How do I transfer the roll and pitch movements to a remote location?

[2raj]

Hi folks,

I am impressed by your guys in the forum, so I am sure you'll be able make some suggestions to help me out.

I need to transfer the roll and pitch movements of my RC plane to a remote location.

I was thinking of using two gyros, a 2 ch radio and two servos. Here is how...

At the remote location I have the receiver and the servos set up to receive and reproduce the roll and pitch movements.

In the plane I attach the two gyros to the pitch and roll axis. I then take a 2 channel transmitter (obviously one on a different channel and even different frequency to the plane radio) and strip it to the board only.

Now here comes the tricky part (hence why I am calling for your help)... Somehow I connect the two gyros to the transmitters the aileron and elevator controls.

What do you guys think? How do I do this? Or can you think of an easier way to do this?

Thanks heap in advance,
2raj.

[tve]

Hi folks,

What do you guys think? How do I do this? Or can you think of an easier way to do this?

Thanks heap in advance,
2raj.

Hi 2raj,

I'm not sure exactly what you are doing but rate gyros might not be what you are looking for. Are you wanting to mirror the pitch and roll, or do you actually want the rotational rate? If you want to mirror the pitch and roll, some tilt sensors like these might be better.
http://www.memsic.com/memsic/data/products/MXD2125GL/mxd2125gl_hl_nl_ml.pdf

They are commonly available in easy-to-use form like this

http://www.parallax.com/detail.asp?product_id=28017

regards,
tve

[2raj]

Thanks a lot tve! Yes, I want to mirror the potch and roll. I had a quick read and I think Memsic 2125 will do the trick.

Tell me something though, this chip is outputting digital PWM. This is the same that is fed to servos and what is transmitted through RC radios, right?

Doesn't the RC transmitter sticks produce PWM signal that is fed to the RF circuit?

In that case, I should be able to take the Memsic 2125 output (may be with a bit of translation/adjustment) and feed it to the RF circuit, effectively replacing the control sticks with the chip?

I like the idea of using an existing RC radio. Otherwise, I'll have to use a data modem and go all digital (sound expensive).

Regards,
2raj.

[tve]

Hi 2raj,

You will probably have to do some kind of conversion to modulate your transmitter. I don't know for sure, but I would assume that the potentiometers in the transmitter are used to create a variable voltage that determines the pwm modulation. You will probably have to convert the pwm output from the memsic ic to a matching variable voltage.

This is one other thing. You may have noticed when reading the specs, but there are 2 modes for tilt sensing. If you use 1 sensor mounted flat, you can detect about +/- 60 degrees on either axis. Is this enough for your application? If not, you can mount 2 sensors perpendicular to each other and vertical to the surface. Then you can take the arctan(x/y) to give you 360 degrees on both axis. I plan to use this method for something I'm working on. You might need a microcontroller for this. I think the first mode could probably be made to work without a controller.

Regards,
tve

[Mr.RC-CAM]

Inertia sensors like the Memsic devices will not be a simple implementation on anything that moves around like a model aircraft or car.

These sensors cannot differentiate between static G-Forces (gravity) and dynamic G-Forces (inertia). In a model R/C application the model moves a lot (inertial force), so the static forces that need to be measured for the pitch and roll data will be gravely affected by these inertial forces. There is some discussion about this issue at the RC-CAM forum: Artificial Horizon Project (http://www.rc-cam.com/forum/index.php?showtopic=294)

BTW, the PWM output of the Mesmic has nothing in common with the R/C signal format. You can either use a microcontroller to measure its pulse width or you can intergrate and filter it for use by an A/D. The data sheet has all the info about either method. I used the latter method and it worked fine in head sensing application: HeadTrack-R Project (http://groups.msn.com/rccam/general.msnw?action=get_message&mview=1&ID_Message=6106)

RC-CAM

[tve]

Inertia sensors like the Memsic devices will not be a simple implementation on anything that moves around like a model aircraft or car.

RC-CAM

Mr. RC-CAM,

These are good points. I found in my application, that measuring the pwm output from the memsic device worked well. I also found that integration was not necessary. It seems that the device is fairly stable even with vibration. I also ended up needing to use a rate gryo in my application to account for rotational acceleration that the memsic device was just too slow to respond to. I think it probably is because of the viscosity of the fluid in the device (just a guess)

2raj,
What exactly are you trying to do? If you need to fly two models simulaneously (mirroring each other), you will need something more complex like a combination of the memsic device, rate gryos, and a controller with a PID loop. The memsic device could drive the proportional term and the gryos drive the derivative term.


edited:

I was thinking some more about this, and realized that mirroring two flying models is a very complicated thing to do. You would probably also need gps, and digital communications, and some pretty fancy firmware to sync them up in addition to the accelerometers and gyros. Sounds like fun!

[2raj]

Wow! You guys are legends! You have been most helpful.

I think I'll just come clean and explain what I want to do. It is no secret. Someone else probably have already done it.

Ever flown a model plane and wondered how it would have felt like sitting in the cockpit and experiencing the flight (well may be not when I am flying) :-) You guys probably know the feeling while watching the video footage of your flights.

What I want to do is to build a moving chair simulator driven by the roll and pitch movements of a flying RC plane, supported by the video and audio form the cockpit.

There are some creative examples of moving chair simulators that the Microsoft Flight Simulator nuts have come up with. My research of these guys tells me for a decent experience I only have to provide about 20 degrees of movement in the roll and pitch axis, as long as I provide audio and video form the plane's cockpit.

For audio and video I am using a HMD and I want to mount 3 cameras and use a head tracking system to switch between cameras for a more realistic view from the cockpit. But for the starters one camera will do. (Thanks Mr.RC-CAM for the pointer to your head tracking project.) That will do the trick very nicely.

BTW, I have found a $200 source of brand new HMDs on eBay. I'll receive mine a week or two so I can tell you how good or poor it is. If anyone interested let me know.

Also thank you for the pointer to the Artificial Horizon Project. Great idea to use the Co-Pilot's sensor. I just need to take the output form the sensor and transmit it my chair to drive a couple of servos.

For moving chair I am using an air compressor and two pneumatic cylinders with some solenoids driven values. I have seen it work well for Flight Sim. Have a look at these two videos:

http://www.jollyeskimo.cjb.net/spacechair.avi
http://www.jollyeskimo.cjb.net/dadonchair.avi

I have the parts for the chair ordered. I need start practicing my welding skills.

I want to mount it all on a trailer and take it to the club for the ultimate RC flying experience. It would be a great for shows.

There you go, that is my project.

So what do you guys think?

Is there enough interest here to create a thread on this? I can sure use all the help I can get with the circuit designs. :-)

Cheers,
Tooraj.

[Mr.RC-CAM]

Yup, for your virtual reality application the Co-Pilot is an ideal solution for measuring pitch and roll attitude.

RC-CAM