OpenxVario Air Speed Sensor - Add on

[tilmanb]

Eageltree driver
https://github.com/paparazzi/paparazzi/ ... peed_ets.c

diydrones driver
http://code.google.com/p/ardupilot-mega ... rspeed.cpp
I'm a bit surprised here with the default value for AIRSPEED_OFFSET (variable _offset in the code)

What I had learned about the diydrones driver a while ago http://web.archiveorange.com/archive/v/ ... V0CeURHam3
I never bothered to finish that train of thought and went with the eagletree sensor.
http://paparazzi.enac.fr/wiki/Sensors/Airspeed

[Rob Thomson]

Looks quite simple really!

So.... the question is. Are you going to jump in and build an openxvario speedy thing!

Now....

Rainer. For sake of keeping things tidy. Do you want this as a sub project? Or merge into the existing tree etc?

Rob

[rainer]

i guess there won't be to much code involved, so once it is working we should implemet a define statement to optionally enable the code needed for it. It makes no sense to create too many different branches for changes like this.

Rainer.

[Rob Thomson]

So...

Been thinking about this a bit more.

Flaps is correct. What is possibly more useful to us is probably an angle of attack sensor - and not an a air speed sensor.

While air speed is an indication of a stall, in the real world - when thermalling your glider at high altitude - what probably is of more use is to know if you are exceeding a certain angle of attack. Too close to the mark.. and the model will stall. (and loose speed)

Really just looking at all options. Is a triax sensor the simple option here?

Rob

[jhsa]

airspeed was, is, and will always be the most important information in an aircraft.. In theory a wing only stall when excceding an angle of attack of more or less 17 degrees.. providing that it has enough AIRSPEED of course

[Rob Thomson]

Agreed.


But an angle sensors can provide a useful feedback... Without then install of a pipet tube. (usefully in many aircraft)

Reality is.. Both would be useful

Sent from my GT-I9300 using Tapatalk 2

[jhsa]

you can have angle of attack and not be stalling.. and you can have a stalled plane with little angle..
It's known that banking an airplane increases it's stall speed and and some aircraft are worse than others..
Bad construction and/or design can make things even worse
I still think that measuring the angle of attack (?in a turn?) makes very little sense..

[Rob Thomson]

Well... With a triaxial sensor, we would know we where in a turn, and in theory adjust the calculation for the relative angle of attack.



Guess the point is. To accurately detect a stall, you need both airspeed, and angle. Without both... Either result is miss leading.

My thoughts on the trial style sensor is an indicator, that your angle of attack is close to the edge. This does not mean you will stall. But is an early warning.



Sent from my GT-I9300 using Tapatalk 2

[Kilrah]

you can have angle of attack and not be stalling.. and you can have a stalled plane with little angle..

Actually not!

What you mean is the pitch, i.e angle in relation to ground, if you have a lot of power you may be climbing 45°measured versus the ground and not be stalling - BUT the thing is that you have enough speed so that your angle of attack (angle between air and wing) stays low, below the critical angle of attack you'd be stalling at!

Angle of attack IS the absolute reference when it comes to stalling. That's basically what the stall warning in any full scale aircraft is based on.

[Rob Thomson]

Yes... Much what I meant. Which is why you need both angle and speed to make anything worth while.

Technically, you can be in a vertical dive - and stalled!

Sent from my GT-I9300 using Tapatalk 2

[Flaps 30]

I know that most of this is off topic with respect to the request for the air speed sensor. This really is pub talk IMO.

you can have angle of attack and not be stalling.. and you can have a stalled plane with little angle....

Hmm... Not as far as I'm aware if you give it some thought. What you are saying only makes sense with numbers. Sure you actually need an angle to be produced between the air flow direction and the wing chord to produce lift. The sensor senses the angle at which that air is flowing too high an angle and the laminar flow breaks and a stall follows close on that point.

It's known that banking an airplane increases it's stall speed and and some aircraft are worse than others....

That is true. G forces increase hence wing loadings rise and the most important thing is that the angle of that air with respect to the wing changes in a turn.

Bad construction and/or design can make things even worse

Yup... And Rob's model falls into that bad design area. ... Not really.. High aspect ratio aircraft all have a greater tendency to drop a wing in a turn. The way that some designs reduce that tendency, is with stall strips on the inner sections of the wing.

I still think that measuring the angle of attack (?in a turn?) makes very little sense..

I would challenge that. If you do the hand flying thing and give it some thought, you will see (maybe) that the airflow wrt to the wing does alter and that is what we really need to know.

Do a search on - 'John Farley angle of attack' - You will find some good reading material.

[jhsa]

To accurately detect a stall, you need both airspeed, and angle. Without both... Either result is miss leading.

I disagree.. Maybe you need both on modern aircraft where the pilots swapped the stick with some white gloves and a computer keyboard.
But on real aircraft (like the one in the picture) you only need airspeed.. The plane I learnt to fly had only an airspeed indicator as the main gauge.. the second one but a little bit less accurate is your face.. the third one and also quite accurate for detecting a stall is your a**

Yes, this is the plane I've learnt to fly, yes it's me at the controls and you can see the little tube that is the airspeed indicator. analog one. it's a little disc inside a tube
on this plane, the drag is so great that without engine you will stall it even with negative angle of attack you really need to put the nose down and keep your airspeed.

[Rob Thomson]

Bad construction and/or design can make things even worse

Yup... And Rob's model falls into that bad design area. ... Not really.. High aspect ratio aircraft all have a greater tendency to drop a wing in a turn. The way that some designs reduce that tendency, is with stall strips on the inner sections of the wing.

That gives me an idea!

