Gyro Stabilized UAV?

[rclinks2002]

Hello,
In designing our UAV I have realized that a traditional Infrared Horizon sensing system (FMA Co pilot) will not work for our application. I was wondering if I could use a helicopter heading hold gyro as a roll stabilizer? I assume that since it is a heading hold it has automatic drift adjustments. Any thoughts? Thanks
Ben Mitchell

[Crazyrcer]

I know gyros have been used before, but I'm pretty sure they won't work like a wingleveler. Some people have used them to stabilize thier planes during gusty/windy conditions. It might be worth buying a GWS gyro ($40) and seeing how it works.

[d_wheel]

Why will the IR system not work? I have used them with success on everything from park fliers, to flying wings, to helicopters.

Heading hold gyros slowly drift over an extended period, and need input from an outside source to correct this drift from time to time. Not sure how this could be done on a UAV.

Later;

D.W.

[rclinks2002]

The UAV is for mapping of forest fires. I am worried that the Infrared unit will percieve the fire as the sun and cause the plane to be uncontrollable. Can you think of an alternative for using gyros or infrared? Thanks

[d_wheel]

The IR unit I have been using (FMA Direct Co Pilot) has very little problem with the sun. I fly it regularly late in the evening when the sun is low, and the only thing it does is try to turn slowly away from the sun. It is easily controllable under these conditions. I would think it would be the same with a fire, that is, it would try to turn away from the direction of the fire.

Their is an autopilot available from u-nav.com that might work for you. When flown without a wing leveler, it must be used on a very stable aircraft (high wing with dihedral) trainer type platform. Their high end PicoPilot includes gps navigation, altitude hold, and airspeed hold.

Hope this helps;

D.W.

[rclinks2002]

We have figured out an electronics option for flying the plane, but a Infrared wing stabilizer will not work because a large forest fire emitts a significant amount of infrared. I am worried about the unit recognizing the fire as the sun and the sun as the sun also, so perpetual rolling....Any thoughts?
Ben

[CenTexFlyer]

Actually, as I understand it, a thermopile doesn't actually register the heat from the sun. It samples the IR temperature difference between the ground and the air. We have flown our wings over a 30 acre controlled burn and had no problems with smoke or the heat signature of the fire itself. It was actually kind of fun "surfing" the heat plume along the fire line.

[kd7ost]

How big is your UAV? There is a modified gyro system out there and it uses some fancy software to make it work. In fact, the gyro is set to sense yaw. It's very, very sensitive. It manages roll in it's operation though. It needs to be flown on a conventional aircraft. http://www.maxxprod.com/mpi/mpi-16.html, scroll down the page for the BTA AS 07. I have one and it works well. It doesn't need to see outside conditions and does not require calibration to outside conditions. It's bigger than the co-pilot by a fair piece but only draws 30ma to operate so it's easier on your power system than a control surface servo. You can't use it with 2 meter communications on board though but I presume you're not. The BLM, Forest Service and other DOI and DOA, freq's at 160 to 170 mhz aren't a problem as long as the radio isn't on the plane. :D

I'm with you on the thermo piles. When those Forest fires get big enough they generate there own weather cells. Huge tree's explode. Of course when they get like that you pull back anyway. I work at the NIFC in Boise in the Remote Sensing Fire Weather Support Unit. I wouldn't mind being involved with your project. The incident control aircraft and pilots are the biggest obstacle. :rolleyes: I would like to loiter an airborne repeater for communications. The grass and range fires don't generate the same kind of geographically isolated heat.

Dan

[rclinks2002]

Dan, I looked at the BTA As07. It has alot of the features that we have already created/put together. Such as the GPS unit, and telemetry. The only problems we are having is a video transmission system, the stabilizer, and a reliable altitude hold. I looked at the guy who converted a RCAP to do altitude and I would like to wait until walter integrates it into one system. (rather than having 2 systems) Any thoughs?

Thanks
Ben

[kd7ost]

I think that's sound thinking. ;) I do have an altitude hold unit sold by UNAV. You can get it as a stand alone unit even though it doesn't appear on their website. It's the Pico Alt 3. http://www.u-nav.com/. I connect between receiver and elevator servo. The systems are connected dynamically then through the airframe and its flight dynamics. There's nothing wrong with using separate systems. They work together. I use a GPS to rudder steering unit. When it puts in a yaw command to steer, it rolls the plane. When the plane starts to roll, the roll control device detects the roll and inputs aileron to keep it level. This makes for nice camera platform because the ground isn't arcing left and right making complex coordinated turns. The cost does run up when you buy separate systems like that though.

Dan

[rclinks2002]

I think the "pico tilt" with a converted rcap would suffice dont you think? (I am on a too low of a budget to afford the full out alt 3 ($250) Any thoughts?

Thanks
Ben

[Cerberus]

Head holding doesnt have compensation in it for drift, for a futaba its just a MEMS so its got low drift compared to a Pizo device but it does have drift and if your thinking of using it as a sensor to determine roll angle your looking at integrating the output and hence integrating the error. You really need an IMU, then you can find the gravity vector (assuming vibration is sufficiently low, and of zero mean(an acceptable assumption) and bleed it into your Gyro reading... i.e. the gyro provides quick feedback to correct for sharp changes due to wind gusts etc. and the accelerometers in the IMU provide a longer term estimate to correct for drift.

