RunRyder RC
 5  Topic Subscribe
WATCH
 1 page 841 views
Helicopter
e-
Elec-Helicopters New or Limited Activity
MS Composit › Making a micro better in windy conditions!
05-31-2004 04:37 PM  13 years agoPost 1
mattijs

rrApprentice

Keerbergen Belgium

My Posts: All  Forum  Topic

I started a thread on the zoom forum asking the zoom guys if somebody would want to try putting a gyro mounted sideways on the collective servo. I wonder if it would enhance a micro helis ability to fly stable in windy conditions. My hornet goes up and down when it is hit by a gust of wind, i figure the zoom does the same, prabably less though because of the increased weight but still. It might be significant enough for one of the zoom guys to try it.

here's the thread: http://www.runryder.com/helicopter/t109918p1/

For those who don't know, the zoom doesn't have the same head structure as the hornet and features two servo's for cyclic and one servo for collective. They can easily just put a (HH maybe even) GYRO between the RX and the collective servo.

If it works good on the zoom i might make a mixer for my hornet with a SX microcontroller as this is the first cool project i could think of other than make a led controller for night flying but that is just a waste of the µcontroller so....

What do you guys think???

Good idea to try to stabilize collective with a gyro or do you think or know why it won't work (good) ?

PM  EMAIL  HOMEPAGE  Attn:RR  Quote
05-31-2004 06:07 PM  13 years agoPost 2
tabbytabb

rrElite Veteran

seattle

My Posts: All  Forum  Topic

I am sure you could put a gyro on every axis, but then you wouldnt be flying anymore, what fun is that ??

You have to look at flying a micro in the wind as an expensive adventure

Tabb

PM  EMAIL  HOMEPAGE  GALLERY  Attn:RR  Quote
05-31-2004 06:10 PM  13 years agoPost 3
mattijs

rrApprentice

Keerbergen Belgium

My Posts: All  Forum  Topic

Over in the ZOOM thread someone gave a solid piece of information proving it won't work. GYRO's sense rotation , not linear motion.

I responded by saying it is possible to translate going down and up into rotation if you make a tilting tray or something balanced by springs (altough that might make the PID stage in the GYRO confused and has one distinct resonance frequency )

Another possibilty is to make a sort of gyro yourself. I am looking for the right sensors and don't think it would be that hard to design something that works.

I don't know if this will work but i figure that a gyro gets its rotational information through monitoring TWO linear motion sensors, if they 'feel' the same accelaration the heli is starting to translate, if the measured accelarations differ the heli is starting to rotate.
This however is just a theory on my part as i don't really know how real gyro 's work.

If my theory is right finding the signal on the PCB from one of the two motion sensors and disabling it will make a linear instead of a rotational sensitive gyro.

In the end it doesn't really mattar what device to use i just wanted to know if it'd make sense trying to electronicly stabilize the up and down motion on µheli's in windy conditions.

Let me know what you think

PM  EMAIL  HOMEPAGE  Attn:RR  Quote
05-31-2004 06:19 PM  13 years agoPost 4
mattijs

rrApprentice

Keerbergen Belgium

My Posts: All  Forum  Topic

Hi tabb,
i have seen someone over on a forum say the same thing once...might have been you. Well as much as i too like flying when it's a bit tricky just too practice them reflexes I see this as a cool project more than a necasity (is that a word???). + You must admit that you can't forsee those gusts and when you are going fast , suddenly going up or down doesn't really look very heli like. You can't do anything about it as it is random and if you think about it....why not? We're all trying to make it as stable as possible without restraining the cyclic (weighing the blades, BELL HILLER mod etc...) and if you listen too what you are saying the pilots who flew heli's 30 years ago should all be laughing because nowadays people use throttle curves, HH gyro's and blablabla this blabla that.

Wich doesn't mean i am trying to say that it is the next big thing like HH once was but atleast the idea to stabilize a heli's up and down movement is worth a shot imho. Especially on micro's.

The up and down movement too me is the only really annoying consequence of windy weather, cyclic doesn't really react too it that much.

PM  EMAIL  HOMEPAGE  Attn:RR  Quote
06-01-2004 07:00 AM  13 years agoPost 5
cyhyam

rrApprentice

So. California

My Posts: All  Forum  Topic

The only way to sense and control the altitude like you want is with a radar altimeter, and that is too heavy for a model.

GPS could hold you to within a couple of feet, but it would also be heavy.