I am going to run some trim tape parallel to the the leading edge, just behind the highest point on the wing.

Believe the proper name is a turbulator?

This will force the air to separate faster in the center of the wing - resulting in a less aggressive stall.

Flight Design 101

[jhsa]

Are you going to put some "dirt" on a glider's wing surface?? You better get that blowtorch out and give it some washout, young man

[Rob Thomson]

If you put the strip un the right place, there will be no drag increase.

It ends up at the point in the air flow where the air separates anyway. (or at least that's how I understand it)

Sent from my GT-I9300 using Tapatalk 2

[jhsa]

If you put the strip un the right place,

Hmm, last time I did that I ended up married and with a son..

[Rob Thomson]

[Rob Thomson]

Interesting reading here:

http://www.standardcirrus.org/Turbulators.php

[jhsa]

here is why airspeed is so important.. forget the AoA. If you don't have enough airspeed.......

[BBvideo 425,350]https://www.youtube.com/watch?v=zxbulrrQVig[/BBvideo]

[Kilrah]

Not enough airspeed is just one of the reasons that causes an excessive AoA and results in a stall, nothing more

[jhsa]

Put drag into the equation and you can stall an aircraft with negative AoA..
As I mentioned before, the brick I'm flying above in that picture really need the nose down without motor or you will stall it..
The only instrument you can't take out of the coclpit is the airspeed indicator and not the artificial horizon that gives you the angle of attack..

[Kilrah]

Put drag into the equation and you can stall an aircraft with negative AoA..

Definitely not

You can stall with a trajectory that has a negative angle with regard to ground yes, but AoA will actually be very positive if you do!

[jhsa]

well, create drag on the wing.. destroy the airflow, and you see if you stall it or not..

All this is way out of topic maybe we should create one at the pub.· but the fact is.. every aircraft has an airspeed indicator and not all aircraft has an AoA indicator.. there must be some reason for that, right? and these ones are easier to build than the airspeed indicator.. Not all aircraft has a artificial horizon, but ALL have an ASI..

But I'm not surprised anymore.. one day I asked a commercial pilot which one was the most important instrument in the cockpit.. He answered, the altimeter
Must have pulled too many banners that day, for sure

[tilmanb]

AOA sensors are a bitch on small airplane.
You can make them with air pressure sensors in the nose by measuring the pressure difference in holes at different sides.
Or more straight forward with vanes on non friction magnetic rotational encoders.

Both options are fun to think about and provide quite useful information for research purposes. But are rather impractical for real model airplanes.
Of someone comes up with a good sensor platform I will use it in my UAV. but not in a model airplane.

I might get a chance to put my vario together next week. But not even that is certain as I host my family at my house and I would disappoint them if I did electronics stuff too much instead of touristy stuff.

[jhsa]

I would like to have IAS if I did FPV, that is for sure.

[Flaps 30]

Interesting reading here: http://www.standardcirrus.org/Turbulators.php

This is what I was thinking of -----> http://en.wikipedia.org/wiki/Stall_strips

but the fact is.. every aircraft has an airspeed indicator and not all aircraft has an AoA indicator.. :

Of course it's nice to have both. Given a choice. I would go for the AoA indicator. It is the angle of the airflow coming onto the wing that decides when it stalls. Yes you can stall it by flying slow. But what is really happening when you do that?

jhsa - How does the stall warner on a light aircraft operate and what is going on with it when it activates. Do you remember what you did on the inspection to test the stall warner?

Worth a read -----> http://en.wikipedia.org/wiki/Stall_%28f ... ty_devices

[DHDSP]

Hi everyone,

about one weeek before Rob started this thread, I had the same idea of using the aformentioned sensor. the reason is I am building my first bigger acrobatic sailplane - the 4m Lunak from Reichard. This sensor is probably used by thousands of arduplane pilots, the code is readily available, and the sensor is cheap. IMHO this is the way to go - even if sensitivity is not optimal. I would really love to have an airspeed indicator on my Sky9x transmitter.

I understand the AoA is probably the better (?) variable, trying to determine stall conditions, but a lot more difficult to measure. So why not start with the easy option?

Actually.. I'm a Pilot too, flying the only aircraft where you can "see", "hear" and "feel" stalls: a paraglider. Actually maybe even smell the thermal blob smelling of grass.That's why I asked Rainer if it would be possible to have an audio output on the OvX, to have it as a backup unit to my Bräuniger GPS Variometer
Although stall detection could actually save lives (landing is always a problem here, going slow AND getting the AoA changed by thermal blobs at the landing spot), noone I know has an airspeed probe attached.

Peter.

[Flaps 30]

Although stall detection could actually save lives (landing is always a problem here, going slow AND getting the AoA changed by thermal blobs at the landing spot), noone I know has an airspeed probe attached..

Yes you would be very well aware of the importance of the AoA with a Paraglider. Wings just love to collapse when the AoA gets too much, especially when one half of the wing hits a strong thermal you have to be quick to pump the wing back up.

More info on stall warners and stall strips here that is worth a read. I love the reed horn. Never met them on the aircraft I flew. ----> http://members.eaa.org/home/homebuilder ... vices.html

[DHDSP]

I have the same problem as Rob: my Lunak has pre-built wings. I've been meaning to put a removable motor into the nose, so the pitot tube should be mounted outside of the influence of the prop airstream.

There are a lot of really small wireless modules for the arduino, like the NRF24l01 or RFM12B. Maybe we can find a module than can be convinced to transmit the output of the airspeed module without the need of an extra arduino ant the wingtip.

[Kilrah]

You still need a wire to power it, so why not run the data through it?