My first 3 axis flight simulator project

2DOF should do for a fighter jet sim. IMHO the third (vertical) axis might be nice to have. Since turn rate of a fighter rarely extends 20°/s the sensation is much less than that of a loop or a roll. Roll rate can be about 360°/s, so this axis has to be very agil. Pitching into a loop is similar to the turn rate, that means the rotary motion itself isn't important. It is the possibility of simulating forces in different directions that calls for a pitch axis. The pitch _motion_ again is nice to have.

So in my opinion a 2DOF sim is suitable for a fighter platform. What do you think?

I do agree that it would be suitable but don't you thing going right over the top would add the icing to cake.My thought is if you were in a dive the simulator would actuality go down at first than slowly pull up although the cockpit view would show the plane going down in the game to simulate the force of the dive.Than as you pull up to level it could be rolled backwards to hold you in the seat.Than in a sharp bank to right the sim would bank sharp to right than slowly go to left while cockpit view show you still banking right.Have a piggyback program to adjust all the values and controllers for the motors.Little tweaks of roll and yaw one way or the other and you well on your way to a masterpiece.
I played fighter-ace for 12 years and now play Aces High,love the old birds!

Go Erik go!


posting.php?mode=reply&f=22&t=2781&sid=b7854b94182c9e40ffe79b94c4294177#

To be shure we talk about the same things: have a look at that video:
http://www.youtube.com/watch?v=yGUcoUUYAx0

We see a 2DOF sim platform which has a roll axis and a pitch axis. And that is all you need, in my humble opinion. If you want to go over the top, you can actually do that by rotating around the pitch axis.

However, your considerations about the banking and going back slowly to 0° (washing out) are absolutely right.

All I want to say is: you don't need a third axis, namely the vertical axis, also called the yaw axis. This makes the design much easier.

Why not build a platform like that on the video? It suits your needs, doesn't it?

Here are some sketches that will help us speak about the same things:

I understand 3 vectors there is another issue called gimbals lock to deal with.
http://www.youtube.com/watch?v=rrUCBOlJdt4&feature=fvw
I will overcome this one way or another.

Well after much consideration I think I will go with this motor system http://www.youtube.com/watch?v=qGQl0c_G ... re=related and controllers.(also work well to sister the slip ring assembly)This will allow me minimal clearance between rings to keep the total height down.The cockpit fame is still getting welded up now but will have pics soon.The next step now I final settled on drive system will be the final circuit requirements for slip ring.I'm still pondering a wireless motor controller to minimize ring requirements on the first prototype.But I will use 3 usb circuits up to cockpit for peripherals and(HUB)for extras,and a 15 pin system for monitor(may add extra 4lines).I'm exited its actually started although I gave myself a year I think I can do it sooner.I have many other plans for this that's why I will plan to spend a lot to get it working right.So wish me luck!

Of course, I wish you all the luck. And it seems to me, you have all the plans already drawn? Or are there still open issues? I'd really like to know which construction you prefer.

You spoke about the gimbal lock problem, indeed a serious problem with gimbal constructions. That's one more reason why I prefer the 2DOF system instead of a 3DOF one. Yes, you got another problem then, you can't simulate flying the curve at all, which you could with a 3DOF system. But, what I said earlier in this thread, the sensation of turning in a curve isn't that much, since the rate of turn is at a maximum of 20°/s. And that is when you pull 9g's! Because the 9g (or even a 3g) experience delivers _much_ more - I repeat MUCH more - sensation then the turn rate of 20°/s you should have a look at simulating the g-forces first. Now that you can't simulate more than 1g (at least with the constructions we speak about) should we rack our brain over the missing turn capability (z axis)? I think: no. Important is the possibility to pitch to give you a feeling of flying over the top. That indeed puts the icing to the cake And even more important is the rolling capability, since roll rate is and exceeds 360°/s - just try to simulate that on your platform - that will give you a real kick... And forget about the z-axis of your platform. My suggestion.

By the way, imho the following link will explain the gimbal lock even better:
http://www.youtube.com/watch?v=zc8b2Jo7mno

I don't like models that don't show the axes but only turning wheels. So, to get a better understanding, I've quickly built a sim platform - with copper wire. Have a look at it, in my model you can recognize the axes. And: the wheels you see in my model don't compare to the wheels in the videos. The latter use the wheels to show a given axis rolling within. My wheels are mounting frames instead.

In this picture I added the axes' names. Note, that the axes shown are not the same as the plane's axes. The plane has it's own co-ordinate system, as well as the platform. When dealing with robots the systems have to be transformed (finding the equations is somewhat tricky). But that's stuff for a later issue.