You might be able to use a pressure sensor. I have heard of them being used on model gliders wing tips to keep them level. They can detect an altitude change of a few feet.

A tilting tray to translate the vertical motion to a rotation that the gyro can sense will also be subject to forces from other flight attitudes, banked turns, etc.

Think it through and you will see it has flaws. We will call it a horizontal pendulum for lack of a better term. It will angle up as the heli starts to fall, this would give positive pitch (through the gyro sensor) to arrest the descent. As the heli stops descending the pendulum will now angle down causing the opposite input…………….negative pitch….ooops now there is a problem, …..better yet a broken heli.


Walt

PM  EMAIL  HOMEPAGE  GALLERY  Attn:RR  Quote
06-01-2004 07:53 AM  13 years agoPost 6
donlynn

rrKey Veteran

New Zealand

My Posts: All  Forum  Topic

Gidday

What about car ultrasonic reversing collision avoidence buzzer, sold here in NZ for ~us$50 made in Asia
I did a little enquiring about 3 years ago to see if it was viable, apparently it might work, The closer to the ground the more accurate, the better the reflection (surface texture ) the better the accuracy ,long grass is not good !

I'd be impressed doing 40 kts at 3 feet in turbulant conditions !

I looked at using a map (manifold absoloute pressure) sensor from a car fuel injection unit but too coarse, we want tolerance of less than 1 foot

Apparently in ww2 the dambusters used a headlight on each wing angled in , when the spot on the water was one they were at release height.

Regards Don

PM  EMAIL  GALLERY  Attn:RR  Quote
06-01-2004 08:21 AM  13 years agoPost 7
mattijs

rrApprentice

Keerbergen Belgium

My Posts: All  Forum  Topic

ok WALT you lost me there...pendulum effect????

Thanks to VITEK over on the zoom forum i realized that a normal GYRO isn't the best idea (seeing it would need to be mounted on some tilting tray to get up and down movements translated into rotation)

I was also thinking about altering a gyro but as it turns out (been reading datasheets from the sensor developer that makes the futuba gyro sensors) that will be somewhat too dangerous.

WHY not just use an accelerometer. Most of them can spot very small movements and i really don't see why an electriconically balanced heli wouldn't work, the control inputs 'should' be what we normally do on our TX's so why would it go terribly wrong like WALT suggests.

Pls WALT, do enlighten me as I don't want too buy a sensor i have no use for.
I think it should be fairly easy to program my beloved SX controllor to monitor both the accelarometer and the collective channel much like a heading hold and than use 3 discrete mixers to get the collective mixed again for full 120 degree control. Later i could program the mixers into software to get the weight down.

PM  EMAIL  HOMEPAGE  Attn:RR  Quote
06-01-2004 08:51 AM  13 years agoPost 8
helicopter34

rrVeteran

New Jersey, exit 82

My Posts: All  Forum  Topic

Insane

Man, you have some pretty high ambitions for your pet projects. If you get it to work though, that would be something. Your #1 problem is going to be weight.

You might end up doing it too because if you add enough weight, it should fly more steady, but thats not really what your going for.

PM  EMAIL  HOMEPAGE  GALLERY  Attn:RR  Quote
06-01-2004 08:59 AM  13 years agoPost 9
mattijs

rrApprentice

Keerbergen Belgium

My Posts: All  Forum  Topic

I ordered two samples from analog.

The sensor is the adxl 202

Here is the datasheet for all you electronic nuts.

http://www.analog.com/UploadedFiles...DXL202_10_b.pdf

Does anybody have an assembly code of a standard elevon mixer or does anybody know a page on the web where the output signal from an RX is explained. I haven't got an oscilloscope at home so it might be necessary for me to know exaclty what signals to anticipate for.[URL=http://www.analog.

PM  EMAIL  HOMEPAGE  Attn:RR  Quote
06-01-2004 09:08 AM  13 years agoPost 10
mattijs

rrApprentice

Keerbergen Belgium

My Posts: All  Forum  Topic

I ordered two samples from analog.

The sensor is the adxl 202

Here is the datasheet for all you electronic nuts.

http://www.analog.com/UploadedFiles/Data_Sheets/70885338ADXL202_10_b.pdf


Does anybody have an assembly code of a standard elevon mixer or does anybody know a page on the web where the output signal from an RX is explained. I haven't got an oscilloscope at home so it might be necessary for me to know exaclty what signals to anticipate for.

