alone would cost something on the order of $10k if you have to buy them retail. Of course, we know that Ebay is your friend.

> - What gearmotor speed and torque ratings should I be looking for? Anything else I should know about gearmotor specs? <

The gearmotors would be uncooled. Most motor manufacturers have ratings for this.

> - What do I need to control an AC gearmotor's exact shaft position, reversing, etc.? <

One word: Home switch. You don't need servomotors for this application. The VFD's will handle spinning the motors and reversing. You could certainly use servomotors, and it may be easier.

> - What is that controller / drive box hooked up to each motor? What function does it provide? <

That is a variable frequency drive, aka VFD. It makes the motors spin.

> - What else would I need to hook all of this together that I may be overlooking? <

If they are using analog control, you need to be worrying about where you are getting analog outputs.

> - Any other comments? Am I nuts? Should I even attempt this? Am I going to kill myself taking a simulated Cessna for a little spin? <

There is a crushing hazard underneath. You have to be willing to investigate the structural requirements. There is some first year undergraduate level engineering
work that you need to do to keep this safe. And that you shouldn't solve by asking questions on the internet.

You need a system that can be shut down with a "big red switch" and limit switches. Things will go wrong just as you are pitching over into a bombing run.

Finally, I wouldn't do it with Windows. Windows is not intended for real-time tasks, and your performance may be really bad at times. The problem is that Windows will occasionally ignore your program for a little longer than is safe -- think crushing hazard.
 
Eric,

How can you hold position with a non servo?
What other motor would you use for this application?

Dritan
 
I just came across this very interesting website and have been reading some of the inputs from your members.

Like most of you, I have been thinking of designing a 2 axis simulator now for some time, but the biggest issue has been an easy way of extracting the necessary data out of FS2004.

There are many but most require additional data processing, etc, which complicates the entire project.

I recently came across an English website that manufactures an external USB based card that allows the connection of real looking intrumentation gauges that work like the ones displayed on the FS2004 intrument panel.

The gauges have a small servo inside, as used in model RC aircraft to operate the display. The one which is of interest here is the attitude indicator which has the X/Y information.

Now, since the gauges use standard RC servo signals, it would be very easy to feed this information into a large servo motor to control the platform.

It is very easy to make a large RC servo based electric motor. I have taken a windscreen wiper motor assembly from a truck and using a high current RC driver circuit can attain upwards of 200kg of force at about +/- 30 degrees/second when geared correctly with a reduction ratio of about 15:1.

That's more than enough force and speed to move a single person [perhaps more if the C of G is set correctly.

Whilst I have not purchased the USB card, I cannot see any reason why this extremely simple and cheap system will not work.

Obviously I'm open to suggestions or other ideas.

rgs

Peter
 
F

Francisco Lobo

Ok, i am trying the same thing... Here are some pointers...

The difficult part is the mechanical part, you need to move your weight and its heavy (in my case). Using linear actuators might be a good choice, but DC motors can put much more power to it than AC motors Although for DC you will need a powerfull power supply.

go to: http://www.npcrobotics.com for possible motors.

How to control speed:
One simple word: PWM (pulse width modulation), this is how you control speed in DC applications with the most efficiency, i will not cover the hard part of theory, but go to:
http://www.roboteq.com for a possible dual motor control unit.

Hot to control position:
Feedback Loop or PID, this is what servo uses. Home switch is not a good idea, overtime you will drift. use a potentiometer connected to the roboteq unit or as a feedback loop to your controller.

How to control the roboteq:
VIa a serial port with simple commands. You can get (in x-plane) the data out by means of UDP, i could write a visual basic software that can control the roboteq unit easily...

In other words, someone design a good base mechanically and i can do the electronics and positioning... We need a good mechanic. I can handle electronics...
 
Hi there,

Yes, window whipper motor should be fine as I have built a 30 degrease moving cockpit all working around c/g, and this is what I am looking at using. The thing is I'm not up with the data side of things, all switches working in my cockpit and are looking at gauges at the moment. What I'm not sure of is what motor controller do I use? It will need to be able to control stalls, weather, guns and cannons, roll rate and pitch rate, etc. Another motor I thought of was a garage door opener, as they to are very torquey motors.
 
M

michael coats

It has been several years since you posted this inquiry. I have the same interests and have spend very recently (last 12 months) investigating all methods and vendors to build a 3 axis motion base for less than $5,000. I have been told and I believe correctly so, it's impossible.

Impossible only with current methods. Therefore some engineering has to be done.

I have developed a dual place 737 simulator with a rudimentary 2 axis motion base for about $3000 U.S. including all computer hardware, enclosure, mechanics and electronics. But I am interested in further developing a better motion base. The one I built is a electrical screw-drive. Using 4 of these I am able to obtain about 30 degree of pitch and roll, but slowly. Like a jumbo jet but not suitable for a fighter or small private prop plane.

If you are interested in working together as well as anyone else we might form a consortium.

email me at [email protected]
 
V

Vince McCarroll

I have a working model of a 3 axis motion base for simulating high performance aircraft and baja race cars. The base uses 3 pneumatic actuators, with external inexpensive resistor position sensors, and a simple 3 axis controller. The mechanical components are less than $1000 and the electronics is less than $1200. The power source is from a small air compressor that can deliver approximately 80psig at 3gpm, enough to provided a reasonable response for +/- 25 deg/sec pitch and roll and 4 ips heave. Higher rates could be achieved with a lager air volume.

I can be contacted at [email protected] PS not avail from 2/6 until 2/27

Vince McCarroll
 
Top