I worked with a team on this approach for an indoor draganflyer (helicopter) with some success. Dihedral or other form of inherent mechanical stability would also be a cost effective option.
Cheers
Will.

[MattChave]

Magnetometers guys .... Are you getting sick of me saying that yet?

Let’s try work out your problem quantitatively...

Ok according to:
http://atlas.gc.ca/site/english/maps/environment/forestfires/1
“woody fuel ignites at around 350°C. The maximum temperature in forest fire flames is in the 800 to 1000°C range.”


The sun has a mean surface temperature of 6,000°C (internal temperatures are around 15,000,000° C)

I can’t find the FMS thermopiles datasheet so I’ll assume their working around the long wavelength infrared (8-15 µm lets pick the middle of this range (10-11) and run through it for that) since the atmosphere has a wide range around these values where IR absorption is a minimum see: http://www.electro-optical.com/bb_rad/atmo_abs.htm

Read up on the plank radiation formula, S(lamba)= ((8*pi*h*c)/lamba^5)*(1/(exp(h*c/(lamba*K*T))-1))

S_sun@10um = 193.5346 J/m^3/lamba
S_forrest@10um = 23.7806


The sun has a radius of 6.96 x 108 m
So area that’s emitting our IR is (suns total surface area) = 4.5655e+018 m^2

So the sun at 10-11µm is emiting:
P_sun = 0.5532*10^23 W

And the surface area of the sphere that it is emitting it over by the time we get it on earth is 3 *pi*r^2 where r = 150,000,000,000m
So surface area = 2.1206e+023 m^2

So the power density from the sun at the this IR wavelength before our atmosphere is about 0.5532*10^+23 /2.1206e+023 = 0.2609 W/m^2

And only about 80% of that looks to get through so:
Intensity_sun = 0.2087 W/m^2
(Guess that sounds reasonable, remember a thermopile will have a wider bandwidth so will collect much more power but this argument should still hold for relative intensity)

Let’s now work it out for the forest fire:

Ok let’s make a fire 100m x 100m and see how we go if we were 1km above it.

P_forest = 1.5266e+003 * 100 * 100 = 1.5266e+007 W

Which will spread over half a sphere of radius 1km (the grounds probably a pretty good reflector)

= 9.4248e+006m^2

so the intensity on the plane 1km above the fire is 1.5266e+007 / 9.4248e+006m^2= 1.6198W/m^2

That kind of corresponds to a continuous crown fire that http://atlas.gc.ca/site/english/maps/environment/forestfires/1 says “can emit 10 000 kilowatts per metre or greater”

So in conclusion:
(Someone should really go back over this, since it’s pretty late when I’m doing this 12pm NZ time)

The forest fire will appear 7.76 times brighter at a wavelength of 10-11µm than the sun above it when 1km away from a 100mX100m raging forest fire.

The sun won’t really normally cause too many troubles because its peak emission is 4.6229e-007 (somewhere near green) which is well above the IR frequencies we are interested in. Our thermopiles job is to measure the cosmological background radiation, and the earth, NOT the suns emissions. That’s why it works at night too. All it is, is basically a thermopile. If it sees space (as in the final frontier) then it might output 1V, if it sees the ground only it might output 3V. If it sees half and half then it might output 2V.
So to level the plane it wants to measure the same voltage on both sides. That’s really easy.

If you want to see how we tackled it with magnetometers see:
http://uav.bravehost.com


Hope this helps,
Matt Chave

[rclinks2002]

Matt,
Wow thanks for the math calculation. How much would one of those units set me back (just for the flight stabilization, not the whole navigation system). Thanks

Ben Mitchell

[d_wheel]

"Magnetometers guys"

Yes! I have felt for many years that magnetometers would be one of the better tools to use for UAV purposes. If my memory serves me correctly, Maynard Hill (who was in charge of the model airplane trans atlantic flight a couple of years back ( http://tam.plannet21.com/) ) used this method to build autopilots for some of his models back in the 50's or 60's. In one magazine article, he strapped a pair of pliers to the tail of one of his airplanes which moved the cg back significantly, but the autopilot flew it as if the cg were perfect.

I can't wait to read all about how you did it!

Later;

D.W.

[kd7ost]

I think the "pico tilt" with a converted rcap would suffice dont you think? (I am on a too low of a budget to afford the full out alt 3 ($250) Any thoughts?

Thanks
Ben


Holly crap, thats a lot of math far over my head. I laughed when Matt said someone should go over it and check it. :D Looks good to me. :p

I'm not much of a theorist and am an applications guy. But boy we sure can get stuff done working together. So a Forest Fire is gobs worse than the Sun for a thermo pile. Especially in the numbers where you listed a fairly small fire at only 100 x 100 meters.

Ben
The Pico Tilt will work to stabilize or keep looking in one axis, a camera on a platform flying with coordinated turns. Is that what you're after? The pico tilt won't stabilize the aircraft.

It might be good at this time to define a typical mission parameter you're looking for. That will help narrow the search to what equipment works best. The Plane need to fly via RC or autonomously?
The airframe we are using is X ?
Roll control is needed or inherent stability around the roll axis is OK ? Like Will said, Dihedral! Maybe Polyhedral?
The camera is a video link? looking out, down etc.
Fixed digital Camera on board? Need to see what it sees for pictures?