I decided the inner frame to be the roll frame (y axis). The reason is as simple as important: the inner frame has to perform the highest angle rate and so must be very agile. That means, it will be highly accelerated and therefore has to be very light. If we make, say, the x-axis the inner frame (coupled immediately to the cockpit) the roll frame has to roll the cockpit _plus_ all the material of the x-frame (could be 50 to 100 kg more!).

...so, GrimReaper, when buying (expensive) motors think of their performance: the maximum revolution speed should not be too high, since torque decreases with increasing revolution speed. And you'll need torque!

I just had a look att the IntelliDrives motors - they seem to have a lot of torque, up to 110Nm - that's quite enough even for the roll axis...

What's the price for such a drive? 600$?

ROFLMAO I made my first model Exactly like that with copper wire.
:

I want to be able to use this simulator in any game.You could play Aces High drive tanks run turret (Yaw axis)'drive a jeep run AA guns ,fly jets, do complete flips in car racing games.If you spend a little more add the extra(none stop rotation) axis and give it the lowest profile as I have (73-75high depending on my final stand),it gives you so many more variables to work with to add realism.
My footprint will be( 7'6*7'',or 7'1''*7') total, with optional 1 side wall mount and stand.So you will need a large room he,hee!That only gives you 2'' clearance to floor.I thought you could use Lazar beams around all 4 sided ( an emergency stop system) with old garage door units,you know,encase the cat comes in !The motors with break very fast in mill a seconds.Even with the fastest of cats 2'of clearance should be plenty of room between the simulator frame and the floor and it's esophagus to draw a life sustaining amount of oxygen till you get out. :mug1: I want it to look good too so I'm going to take time on frame construction get the shape I'm looking for.Still waiting on price of motors,Ill wait till the frames finished welded ,mount the seat and monitor get inside it . Than I'll figure out exactly what i want in side for everything(Lights,sound,sticks,comms, ect)I think it will be in my best interest.It will have a slide forward cockpit opening with a latch(emergency stop)system not automatic just sliding hatch.I wont be able to do a final circuit count until I'm sure of what i need.

Sound very interesting! I think you are on the right track, and I will be interested on how you solve gimbal lock (if possible).
Interfacing all this to the computer and controlling properly... that will be an other chapter.

Sorry for missing your question before, I'm far-far away from you, Niagara area. When it's all done I will definitely make the trip to try it though! :

Ya that would be a blast.When its all together and working I'll send ya a personal invite.I like that area its nice only 5 hours away.

WELL THIS IS PROTOTYPE!Mini-Me

Instead of reducing the number of DOF (by following my suggestion) there can be another solution. You can now call me a show-off, but it was my first approach to the gimbal lock 8) . I gave it up since it's too complicated to be handled in a home made sim platform. At least for my requirements. But a man with your ambitions...

Have a look at this very interesting text, found at
http://www.hq.nasa.gov/office/pao/Histo ... mbals.html

See the picture
http://www.hq.nasa.gov/office/pao/Histo ... lm_imu.gif

I quote:
==================
The following is taken from an April 1963 MIT Instrumentation Laboratory Document, E-1344, 'Apollo Guidance and Navigation: Consideration of Apollo IMU Gimbal Lock' by David Hoag. The difficulties near gimbal lock can be avoided by the addition of a fourth gimbal to the IMU. This will be called here the redundant gimbal since it provides more degrees of freedom than theoretically necessary. This redundant gimbal will be considered in this memo to be mounted outside the normal outer gimbal. The order used in this description is then: inner, middle, outer, and redundant. The most likely operation would use the inner three gimbals to drive the stabilizing gyro error signals to zero while the fourth if driven so as to keep the middle gimbal near zero and away from the gimbal lock orientation. This can be done by generating a redundant gimbal rate command by expressions similar to

A(redundant) = k sin A(middle) / cos A(outer)

so that a negative feedback on middle angle occurs to drive middle angle towards zero. It should be possible to make the inner three gimbals have the same dynamic performance as the simpler three-degree-of-freedom system.
==================

Maybe, this could be your solution?

The typical case of solving a problem by creating an other one.

This is getting more and more complicated...

I would stick with 2 axis for starters, then see how it works out.

I like to think so.

Grimreaper has got quiet - I imagine two possible reasons:
- the challenging Complexity of Physics has knocked him off
- the annoying Prophesy of Doom has knocked him off
What will it be?
Or has he just been out for a ride on a swiftly built platform and doesnt find the emergency switch?
Ok, Grimreaper, no offense!
:brows: