Help building a full motion enclosure using MSFS2002

[Guest]

Having had a fascination with flight simulators ever since I first played with a simple simulator on a Sinclair ZX81, I have always wanted to go that extra step to realism (You do get utterly hooked on realism over time, especially if you have flown for real!!!!)With the latest incarnation of MS Fight simulator, you can get pretty close (barring the odd glitch!) You can now run multiple screens on a pc, ie have 2 for outside views, and 2 for instrument views. Using part of the MSFS SDK you can use netpipes.exe and run multiple pcs per flight in real time across a LAN, or use wideview or similar, and hence have more than 4 monitors hooked up to give a very real view inside and out of the cockpit. Add plug in switch gear from 'goflight', as well as realistic controls (including throttles etc) and things get pretty cool. Put the lot in an enclosure painted and shaped like a cockpit and things start to get very cool. Check out this static simulator at www.rcsimulations.com for an idea of what an be built.However, I have always hankered after the simulated movement that you get in a top end simulator (haven't we all!!)Having finally sold some other toys, I now have some spare dosh to at least have a go at constructing a simple movement enclosure. The idea is to build a cockpit box, populated with all of the above (4 monitors for outside view, 1 or 2 for instrumentation, additional switchgear controls etc) and then put this into a simple frame that allows 2 axis movement, pitch (probably limited to about 30-40 degrees, and roll (upto 360 degrees depending on weight and height of enclosure)I have already designed the mechanical systems, and have the movement control software ready to go. The enclosure design is also completed.The issue now is geting the dynamic in flight plane pitch and roll parameters out of MSFS in a useable format. I have downloaded the "netpipes" exe, and samples, but cannot make head nor tale of the code that is there ( I am not a c++ programmer)What I want to do is write a little program that dynamically writes the pitch and roll as a simple number (current poisition in degrees for both pitch and roll) and then pipe this to a file, or file handle, so that I can use my hardware control software to move the motion enclosure. It doesn't even need that many updates, just 1 or 2 a second would do.I have already put this request into the panels and gauges forum, and I am getting hep using the attitude indicator gauge to provide some of the information.has anyone here done something similar?please let me know.

[Guest]

TJ.DALE,You need a hardware to act on your mechanism and this move your frame.You say that you already have the software , if so , you have the hardware , right ?If the software is compatible with the hardware ,I don't understand what you need.I have a software that works with the chip FT639 from Ferrettronics, which moves RC servos for pitch and roll, the servos follow the pitch and roll of the plane ( FS2002).rgdsAlberto Kunzel

[Guest]

Hi,Have you checked out the FSUIPC SDK? There is a large list of the available offsets in a word doc that has pitch and roll listed multiple times. It's pretty simple to interface to FSUIPC, even I've been able to do it and all I've done is read through a vb book. http://www.schiratti.com/dowson.html you can find the SDK about half way down the page and it includes code samples for C/C++, Delphi and VB.Hope this helps,Spencer

[Guest MikePowell]

Hi,By all means, take a look at FSUIPC by Peter Dowson over at www.schiratti.com, and FSConnect by Russel Dirks posted in the Avsim library. Both utilities have sample code with them. Pulling pitch and roll out of FS is not hard. What sort of algorithms have you developed to process the FS data?Roland vanRoy has experience with a home brew motion base, and is a good resource in this area. He posts here on occasion.Mikewww.mikesflightdeck.comInfo for simpit builders

[Guest]

Hi Alberto,many thanks for the reply. I have not written an interface with FS 2002 yet that extracts the pitch and roll information so that I can control movement hardware (motors)I have designed the movement enclosure, as well as have the hardware to move the enclosure. What I need is the ability to control the large capacity motors using output from FS2002. Your software / hardware system that moves RC servos sounds ideal. Is this something you wrote yourself? Could you explain more? Could I have the code, as well as understand more about the ferrettronics software and hardware you have used to do this?many thanksTJ

[Guest]

Thanks for this Mike, do you have an email address for Roland, or do you know if he has a website?many thanksTJ

[Guest]

Tj,First of all go to www.ferrettronics.com and read about FT639 chip.The soft I have extracts from FS2002 pitch,roll,air speed , position of the throtle and hydraulic pressure of the plane , I am doing a 737 cockpit.The RC servos move as the conditions change and you can connect the output of them in what you need, as in one switch which will on an electric motor which will move the frame, the potentiometer of the servo will be replaced by one coupled to the frame ,so the frame will stop in the right position. Hope you are familiar with servo operation, if not, look for servos operation.The soft was made by a friend and I have no problem to send a copy for you.RgdsAlberto Kunzel

[Guest]

Alberto, sounds perfect.please email the softare and any other information you have (controller board build etc) to tj.dale@dial.pipex.comI have no problem with servo operation etc.many thanksTJ

[Guest MikePowell]

TJ,Roland does not have a web site yet. He posted a reply on the Aural tactile transducer thread on July 10. You can send him an email by clicking on his name in the thread then on the send email link within his profile.Mikewww.mikesflightdeck.comInfo for simpit builders.