MEMS sensors would be the way to go but off the shelf products using them are going to be priced indicative of supply and demand. (They won't be real cheap) The true cost of these is not the sensors themselves but the design time and software needed to get something off the ground.

I can't find any current information on Maynard Hills early stabilizing efforts nor do I recall how involved he was with those historical works. I do recall he used what was called "Electrostatic stabilization" which worked by detecting the earths magnetic field somehow. But there had to be some rather benign radio active sensors on the wing tips as will as nose and tail. Today these might be the kind of sensors found in some fire or C02 detectors. There was a simple circuit that went with it. I have no idea why it isn't in use today or where to find information on it. Perhaps you could google it.

Dan

[rclinks2002]

Dan,

Ok following your last post I will answer the questions that I have been avoiding for quite some time. (mostly because we havent thought about it too much) ok here goes...


1 "Is that what you're after? The pico tilt won't stabilize the aircraft."
I guess that the pico tilt will not be a viable option, I need something that will act the same as the FMA co pilot but doesnt require Infrared signiatures.

2. "The Plane need to fly via RC or autonomously?
The airframe we are using is X ?
Roll control is needed or inherent stability around the roll axis is OK ? Like Will said, Dihedral! Maybe Polyhedral?"

The plane will be using a basic RCAP with a real time GPS map overlay telemetry. So the RCAP will be used to fly the plane to the fire then using a rather creative comp program will map the edge of the fire. The air frame is one that we are developing ourselves (we are AE students not EE students). It will be a polyhedral wing and there will be initial wind tunnel testing on scale models.


3. "The camera is a video link? looking out, down etc.
Fixed digital Camera on board? Need to see what it sees for pictures?"


A classmate (pikeater) and I are actually talking to you in a different forum in relation to the video transmission issues. Using a black widow av camera system will work for testing but as stated, we cannot use it for commercial purposes. There will be a standard digital still camera looking through the bottom of the fuse and will automatically be taking a picture every 5 seconds (time is adjustable) and storing them on a standard SD card. This is in case we manage to miss something with the fire camera. (These pictures do not need to be seen until the plane is on the ground.


4. I looked at the http://uav.bravehost.com website and I sent them an email about it. I think it might work but I am unsure. I have looked at other commercial devices and they seem to use a combination of gyros and accelerometers to stabilizer the aircraft.


Any thoughs?? Anyone???

Thanks
Ben M

[rclinks2002]

Dan,

I also saw your UAV on a website it said it was 25 lbs What sort of engine did you use on yours?

Ben

[kd7ost]

I use the Fuji BT32A. They now have the BT32B. It has a little more power. It's 32 cc or 1.95 cid. I have very specific tastes in a UAV power plant.

Gas powered with a Spark plug. This is cheap to run and is highly reliable.

It is economic in fuel consumption too. I can fly for 1 hour on 1 quart of gas at 50 to 60 mph. Fuel is a standard low octane (regular) and 2 cycle oil mix. It burns clean and is available in any little town or fire camp.

I only select engines that have a magneto ignition system. This means it's a stand alone power plant. I don't need an electronic ignition system, (which in and of itself is OK) or the battery pack needed to power it. That extra battery pack is what I don't like on board. More to go wrong, more to charge.

Finally, I select an engine that you can get a spring starter for. The Zenoah G26 is a great engine and is smaller than the smallest fuji. But you have to drag a battery and a huge starter, I mean a Sullivan Dynatron to start the engine. You also have to have a spinner to fit the engine for the starter cone to fit to. With a spring starter you don't need as much field gear. Just a big heavy duty padded leather glove. ;)

I would replace my Fuji, which is a wonderful engine btw, with a Zenoah G38. It's just a little bigger at 2.3 cid, you can buy a spring starter for it and it's cheaper than the Fuji or even it's smaller sibling the G26. It's been around for years and is a great highly touted and trusted engine. I really believe it would give the most bang for the buck. Props are expensive though if you design a pusher. Keep the engine high enough so it doesn't suck up rocks on the take off run.

My GPS onboard the plane shows over 1500 miles of travel. Never a dead stick. You do have to select soft mounts for those thumpers. If you guys decide to head that direction I can stear you to the mounts I used.

Dan

[rclinks2002]

The power plant section of out group was looking at using a axi 5130 for power, but I dont think it will be enough power to get our plane off the ground. Also the website said your plane was 25 lbs, what was the wingspan?
Ben

[kd7ost]

It's actually 22 pounds but thats close enough. Must have been on the UNAV site in the Gallery?

It's 102 inch span, 87 inches long. It comes apart to pack smaller.

Electric would be an arguable good choice for many missions. But for me I wanted a plane I could land, re-fuel and put back up. I can carry double the gas for 2 hour flights at WOT. Throttle back and I can get 3 hours. My battery power systems allow me to fly in excess of 12 hours but I have to land to re-fuel. Electric won't do that yet at a cost we can afford. A plane this size has a lot of interior room and heavy lift capacity to allow growth and development of systems. It is ambitious to put together though.

Dan

[kd7ost]

Dan, Ok following your last post I will answer the questions that I have been avoiding for quite some time. (mostly because we havent thought about it too much) ok here goes...
Any thoughs?? Anyone???
Thanks
Ben M

Ben,

I'm not going make any recommendations to the approach you guys take. The reason I was asking all the questions was trying to isolate what the specific mission was about. In the end, it's all quite similar to what the rest of us are doing. You are remote sensing with a visual product. There are questions I never asked back when I started on this path of my own. If I had a better idea back then as an end result, I would have saved my self time and money in building and R&D.

Philosophically speaking of course.

When I started I was stuck on the idea that I had to have my plane fly coordinated turns, that is roll with pitch and yaw, all thrown together. Heck, that’s how we fly our models and full scale planes. Then I would have to develop a camera platform that would keep all the unwanted flight maneuvers out of the camera. I wanted it to look straight down all the time. The Pico Pilot line of products was not available back then. I did what I could to use dynamic systems. Dihedral, polyhedral, well balanced around pitch axis, Co-Pilot, BTA-AS-06, then BTA AS-07, PDC-10, PDC-15 and PDC-20, GPS etc. None of these did the job the way I pictures them and I spent months tweaking gains and test flying to try to get the coordinated turns. I came up with some amazing Rube Goldberg camera platforms. In all the goofing around and watching the plane, it finally dawned on me that I didn’t need coordinated turns. Once I cranked the gains back up I could see that the plane would fly just fine while being stabilized in the roll axis. The Co-pilot as well as the BTA units would cross couple aileron for me so when a turn was being made, the camera stayed looking straight down even though it was on an easy to make solid mount. I was able to build tilt into it and learned to fly by video using rudder only to steer around once I got up to flight altitude. It was all so simple. A three axes plane was a camera stabilizing mount. Don’t coordinate the plane then stabilize the camera. Just stabilize the plane. The camera is steady then.

I have a real problem in this philosophy with non aileron aircraft then. It’s been a spot of contention with me and others at times. I don’t use dihedral or polyhedral for stability. I have pictures to take. I don’t want my yaw steering to induce a roll. I want the platform to stay flat. It takes up a little more space. But we aren’t talking big planes. The roll control devices can all be overridden at anytime by simply putting in aileron. I do this for take off and landing. It’s slower to roll this way but we’re basically flying cargo planes at this point. My plane won’t perform a loop or a roll the way I have it set up. They need to be stable so that’s what you’re looking for. The plane is built as a camera platform instead of just hanging a camera on a plane.

I don’t know if any of this is helpful or you guys may have already though it through. Defining a final mission parameter helps you back up and design the platform that meets those parameters.

Dan

[rclinks2002]

Ok, I will write a long reply to your post in a few hours...(I have to get going soon) what about this idea....our plane will have a 144inch polyhedral wing so roll stability will be pretty good.(Its basically a large motor glider) then we add the RCAP unit with both navigation and altitude hold. If I remember right on 2 channel gliders when you turn you have to apply back pressure to keep the plane from dipping the nose. So the altitude hold will prevent the plane from losing altitude while executing a regular rudder turn....Any thoughs?

Oh yea, I will try to PM you a preliminary drawing of the plane....
Thanks
Ben

[kd7ost]

Ok, I will write a long reply to your post in a few hours...(I have to get going soon) what about this idea....our plane will have a 144inch polyhedral wing so roll stability will be pretty good.(Its basically a large motor glider) then we add the RCAP unit with both navigation and altitude hold. If I remember right on 2 channel gliders when you turn you have to apply back pressure to keep the plane from dipping the nose. So the altitude hold will prevent the plane from losing altitude while executing a regular rudder turn....Any thoughs?

Oh yea, I will try to PM you a preliminary drawing of the plane....
Thanks
Ben

Sounds good. You don't owe me any explanations though. Who's the pilot in the group? The back pressure is typically refered to in aircraft jargon as "up elevator" for pitch. :)

The design sounds like it will do the job alright. Dynamic roll stability will prevent you from having to purchase and deal with a roll control device. I concider that a substandard method for this type of work though. It is a good starting or entry level place to begin though. If you're just capturing video, it will do fine. If you're trying to geo reference or map out spots on the ground with digital stills, you need to get the roll and pitch component out of the camera, or figure out a way to determine where the camera is, what angle it is taking the picture at, how high up it is etc to locate the spot of interest on the ground with any accuracy. If the camera is looking straight down when you take the picture, the lat and lon of the camera is the lat and lon of the center of the picture. (Much much closer anyway) If location isn't critical, this is all moot. :rolleyes: The higher you fly, the bigger the error.

Your assessment is correct all things being equal. If you have an altitude lock and you induce yaw for a turn, the nose will typically drop. A small measure of up elevator will be needed to keep you at altitude if you employ the system like an RC plane. Another option is to use full scale techniques. By having a small degree of up thrust in the engine mount, your plane will climb at full throttle. It will descend with less. Some people use engine power for altitude control instead of elevator. The Pico Alt 3 from UNAV comes in a T or E version. The difference is each has software tweaks that optimize E = Elevator, or T = Throttle. In my mind there are pro's and con's to each and most of the choice depends on the power plant you use and how optimized it is to your airframe.

Over a fire environment you will have extreme's of lift from convection currents. You need to be aware of the potential for extreme thermal activity and high wind where the hot air and cooler air mixes.

