Howdy from a new forum member!

Paul Frymire

Hi everyone,

My name is Paul Frymire and I just recently discovered the Parallax world and am very intrigued by what I've seen people do!· I am a complete novice when it comes to electronics though I do have coding experience, and was hoping to get some advice as to the best route to take to reach my eventual goal.

Basically I love to fly radio control helicopters.· On a whim I recently purchased a coaxial electric helicopter (two counter-rotating propellers) and was amazed at how stable of a platform it represented compared to a traditional helicopter.· Because of the inherint stability of the counter rotating blades and the lack of a tail, I started thinking it might make a decent platform to explore simplistic vehicle·automation.

I've spent a few days reading the forums and trying to best familiarize myself with the Parallax technology and have a few thoughts that I'd love to get feedback on.

From what I've read, the·more significant·tasks will be keeping the helicopter's·position stable with minimal left/right/forward/back movement.· Because of this, my first objective would be to gain an understanding of how to implement a two dimensional Accelerometer and understand the relation between it's feedback and the controls of a helicopter.· My thought was that the easiest first step would be to implement an Accelerometer that controls the forward/back and left/right movement of the helicopter, but leave the throttle/altitude control in my hands for initial safety and control.· This would involve relinquishing control of two servos.· My objective would be obtaining a safe hover.

What is everyone's thought on this as a first goal?·

Also, what is everyone's thought on a first kit to start out with, with this goal in mind?· Like I said, I'm new to this area, but not new to code, so I would like something that will challenge me and hopefully carry over equipment wise to my project.· I thought about getting the new USB Boe Bot kit along with an Accelerometer for future use.

Thanks and nice to meet everyone!
Paul Frymire


·

latigerlilly

To hover at a predictible low altitude, I suggest getting the sonar kit: www.parallax.com/detail.asp?product_id=28015 . If you want to fly higher than a couple of feet, then an alimeter will be needed: www.geocities.com/jpadfield13/altimeter.html .

Paul Frymire

Becaue I'm completely new to this, I am not going to try to control the altitude of the hover at this point.· As I mentioned, my initial goal is to control only the pitch and angle of the helicopter through an Accelerometer controlling the servo's that handle that, while I stay in control of the altitude/throttle for safety.· Basically my test scenario would be to increase the throttle and taking off with the helicopter staying in a stable hands off position in regards to movement.· After I have achieved this, I would then move on to other areas.

Larry

I don't have a helicopter, but from what I have read, there are rate gyros that will do what you want, that is, maintain any orientation until commanded otherwise,

Very simply, the control signal goes to the rate gyro, and it changes the signal to the servo depending how the helicopter has moved.

Does your heli have rate gyros?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

Paul Frymire

Actually my heli does not have any gyro, which is what made me think about this project.· The helicopter uses two counter-rotating propellers on a single shaft that counteract each other's torque removing the necessity for·a tail rotor and gyro and as a side effect give it a decent lifting ability and great stability.

Larry

Interesting problem.

There was an autonimous flying vehicle contest a few years back that proved to be a pretty tough nut to crack. It involved flying over a fence into an square, picking up an object, and flying back out. It took several years for university teams to crack.

The problem, of course, is that there aren't good sensors for guaging position in the air. Airplanes use beacons,GPS,and inertial guidance.

There are IR beacons, (Pololu comes to mind but they probably won't work well outside)

You could try sonar, but that may not work well under a noisy heli.

Accelerometers may work, but I suspect they won't resolve well enough and won't counter a breeze. In addition the vibration of the heli may be hard to counter- accelerometers are often used as vibration sensors. at a minimum, you will have to adjust the bandwidth slow enough or use a low pass filter.

Gyros are better at holding attitude, not position, which I guess is really what you want. In addition they are subject to drift.