[Avcomware 1]

Hi.For TJ: Before you go FSUIPC, consider the implications, compatibility and now cost etc.What I have, but I am not completely satisfied with, is a Controller / Joystick that I built, basically 8 micro switches that drive the motors and a joystick located under the seat that actually drives the FS joystick / Gameport / USB port device. The idea was to make it universal and without the need of software interface, This may be something that may work, but I think it

[tigert 0]

Also note that just moving the platform like an artificial horizon is a lot different from what for example the Big Bucks simulators do - you should investigate "motion cues" from google or such, there must be some research going on and maybe something published too. Basically, to simulate acceleration, you can tilt the base back slightly. This combined with the visual image staying horizontal in your "window" gives an idea of acceleration since you get pressed against your seat. Also since the base cannot turn indefinitely, one needs to "wash out" the movement so that it starts crisp, and slowly fades - this should give an idea that the movement continues. Same with many things - my impression is that it is a lot more than just taking the plane's attitude from FS and moving the platform like that. Plus I would guess if you want to do motion, an absolute must is a completely enclosed simulator to eliminate all external visual cues.Not that I know how to do this, but this is my impression on how the motion systems work.Tuomas

[Guest]

Tuomas, many thanks for this.you are quite right, there is far more to this than first meets the eye. Feedback from several forums has shown that duplicating the planes pitch and roll fully will only result in making the occupents feel sick...!It seems that most of the big buck siumlators do exactly as you suggest, apply a sharp direction movement that then washes out rapidly (returning the motion platform back to centre), so that an opposite movement can be made when the plane attitude changes.The actual movement of the motion platform can be as little as 10 degrees in any direction to accomplish this succesfully.The hard bit is monitoring the planes attitude and velocities on all 3 planes, and then calculating what 'motion cue' is required before moving the platform.More research is needed I think.

[tigert 0]

>The hard bit is monitoring the planes attitude and velocities>on all 3 planes, and then calculating what 'motion cue' is>required before moving the platform.>>More research is needed I think.Yeah.I guess it would need a dedicated computer for this.My own setup is more like a basic IFR/VFR trainer which will soon have twin prop quadrant and basic instrument panel + avionics (radios etc) as switches. If I ever decide to go full enclosure, I think the "bass shakers" in certain places, good visual (a collimated mirror, and probably a fresnel too gives a LOT of depth, this adds realism too, a video projector sure gives a big image but it probably isnt as good in that sense. Focusing between the "outside world" and your instrument panel is one of the things that feels "out there".I bet that plus a few good bass shakers and a nice audio system could do a lot. A full scale motion platform is something that would definitely rule, but all these others can also work wonders. And without them, the motion is kind of wasted too. It all needs to be balanced I think.Also, small things seem to fool the brain - how about small air vent holes that blow cooler air at your face, it could even make the cockpit colder the higher you go :-) It might all sound silly, but I think these small things can do more than one first thinks.Best, and good luck ;-)OH btw, there was this australian guy who was working on a motion platform, they have a 747-400 sim for "World Flight Australia", maybe you can contact him and discuss stuff, he probably has been thinking about the same things. The worldflight address is http://www.worldflight.ozflight.com.au/ .Tuomas

[Guest]

Thanks for this Spencer,Yes I have now checked out FSUIPC, and through the use of a an EPIC card along with the gauge driver card, it is possible. This is now the route I am taking.many thanks for thisTJ

[Guest]

HiExcuse the delay in joining this discussion as I'm in the middle of an overseas move. (just arrived yesterday, presently living out of a suitcase at my new address in Holland, awaiting the arrive of the sea container)I've posted some info on motion drive and stuff before on this forum, that may be of interest. (a search for items posted under my name should turn them up)Motion drive software should actually be quite fast, to make the nice turbulence cues felt. X-plane is one of the platforms that has more build-in support for motion drive parameter output, and they deliver 20x - 50x per second updates. You can download the demo and see in realtime what various parameters do. There is also some freeware software available at http://www.x-plane.info/ that you can run on a networked 2nd PC to extract the parameters. Indeed the hardest part of the motion system is the software manipulation to make the motion "feel real". Parameter mixing and differentiation, wash-out (return to zero), max motion range limits etc. requires lots of work on the software side. I had some discussions with people who were experimenting with motion via X-plane, and it's definitely on university mathematics level. I'm nowhere near that point, so can't help you with that yet.For the PC to hardware interface there are commercial packages available. Even Dirk's FSbus has D/A outputs for driving servos, that could be useful. PIC's are indeed great IC's, so if you are up to programming those types, you could make your own D/A with those. I'm persuing the I2C bus approach with VB6 that should be able to give me 8 fast analogue outputs with simple connection via the parallel port and an I2C bus D/A IC like the TDA8444. Unfortunately, the overseas move has put a big delay on the whole project, so it will be a while before I can continue. I did some simple experiments with a totally different interface: A combination of signals from yoke movement, mixed with signals from force feedback, rudder pedal, brakes and throttle movement. Each with some differentiation networks and gain adjustments. It was quite some fun, but in most cases still quite fake. The rotatation and climb was quite nice though, with vibration and back pressure quite realistic. Only pitch and roll motion (my setup) is extremely limited. Heave, longitudal and sideways motion are equally important, but hard to achieve in a homebuild setup. Fully enclosed is a must. Delays & errors in motion wrt the screen will make you slightly sick. My Taiwan e-mail address is not really useful anymore. I suggest we keep the discussions in the forum. I can't post any pictures or schematics now, as most info is in the seacontainer (expected to arrive within a month)Hope this is of help. Roland