Dan

[rclinks2002]

Dan,
With the back pressure, I was rather tired when I was typing the last one so I wasnt thinking too clearly. As it turns out 3 out of the 4 of us are pilots, But pursuing AE degrees (even though we have only had 2 AE classes) not the infamous riddle air taxi degree. I have looked at the picopilot sp which aparently has a one axis wing leveler but I am not a fan of having to program the flight path into the unit. Really the RCAP in our situation is only for a flyback feature. So if something happens or we lose contact with the plane, it will automatically activate the RCAP and fly the plane back to home base. With the camera, and mapping functions, standard forest service maps only have a 10 meter degree of accuracy 1:50000 map. With the IR package we are using the camera and program will fly to the edge of the fireline and run sort of like a line follower robot that you see in science catalogs, and will trace the fire line onto a map overlay back on the ground in real time. Right now a good portion of the FS offices use a person sitting in a helicopter staring out a window with an IR camera at 4 in the morning. (when the fire is least active) This is very expensive. With regards to changing weather conditions, since the fire is lease active at 0400 fire weather will still be a problem but not out of reach. The UAV will not be flying during the times when the fire is most active. (1300-1700) Using full size planes to map fires during these times are generally frowned apon as to not conflict with airtanker operations and turbulance. As with the still camera on the bottom, that is mostly for trival reasons. Just a way to take pictures from the air. I havent fully researched it yet but, there really isnt a need for a std still camera on the airplane anyways. Its just something to look at in awe....



Ben

[kd7ost]

Understood. That explains it quite a bit more. You're right the incident control aircraft won't like anything in the air besides them. Too much happening. Sounds like you guys have it well thought out. It sounds pretty ambitious and I wish you the best of luck. If I make a trip down your way this fire season I'll have to look you guys up.

Dan

[ElectroLawndart]

Hi All,

This looks like a job for subsumption behavior. Instead of one sensor to keep the aircraft level, you have numerous sensors. You might have a tilt sensor to tell you if the force of gravity is perpendicular to the wings. Since the tilt sensor can be fooled in a coordinated turn, you have a magnetometer to measure heading and heading changes and maybe the GPS to backup the magnetometer. This also works for the pitch side of things, just have another tilt sensor measuring pitch and substitute the magnetometer for an altimeter. In essence, there is data from a primary sensor, a secondary sensor and maybe a tertiary all weighted differently. Sum all the data. Divide by the number of sensors...Take that data with a grain of salt. :p

I can only imagine the turbulance that will be encountered over a forrest fire. I've only flown in a Piper Warrior from LA to Vegas on a hot summer day and that scared the crap out of me.

Maybe pull the IR filters from a couple of CMOS cameras and mount them facing straight out and look at the fire from an oblique angle. At least that way you don't have to fly directly over the fire line. I believe that even the attack bombers don't overfly the fireline but come in from the front and throw the water/retardant while in a banked, coordinated turn. Not barreling through the flames like in the movie 'Always'

The views given here are not nessesarily that of the informed and it's late.
Dart

[d_wheel]

kd7ost,

Where did you get the shock mounts for your engine? How well do they work? I am mounting the same engine on a new bird and would like to have some kind of shock mount system.

Later;

D.W.

[Tuner]

Have any of you used the RCV engines. Their shape and 2:1 geraing in the Inline version is a tempting powerplant.

Is gasoline based that important in uav's.
I can see for military use but for industrial/commercial????

Scott

[uthere]

I've thought about using a 3-axis magnetometer for pitch/roll stabilization as well, but realized that a best a magnetometer can do is tell you how the plane is oriented in relation to a magnetic "axis", yielding a circle of possibilities, not a fixed orientation. For example, if the vertical inclination of the earth's magnetic field is 70 degrees in your locale and your magnetometer shows the plane lined right up on that field, you know that the plane is pointed 70 degrees from the ground, but it could be rolled in any orientation along that axis including upside down.

BTW, there's a great tool for finding local magnetic inclination and declination at: http://www.ngdc.noaa.gov/seg/geomag/jsp/struts/calcPointIGRF



(http://www.ngdc.noaa.gov/seg/geomag/jsp/struts/calcPointIGRF),


-Jim

[kd7ost]

kd7ost,

Where did you get the shock mounts for your engine? How well do they work? I am mounting the same engine on a new bird and would like to have some kind of shock mount system.

Later;

D.W.

D.W.

I got them from Estco Enterprises. http://www.estcoenterprises.com/display_items.php?cat_id=9&line_id=47
The Web Site doesn't have all the parts. I used style mm312. These have a 185 pound compression and a 67 pound shear load rating. They are very stiff but do the job for me quite well. Idle is the biggest problem to solve. These mounts have 5/16-18 thread so it's outside the local hobby shop supply for blind nuts etc. I went to my local Home Depot or Lowes for hardware. They have a 1-800-696-7828 phone number or better yet just request a calalog from the web site. The isolators are 5/8th inch thick and 1 3/8 inch diameter.

Dan

[kd7ost]

Have any of you used the RCV engines. Their shape and 2:1 geraing in the Inline version is a tempting powerplant. Is gasoline based that important in uav's. I can see for military use but for industrial/commercial????
Scott

I've never owned an RCV but it looks like a good engine. I don't know anything about a 2:1 gearing unless you mean it's a 4 stroke. I do use glow engines but my long time use gasoline engine is a lot more reliable and very cheap to run. It's also very clean. Over 1,500 miles logged and never an engine out. All our glow engines leave quite a residue of lube. My plane is pretty big. Gas is the logical choice unless you have real deep pockets.