That leaves vision. A really good vision system would work fine, measuring "optical flow" see if the heli was going up, down, or sideways. What's optical flow? Think of the last sequence in 2001-a Space Odessey.
The problem is that there aren't good vision systems for hobbiests-yet (I'm waiting to latch on o a PROPcam)

The CMUcam will sense color, but not flow - You could put a target on the ground and have the heli hover over it.
The CMUcam2 will sense movement, but I don't know if it does it well enough.
I recently saw a French product called the BOP-eye - I'm not sure of its capabilities.
www.pob-technology.com/pob/crbst_0.html



How about you tie a string to it?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


Post Edited (Larry) : 2/1/2007 1:22:00 AM GMT

Paul Frymire

Though it would be far from precise, could you use the Parallax GPS Receiver module to calculate the velocity in a given direction to them keep the helicopter in a box of sorts?· It claims to be accurate within .1 meter/second.· A combination of this data along with direction of movement might allow you to keep your helicopter in a virtual box of maybe 1 meter by one meter?· Along with this functionality, it could be used in conjunction with the sonar to handle altitude.· Once the helicopter has exited the threshold of the sonar, the GPS altitude would take over.

Kevin Wood

What kind of programming experience do you have?

Paul Frymire

I'm an engineer for Microsoft so I work with C# and to a lesser degree C++ on a daily basis, and I have VERY minor experience working with Pascal, Basic and Java·from my college days.

Kevin Wood

If you have that kind of experience, you might also want to consider the SX28/48 Protoboards. They would be a little more difficult to work with, since you would need to work at the junction of SX/B Basic & ASM, but if you can handle C++, you won't have any problems.

The downside of starting with the SX is that it's a bit more complicated to get going. The upside is that you get access to the full capability of the hardware. As for the electronics portion, the circuits that you would build would be mostly identical. Although there are less examples available for the SX, most of the BS2 schematics and code can be adapted to the SX without much trouble.

Here is an SX-based project similar to your goals:

www.parallax.com/sx/contest/contest_airplane_autopilot.asp

This isn't to say avoid the BS2 or Boe-Bot, just that you might have more options than other non-programming beginners.

Paul Frymire

My biggest concern is my lack of any real experience with embedded design. I was looking at the Boe-Bot for Microsoft Robot Studio kit that Parallax offers. Though I'm not sure how expandable it is, I think it could not only provide me the initial experience working with the Boe-Bot, but also be adaptable to a little of my helicopter ideas. It would be interesting to control and tweak the helicopters movement via the bluetooth interface and Microsoft Robot Studio.

Kevin Wood

In that case, go with the Robotics Studio Boe-Bot kit. It will get you up to speed more quickly than the SX route would.

zener

Hi,

I am beginning to work on a simular project.· I intend to use a regular electric Helicopter as a platform.· I intend to use the ping for low level accuracy and the GPS module for anything above ~ 6 ft?· Haven't decided yet as I need to do some experimenting.·

The GPS module will be very handy in location control and speed.

I expect I will need to use a compass module to keep the Heli pointed in the right direction.

I will be looking forward to reading about your efforts.· I will keep mine posted too.

Bruce Bates

zener -

I don't mean to stop your project before you start, but I rather doubt you'll be able to hold the helicopter in a perfectly steady 3-axis hover, which is the only way you will be able to use the PING))) module effectively and accurately.

You also may want to check the positional accuracy and repeatability of any GPS module you might intend to use, before purchasing it, so there is no surprise later after it's purchased.

Regards,

Bruce Bates

Post Edited (Bruce Bates) : 2/27/2007 5:04:03 PM GMT

zener

I will start with the Parallax GPS Receiver first so that I can experiment a little without having to invest a great deal in hardware I don't need. Weight is also an issue when I am selecting hardware.

I have seen some examples of autonomous Helis on the web from some Universities that seem quite stable. I intent on building a "simulation stand" to help reduce the damage from crashing. Hopefully, by the time I get the algorythms perfected, I still have a helicopter left.

I will start with experimenting with the stability of my sensors under extreme vibration. That might slow me down a little. I need to wait till summer too as Alaska is not the best location for winter outdoor projects. I can do some testing indoors.

Thanks for your input. I will keep my progress posted here.

Robocop

Paul,

I also like to fly rc helicopters and have a uav heli project as a goal. I currently have a T-Rex 450x heli (and an X-ufo) which I plan to use for my experiment. I also plan to use the basic stamp for electronic control and programming. Currently i'm learning with the Boe-Bot (which I've had for about 2 weeks and i'm totally pleased with) and plan to experiment with sesnors that can contribute to my goal (Parallax GPS Receiver Module, Hitachi H48C Tri-Axis Accelerometer, PING)))™ Ultrasonic Sensor, etc).
About your first goal you might get some ideas from this self balancing robot project -->(http://www.parallax.com/html_pages/resources/custapps/app_bway.asp)

Josh