PM  EMAIL  HOMEPAGE  Attn:RR  Quote
06-01-2004 02:00 PM  13 years agoPost 11
dnts

rrVeteran

Israel

My Posts: All  Forum  Topic

Mattijs,
I was going to offer you the ADXLs. They are quite small, and with a PIC16F676, you can do all the mixes etc (you need 3 inputs for CH1,2,6, three outputs for 3 servos, and two inputs for the ADXL).
Nir

PM  EMAIL  GALLERY  Attn:RR  Quote
06-01-2004 07:08 PM  13 years agoPost 12
mattijs

rrApprentice

Keerbergen Belgium

My Posts: All  Forum  Topic

Say dnts, you seem to be knowledgeable.
What are your experiences with that perticular sensor and what do you expect from this little project? Doable?
And why if I may ask would you be so kind to offer me these sensors? Got some spare from a project you didn't finish or do you get them for free where you work or something as i see you also are an engineer.

PM  EMAIL  HOMEPAGE  Attn:RR  Quote
06-01-2004 07:20 PM  13 years agoPost 13
dnts

rrVeteran

Israel

My Posts: All  Forum  Topic

Mattijs,
I meant "offer" in the sense of "propose", not in the sense that I send you mine (I have just two of them). I made a tilt sensor out of them and it worked to some extent but they are very sensitive to vibratios and you may need to filter the high frequencies with a very tight low pass filter. I don't know how slow you can make the feedback loop so that it will both respond to the change of altitude and at the same time do not introduce noise to the servos (which will jitter them and can cause premature failure).
It's nice to try, though.
Nir

PM  EMAIL  GALLERY  Attn:RR  Quote
06-01-2004 08:02 PM  13 years agoPost 14
WOS

rrApprentice

NW INDIANA

My Posts: All  Forum  Topic

why couldn't you use a FMA co-pilot to stabelize the up and down as that is what the co-pilot does. I think futaba makes something similar as well.

PM  EMAIL  GALLERY  Attn:RR  Quote
06-01-2004 08:25 PM  13 years agoPost 15
dnts

rrVeteran

Israel

My Posts: All  Forum  Topic

WOS,
Both of these devices stabilize the craft over the X and Y (yaw and pitch) axes. Mattijs wants a Z axis (up/down) stabilization which neither of these devices offer.
The latest from FMA can tell only if the craft is inverted (up-side-down) or not, and roll out of this situation. It does not measure the height or acceleration in the Z axis.
Nir

PM  EMAIL  GALLERY  Attn:RR  Quote
06-01-2004 08:45 PM  13 years agoPost 16
cyhyam

rrApprentice

So. California

My Posts: All  Forum  Topic

Mattijs,

As I understood the “tilting” platform idea, it would pivot on one end and spring loaded to sit horizontal. If the heli tried to change altitude the platform would tend to stay in its position in space and the heli would go up or down, depending. This would cause an angular change in the platform momentarily until the motion is stable. The gyro would be attached to convert the angular change into servo correction.

The platform is going to be using gravity as a factor in the centering method., that is why I called it a pendulum, even though it would be horizontal.

The problem I see is finding a sensor (s) that can distinguish between a descending change in altitude when starting from a hover (steady state), and a deceleration from a steady state descent.

In a steady state descent a force sensor will be at zero and the reduction in descent rate will look the same as a climb. The sensor will try to stop what it thinks is a climb. Airspeed indicators could be used to indicate that a steady state descent is actually movement.


Walt

PM  EMAIL  HOMEPAGE  GALLERY  Attn:RR  Quote
06-02-2004 06:09 PM  13 years agoPost 17
mattijs

rrApprentice

Keerbergen Belgium

My Posts: All  Forum  Topic

WALT, is this what you are saying:

Because the sensor measures acceleration, it doesn't know when it is descending steadily and when you stop descending it thinks you are going up?

If I understand you correctly that is what your pendulum explanation comes down to correct?
I don't think however that that would pose a problem. I will program the controller in such a fashion that it resembles a HH gyro.

If the above situation occurs, it is you who started the motion (going from descending to hover) so the controllor knows it should expect something. It should only work when i didn't do anything and the heli did. If you integrate the difference from the moment the sensors reading is different from what your RX says you want, you approximatly know or can find out where and how hard it should counter act.