The Military uses heavy fuels in all ICE's. Even in their small ICE UAV's. They no longer use, transport or store gasoline in tactical environments. Military wet fueled UAV's burn kerosene to diesel to JP.

At the end of the day, there are several good choices in power plants for a lot of good reasons IMO.

Dan

[Tuner]

Good points Thanks!

[MattChave]

Well done Jim, welcome to these forums, Actually you're right, it is in fact worse than you describe (but fixable), occurring at all orientations. I think you understand it, but just missed a bit in your explanation there.

When you measure the magnetic field vector, with the plane at some arbitrary orientation, you can deduce that the planes orientation must lie somewhere on a cone of rotation with the magnetic field as its axis. So then you either need one more measurement on one of the axis, such as an accelerometer/horizon sensor/gyro etc.. or you make an assumption. By knowing one more state you get a unique solution. What we have done so far is use an airframe with inherent roll stability and thus assume the roll to be zero (not quite true, but it works very well). If the roll, pitch or heading is known then there is only one possible unique solution on the cone of rotation which we can be.

In your example we are very happy because we now know the pitch and heading of the plane, and so only need to calculate the roll if we want too. When the aircraft isn't lined up as nicely as in that example though, the roll needs to be determined, since because we are somewhere on a cone of rotation a different roll, gives a different heading and pitch.

Yeah that government site also has some source code for figuring out the magnetic vector at your location.
Unless we're going long distance we don't need to calculate the vector onboard rather just upload it before the flight, if we decide to go further we can use the calculator you pointed at and make a look up table and adjust the reference as necessary, or as I'm coding now, calculate the reference from the gps at the start of the flight and then again at some predetermined radius to that, or time, or waypoint selected by the user (haven't decided yet), so that it will work straight away no matter where you are in the world.

Hope this is clear enough,
thanks, Matt

[d_wheel]

D.W.

I got them from Estco Enterprises. http://www.estcoenterprises.com/display_items.php?cat_id=9&line_id=47

Dan

Thanks Dan,

I will give them a call. Also, what size fuel tank are you running? Have you had any trouble with cracks in the crank case? I've read that a few of these engines have had this problem

Later;

D.W.

[kd7ost]

Thanks Dan,

I will give them a call. Also, what size fuel tank are you running? Have you had any trouble with cracks in the crank case? I've read that a few of these engines have had this problem

Later;

D.W.

The only trouble I have had is with the stainless steel mount. First, in order to fit those softmounts you'll have to drill out the mounting holes. It takes a real good bit. Not a cheapie. I went through a few mounting iterations to finally settle on the soft mounts I mentioned. I used 4 each 3/4 inch tall by 3/4 inch diameter mounts at first. They did great at higher RPM but at an idle, wow, it was ugly. They "walked" too much at low RPM. That might be where this damage began. After about 1000 miles on the plane I found these in a breakdown and clean up engine inspection. I replaced the mounts with a new set ordered from Tower and made extra care to ensure everything fit well without straining or stressing the stainless. They're hard as heck but brittle if you try to bend them so I advise taking care with ensuring tightening bolts don't draw the mounts tight. You need to make them fit well. I haven't had any repeat problems with the new set but I watch them like a hawk. Fuji should have stayed with an Aluminum mount IMO and as far as I can tell the rest of them are. I also use the Fuji spring starter on the back shaft so had to fabricate a mount to accomodate that.

I use a 32 oz Du Bro tank with the gas fittings. I get 1 hour of run time at WOT with that. I can stretch it out to 1.5 hours by dropping to cruise RPM.

Dan

[jescardin]

Hi friends:

YES, a RPV may be roll stabilized with a model gyro. For this, you must set up the aircraft as follow:

1.- You can ONLY use normal gyros (mechanical or piezo) but NOT a "heading lock" one.

2.- The giro must be mounted with sensibility axis VERTICAL (it "feels" pure yaw).

3.- The giro must control AILERONS (not rudder).

With this setup you get an incredible autoestable platform and you will need an specific aileron control procedure with more stick movement and KEEPING the stick position during the turn as if you release it the plane will immediately and briskly self autoestabilize.

The above setup may be improved on two points:

A.- For avoiding inverting the aircraft during turns you have to greatly reduce aileron rate while on "assisted" mode. I suggest you using "Flight Modes" on your transmitter to switch the "assisted" mode on and reduce aileron rate at the same time (the idea is to have less aileron movement on the stick than that available at the gyro for autostabilization).

B.- If you want to incorporate some roll sensibility (of no much use) to the gyro and reduce the powerful estabilization of the above setup, you may install de gyro 110º from the nose (this is its sensibility axis a bit rearward inclined to the tail). This way with just one gyro it will detect roll and yaw deviations. It is very important the gyro is mounted a bit rearward inclined or it will respond just in opposite ways to a same side yaw or roll turns.

Above explained has been not only tested but used operationally by many commercial RPVs and my knowledge is from personal experience during 11 years working on design, production, development and operation of target aircraft for the Spanish MoD and, in fact, this system is still used by the economical configuration of the most popular Spanish propeller driven target drone, mainly used for Mistral missile firing exercises.

