BS2 and gyro- heading hold

iamscottym

Hi everyone, I'm working on a robotics competition at Northwestern University.· The robot has to get from one 3x3x5' platform to another, with a 3.5' gap between, and then go down a ramp and knock over a plastic cup 15' away.· I'm using compressed air in 2 2L bottles for power, its basically a rocket powered plane.· Anyway, there's a stipulation that I have to touch the second platform, so I've just been bouncing off at a low angle of incidence and going on my way.· The problem is its not 100% accurate.· Rather than try and mechanically fix the problem, I figured a heading hold gyro like on a helicopter would work great.·

I was looking to use the I would like to use the GY401 Gyro with a S9254 Digital Servo to control the rudder.· However, all the gyros I've seen are for RC helis, and I'm not sure if they will function without a radio.· From what I've found on the futaba website the gyro considers neutral whatever position its in the first 3 seconds its on.· Logically, no signal would only cause a problem if it established a no signal neutral, whatever it may be, and then got a signal from the receiver.· I'm thinking that I could sent the appropriate signal from the BS2 I'm already using, since its just an analog servo position signal, which the gyro converts to digital for the servo.·However, there are other functions that are controlled by the receiver(transmitter), including turn heading hold·(AVCS) on and off, which I need to do, and am not sure how to do.· I have emailed futaba about this as well.· Also, I dont want to buy a radio just to put a scope on the·signal wire and figure out the signals...I've spent·enough money already on this·robot.

If anyone has any·other recommendations on gyros, etc, I'm all ears.··The only thing I have right now is the BS2 and servos and such for the rest of the robot.


Thanks,

iamscottym

Paul Baker

I dont have any experience with gyroscopes, so I can't answer your question about RC helicopter gyros. But Analog Device's ADXRS series of gyroscopes do not have such requirements. Sparkfun sells break out boards for the series here: http://www.sparkfun.com/commerce/product_info.php?products_id=393

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
1+1=10

iamscottym

Ya, it seems a lot of people use that gyro for robotics.· Some guys I know here tried to use it on a Boe Bot, with some success.· However, it is probably at the edge of my ability to get it working as well as I need it to, and even then, I have yet to see it applied somewhere requiring the high precision of my robot, vs a robot driving around.· I think I'm just going to get it and hope mode change is achieved by a jumper like it is on some others I've seen.
·

LoopyByteloose

That bounce might be a significant problem with any navigation device.

Generally you have to keep the vehicle unmoved while the gyro intialized, after that it will know where it is.·· But you need to find how many Gs of a jolt the device can handle or ignore.· It is possible that little bounce might pull 50Gs or more.

I don't think you need radio unless you want a remote control link.· Still, you need some way for it to nagivate and recognize its target.

I suppose you are just programing it for one flight path and hope the plane is bigger than the margin of error.


▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)

······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan

Post Edited (Kramer) : 4/11/2006 8:12:24 AM GMT

iamscottym

The gyro will initialize for 3 seconds, then it will take off. It has to unfold its wings and drop its angle of attack anyway(it has to fit in a 1 cu ft box at start, so its inclined 45 degrees and the wings fold down and then the whole wing assembly pivots up).

I know I dont need the radio, but I'm not sure how to get the gyro in the heading hold mode without it. The bounce is not very significant at all. I would have to check the math, but at a 10 degree trajacetory, the robot will have only have a few inches of potential energy in the z axis.

I dont really need to navigate, I just adjust the pressure and trajectory. I've gotten the robot to work without the gyro or any in flight adjustments at all, its just that it doesnt always work. Sometimes it takes an odd bounce and goes off in the wrong direction. I will need to make the robot correct its heading fast enough so that it doesnt just take off on a parallel path and still miss the target, but I think I can do it with just rudder. If not, I'll take a page from the hei pilots and use an esc and tail rotor as well(which the gyro can do).

I thought about actively seeking the target, but I'm pretty new to basic stamp (I've been coding forever, but I havent worked with PING))) or anything like that yet, and I dont really have time to learn now).

Paul Baker

Scotty, Yeah the device seems to be not the easiest in the world to interface with, the output scale is an average of 15 mV/°/s. A 10g random vibration in the lateral axis will produce a variation between 2.47 and 2.49 V when sampled at 1/60 Second sample rate (I am assuming you will be measuring the Yaw of the plane and the shock will occur in the Pitch (longitutinal) direction). This corresponds to a potential skew of the reading around a degree or so (though this may be well within your tolerance). Since the variation in voltage is small, you would have to employ an op amp which has a low input offset voltage. One benefit of the package is that it provides a precision 2.5V reference which can be fed into the + input of the op amp.


I got all the specs from this file: http://www.analog.com/UploadedFiles/Data_Sheets/279107307ADXRS401_0.pdf·(thier low cost version).

The higher spin rate ADXRS150 provides additional test data more applicable to your situation: http://www.analog.com/UploadedFiles/Data_Sheets/778386516ADXRS150_B.pdf#page=8

Shows a Shock test in the longitudinal axis of 100g for 5ms, where the output starts to stabilize after 50ms and returns to near normal after 100ms.

Regardless I think using any gyro will likely require a bit of experimentation to produce a reliable method of use. Using a R/C helicopter gyro may not be any easier since they may not be well characterized for shock conditions. The benifit I do see from using a RC gyro is that it will likely be easier to integrate into your system. In theory if you mimic the RC reciever by·sending a servo centering pulse (1.5ms pulse in 20ms window) the RC gyro should autocorrect the bearing of the servo. Though unless you monitor the output to the servo, you wont get any info in order to determine thrust.

Whichever way you go, good luck.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
1+1=10

Washer Medic

would the Hitachi H48C Tri-Axis Accelerometer Module due the same thing as a gyro???

····· Brian

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
There's nothing a good wack with a hammer won't fix

Darn I let the white smoke out again

Paul Baker

No, because it only measures to 3g, an accelerometer would have to be able to measure the entirety of the shock force in order be sure a measure of rotation would be accurate. Also there is a difference of option on how well an accelerometer could measure rotation, theres alot of factors involved such as the pivot point, the distance of the accelerometer from the pivot point, and if the pivot point is variable, it leaves the calculations unsolvable due to too many unknowns. Though if one were use one in the tail and one in the nose, the pivot point could be calculated, but the shock force is still a potential problem.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
1+1=10

Washer Medic

Hi Paul thank's for the explanation
My 401's take a lot of punishment in my·RC helicopters

Brian

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
There's nothing a good wack with a hammer won't fix

Darn I let the white smoke out again

Post Edited (Washer Medic) : 4/13/2006 3:34:32 AM GMT