I know you probably know the theory behind all this better than me but you might not so pls don't think i am pretending to be the guru while we all know you are ...i just don't see the whole in my theory.


In addition i was also planning on integrating what comes from the RX and see to it that once you start slamming the collective the circuit doesn't intervene. Only when you want the heli to not accerate by almost keeping the collective stick within a certain travel and / or speed **of movement the stabilization should occur.

I had 2 remote adjustments in mind, ofcourse the loops gain and the travel** you can use without turning off the stabilization.

PM  EMAIL  HOMEPAGE  Attn:RR  Quote
06-02-2004 07:34 PM  13 years agoPost 18
mattijs

rrApprentice

Keerbergen Belgium

My Posts: All  Forum  Topic

DNTS, so you've tried the tilting method with a normal GYRO or did you use the sensors and if it's the latter, why a tilting tray when you could've just fixed to the fuselage?

PM  EMAIL  HOMEPAGE  Attn:RR  Quote
06-02-2004 08:21 PM  13 years agoPost 19
cyhyam

rrApprentice

So. California

My Posts: All  Forum  Topic

What I am trying to say is the gyro sensors react to angular change. Angular change will continue to occur on a yaw, pitch, and roll axis until the motion is stopped. How will you convert a linear motion (collective) into an angular one? A tilting platform will return to center once the motion reaches a steady rate, not necessarily stopped.


Walt

PM  EMAIL  HOMEPAGE  GALLERY  Attn:RR  Quote
06-03-2004 12:04 AM  13 years agoPost 20
mattijs

rrApprentice

Keerbergen Belgium

My Posts: All  Forum  Topic

Yeah but me and dnts aren't talking about gyro's anymore, we are talking about acceleration sensors

(atleast i am not, that's why i wondered in my last post wether he used the sensors are a gyro with the tilting platform).

And even with a gyro it would be possible (if you physicly dampen the tilting but by doing so you introduce another set of variables and you don't want that) because if you suddenly have a big acceleration (tilting) you know you are still moving if that acceleration hasn't been counteracted. The damping on the tilting tray takes care of the fact that the tray going to its center position isn't going anywhere near the speed of the heli affected by gusts of wind...otherwise it'be harder to remain stable but however you put it, if a gust of wind makes your heli go up the netto acceleration any device (through rotation or plain measurement of acceleration with sensors) measures is going to be positive. Rendering it possible to give a proportional counter collective input.

+ it's only the time you don't make any control inputs (more than some limited amount/travel) the circuit has to monitor the acceleration, continually adding the last of the sensors samples to a sum. That sum gives a coarse figure of the Z axis speed and it's only because giving collective resets the circuit in a way that it can be accurate enough (as it doesn't have much time to get a large error) without a dsp and more than 8 bits integer resolution.

When i compared the system to a HH gyro what i wanted to say was that if you don't change the height you want the fly at the system should do anything possible to keep it at that height.

Actually i don't mean to try to make it that strict. It would get far to complicated initially to make an entire PID so i'd be happy if the system just worked proportionally alone. Say a gust of wind normally would have thrown the heli 1 meter up, with stabilization it just goes up 5 or 10 centimeters. That'd be perfect. Not too complex to program as it's just proportional. Coming to think about it...that isn't like a HH but just what a regular damping gyro does with a tail. And that's exactly why it isn't so hard to make...it doesn't need to be FULL height control, that's ambitious, it's just counteracting a netto acceleration that has taken place over a period of say 1/5 of a second when your TX hasn't asked for it. A real height control would start messing with the collective once the minimum measurable acceleration is being 'felt' by the sensor, always trying to do anything to prevent change in height. The bigger the disturbance in balance the bigger the reaction...but as it feels it needs to get the heli back into its original position the action has to be much bigger and will continue not only when the netto acceleration is 0 again but it will send the heli down again intentionally to go where it was. A system prone to instability, much more so anyway than a plain damping mechanism.


Cool and usefull nontheless!

PM  EMAIL  HOMEPAGE  Attn:RR  Quote
WATCH
 1 page 841 views
Helicopter
e-
Elec-Helicopters New or Limited Activity
MS Composit › Making a micro better in windy conditions!
 Print TOPIC  Make Suggestion 

 5  Topic Subscribe

Monday, December 18 - 5:16 am - Copyright © 2000-2017 RunRyder   EMAILEnable Cookies

Login Here
 New Subscriptions 
 Buddies Online