Sorry is somebody already suggested this configuration on the thread as I have not read all replies. You may be sure this configuration I personally developed and tested about 10 years ago is totally effective and cheap.

Jesus Cardin

[MattChave]

Hi Jesus, thanks for the post.
Thats a very nice simple and cheap solution. How accurate do you think it can fly an aircraft to in degrees?
How much have your systems tended to drift just doing that? and how long would it take for that amount of drift?
Do the military control systems you've seen work in euler angles or quarternions? (or something else)
thanks matt

[jescardin]

Matt,

Thanks for your comments. Please understand we used the gyro just for STABILIZATION, not for navigation. During 10 years we used different makes and model gyros -all from RC manufacturers- but got best results with IKARUS ones. Last model used for 2004 and 2005 production runs was IKARUS Wing-Gyro model.
This brand produced no appreciable drifting -against some Futaba models- during our one hour missions and high South Spain high temperatures getting total stabilzation precission.
For navigation, and before adopting Micropilot MP2000 system about three yeras ago, we produced an intermediate RPV with gyro roll stabilization (as above explained), barometric altitude-hold and control and GPS navigation, but it rendered on human commands over a potentiometer to "veer" left or right based on course over a GPS navigation program on a computer. The pilot also demanded climbing and descend from other potentiometer commanding over the barometric unit aboard.
I hope this may answers your questions.

Jesus Cardin

P.S.: If you would like to see the models you may check de following links, and the RPVs and UAV are related as SCR and INTA. Aircraft names are ALBA and SCRAB.
http://www.micropilot.com/clients.htm
http://www.inta.es/doc/programasAltaTecnologia/avionesNoTripulados/INTA_T1_04a.pdf

[d_wheel]

Jesus,

I am very jealous of you! I have been experimenting with model aircraft autopilots for some time now, but only on a small scale. My dream job would be working for a company that does research and development on autopilot systems.

I have also used gyros successfully on the yaw axis. In a couple of cases, I controlled the rudder instead of ailerons. However, these models would roll almost as well using rudder alone as some others do using ailerons. One of the most important things I discovered is that you must design adverse yaw out of your model if you want it to fly and turn smoothly. I was never able to get one to work without using a microprocessor to control the rate of turn, and to help integrate pilot input while commanding a turn while in autonomous mode. One can be made to fly a short distance, even in heavy gusting conditions, with only a yaw gyro in control, but it is really not good for much. It can't be trusted to navigate over a long distance because it will change directions just a little each time a gust hits it. Using a GPS and microprocessor to correct it's heading from time to time makes it a genuine autopilot. Of course, it is necessary to be in control of altitude also!

I am presently experimenting with using GPS data alone to control altitude. Most have told me this will not be possible, but it is part of me that I must try it anyway.

You have a great day!

D.W.

[MattChave]

hey Ben/rclinks2002
how'd ya go in the end, did you come to a solution?

[jamesshaw89]

What a GREAT thread... thoroughly enjoying the read.

Jim.

[rclinks2002]

Matt,
I havent fully solved the problem yet, but I think the Autopilot from Maxxprod will work, It is essentially the same as the FMA co-pilot but doesnt incorporate IR and it costs 3 times as much... In the long run I would like to use a more compact version that has a lighter weight. Right now I am really only focusing on final exams and final projects....Thanks

Ben Mitchell

[kd7ost]

I have the BTA AS-07 from Maxx Products. Also the older AS-06. They are spendy but do work well. The -07 unit needs a few seconds cal time on the ground before you move the plane, but once it's done it does well. The -07 uses a piezo gyro and only draws 30ma during use. 3 times the Co-pilot current draw but as you stated, no issues with weather or temperature differences. It's also good in canyons or where the skyline rises around your flight area to as well as when it’s winter and the temperatures don’t work for the co-pilot. The co-pilot is a fair weather unit and that suits most flyers. The 06 unit uses a mechanical gyro and uses 100ma during ops.

I have noticed at full gain, the 06 and 07 roll a little bit more than the co-pilot on a full rudder turn but not much.

Also, just for edification, neither BTA unit likes to work with a 2 meter radio source on board. I used a small Alinco DJ-S11 radio sending Ham radio APRS packets and every time the radio keyed it nailed the roll and elevator full deflection. Most guys aren't mixing the two though so it shouldn't be an issue.

Dan

[jescardin]

Dan:
You are totally right regarding BTA autopilots. I also have experience for both the 06 and 07 models from the firm for which I worked. In fact we originally used the 06 on a propeller driven 150 mph top speed target aircraft and found it unsuitable for such speed as well as a defective product quality control as if some worked others drifted quite a lot making neccessary to continuosly trim the aircraft in flight.
From our needs, BTA developed the 07, but our boss was already dissapointed with the product and we returned the leveller to Avico, then BTA European distributor. Anyway, from our ground test we noticed the electronics were easily afected by radio transmisions (we used 400 Mhz for plane control).
On normal RC models we found it quite adequate, but sure we were dissapointed with the technological level as most regarded Israel as the low cost UAV most advanced country at that time, 10 years ago.

Jesus Cardin

[rclinks2002]

Do you know of another place where I can get a BTA-AS-07 other than Maxxprod? Cheaper? Also, dan have you opened it up and had a look at the inside. Can the case be reasonably stripped off to be made lighter? With regards to the 2 meter equipement, we will only be running 439MHZ and 900Mhz so it shouldn't be a problem. Thanks

Ben Mitchell

[jescardin]

Hi, Ben:

Here from Europe I do not know of anybody selling BTA products. The German European distributor was for professional/military use only and he gave up some years ago as he found there were not enough demand for running his bussiness.
In fact, originally BTA didn't offered its products to the modellers; it wasn't till they launched the BTA-AS-07 that they offered the AS-06 to the general public. I think that due the 07 was not a success to the professional UAVs market they sure decided selling it also to modellers.
Price for the AS-06 ten years ago here in Europe was arround 350$/each for orders of 10 or above.
I remember BTA also produced at least two other items they have never offered to modellers (as far as I know):

1.- A magnetometric device which conected to the AS-06 and which allowed navigation N-E-S-W and intermediate courses selected from a dial box which was conected to a customized transmitter. I do not remember the price of this item.

2.- Full autopilot allowing GPS and magnetometric navigation which consisted of two boxes of 12 x 5 x 3" size. The ground station was no so nice as moder ones as only showed the airplane aim and course but nothing of showing maps or something like that. The cost of each autopilot was arround 25.000$ and the autoleveling device was also an AS-06 integrated in one of the boxes.

Some 5 or 6 years ago, one of BTA owners, parted company and founded a new one which has it own web page, offering electronics and full RPVs and UAVs systems but I do not remember it. If I find again, I will post a link here.

Jesus Cardin

[jescardin]

For those interested, BTA new name is Wireless Avionics and its web page is:

http://www.cellrad.com/

Best Regards,

Jesus Cardin

[kd7ost]

Do you know of another place where I can get a BTA-AS-07 other than Maxxprod? Cheaper? Also, dan have you opened it up and had a look at the inside. Can the case be reasonably stripped off to be made lighter? With regards to the 2 meter equipement, we will only be running 439MHZ and 900Mhz so it shouldn't be a problem. Thanks Ben Mitchell

Ben,

I have opened up the 06 but not the 07. The 07 is smaller and uses the piezo gyro on roll so you might save a little weight if you took off the case. The 06 used the mechanical rate gyro. When you power up the 06 you can hear the high pitched whine of the gyro motor ramping up.

The 06 had a pretty odd implementation for pitch control. They used a small adjustable mechanical resistor inside a little sealed box with a rubber diaphragm over it. The center of the thin rubber was attached to an arm that looked like a straightened paper clip. When altitude changed it would expand or contract the air inside the box and cause the diaphragm to expand or contract thus moving the arm and changing the resistor value. It wasn't an altitude hold though, only pitch control. It allows gains or losses in altitude.

To be honest, I have no idea how the 07 does the pitch control. In a true UAV type design I would use a BP sensed true altitude lock like the UNAV pico Alt 3. That's what I use now. It will keep the plane where you set it at.

I don't know of any other source for the BTA units in the US. They actually showed up in RC Modeler ads for sale through a Hobby Shop in California back in about the mid 80's or so. At that time there was a version that also had true altitude hold but was about 500.00 or better. I bought one back then that is no longer around. The unit's disappeared for a while as the Hobby shop closed after the owner died. After what seems to have been a several year hiatus they showed up again at MAXXproducts. They are rather obscure products and a little big and spendy for hobby use. They are also pretty inadequate for professional use. Mine work just fine in 60 mph aircraft with 8 to 10 foot spans. I have that interference issue show up but not when I run 72 MHz RC, 6 meter RC from the ground, or 900 MHz or 2.4 GHz from onboard the plane.

Sorry I wasn't all that helpful; This information is more historical than valuable. I don't think removing the unit from the case would net you a large gain in weight savings.

Dan

[rclinks2002]

I found an as6 on ebay and I think I am going to get it, I guess the weight issue will just be a problem that I will have to work around. more info to come...

Ben

[Hovertime]

Great info in this thread, i was shopping around for my first gasoline engine recently, and wanted one with magneto ignition for same reasons as kd7lost , but ended up ordering electronic ignition engine .... :rolleyes: (Zenoa 20Ei)

This engine will be used for sport RC flying, so will have to do just fine for now.
I guess i will have to either come up with on-board generator/charger or get another engine somewhere down the road when I'm ready to try myself on a longer range project.
Dan - interesting info about BTA's are you still using it? What stops its gyros from drifting and steering airplane into the ground?

[awmeade]

Sorry to dig this one up guys, but my searches through the forum yielded little in the way of results on the BTA device.

I just aquired one for the princely sum of £20 through ebay, and looking through the documentation it looks like there was a plug-in to accept heading commands from a GPS unit - http://www.maxxprod.com/pdf/acc465-2005-2.pdf

Has anyone hacked their own GPS / WPS type instruction into one of these units? I see the GPS unit had serial comms as well on board for setting waypoints. Wish I could find one!

[rclinks2002]

I also have a BTA as06 which is the old version with a standard gyro. It weighs a considerable amount and it proved to be too heavy for my application. When I get back from Afghanistan I will probably buy a AS07.

Ben