model flying saucer - help/advice needed.
Nicholas Outram
Hello,
I am not a regular reader of sci.electronics but would like
to ask your opinion on a rather large project I am about to start.
My project (I'm not mad really !), is to build a flying saucer...
Well, it will be a model flying saucer about a metre in diameter, with
four model engines keeping it up.
it would look something like this (ascii does not do justice!);
___
TOP VIEW : / O \ where 'O' is an engine.
|O + O|
\ O /
---
Has anyone made one of these things before ? What difficulties where
encountered ?
I have many ideas about the control and powering of this model and
would like to ask for any comments you may have on the suitability
of the ideas. Such ideas are;
Power.
------
1. Use RC variable speed engines (this could be difficult to control
needing seperate servos etc.
2. Use electric motors and supply the power through wires to the craft,
as I doubt it could lift the power supply too (easy to alter the
speed, but will there be enough power to lift the device ?)
Control.
--------
3. Build a complete R-C interface including onboard micro to control
the servos and/or system in 5 below.
4. Use standard 4 function RC. (How do I then control the total
power to the engines if I use 1 above without going to the
much more expensive >4 channel RC systems ?)
5. Build a 2 axis stabilization system to keep the craft level in the
x and z axis. Some thoughts on this are: use a gyroscope (maybe
heavy and overcomplicated). Use 4 height measuring devices, use
an upside down joystick with a weight attached. This device would
have to be light to keep the all-up weight low.
The device could only be used if I use 3 above.
I have experiance of electronics but have never built a standalone
microcontroller. I also have plenty of high and low-level programming
experiance.
Thanks in advance for any hints tips on RC and suggestions for
controler chips etc.
Nick Outram.
David J Bell
>Control.
>--------
>5. Build a 2 axis stabilization system to keep the craft level in the
> x and z axis. Some thoughts on this are: use a gyroscope (maybe
> heavy and overcomplicated). Use 4 height measuring devices, use
> an upside down joystick with a weight attached. This device would
> have to be light to keep the all-up weight low.
Interesting approach; probably the only really practical way to handle
controlled flight in a system as inherently unstable as this would be.
Conventional remote control of unstable aircraft is usually chancy, at
best, and disasterous in many cases. A gyro (or a pair of them, for
pitch and roll axes) would be neat, but expensive, heavy, and probably
more accurate than needed; you'd need only rough control of attitude,
nut couldn't tolerate much accumulated error - like an inexpensive,
drifting gyro! A weighted joystick would respond to acceleration as
well as to gravity (or so Einstein told us...). This would make the
platform dip down in the "leading" direction whenever you poured on
the power...
A different method I read about quite a number of years ago may work:
This works best in normal flight, not ground-approach, but that can be
overcome either by reverting to manual control, or adding some filtering
to the control electronics. In it's simplest form this was used to
stabilize a model plane the investigators were using to make temperature
and humidity measurements at controlled altitudes; it was too hard to
control a tiny model at long range, and still get stable telemetry. They
came up with a scheme to take advantage of the Earth's *electrostatic*
field, which can be on the order of hundreds of volts per meter change
in altitude. (A LOT more, under a storm cloud, but more on that later!)
It seems that if one places some sort of "air contact" on each wing tip,
then connects them to a high input impedance differential amplifier, the
output of the amp very neatly corresponds to the roll angle of the
aircraft. Actually, the error voltage approximates the sine of the roll
angle, times the wingspread... Feeding this error signal back, summed
into the aileron control command coming from the receiver made the
plane inherently stable on the roll axis, as it would level it's wings
in the absence of a command from the ground. Similarly, another set
of "contacts" could be put fore-and-aft, gaining pitch stability. In
practice, they took the average of the two wing-tip inputs and compared
that to the input from a tail electrode. They further refined the
pitch controller by raising the tail electrode to the top of the rudder
column. This gave them the *very* nice result of providing almost
automatic coordinated turns, with the plane banking and holding the
nose almost level, with only one control input! The electrodes took
some experimentation, but as I recall, they ended up using the little
Polonium-impregnated strip of metal taken from an anti-static brush
for phonograph records - remember vinyl disks?? :{) The Polonium
emits a charged particle (Beta, isn't it? An electron?), and builds
up a charge of the opposite polarity. This charge dissipates at a rate
dependant on the humidity, but also on the atmospheric electrostatic
potential. The amplifiers were common IC op-amps, LM709's, I think.
Later experiments were planned to test the feasibility of the control
system on a private plane, in the Piper Cub class. There were a few
problems, though. I feel they could be removed today, by judicious
software filtering of the data. One was when flying very low, as in
a landing approach; any object that sticks up a significant fraction
of the plane's altitude - like a fence or even a stalk of wild grass -
causes a "bump" in the iso-potential curve in the field. This sometimes
caused unexpected swerves or drastic tilts that couldn't be corrected
before unintended touchdown. As in crash... This was solved by using
an extra control channel to switch off the flight control system
during take-off and landing. Another glitch appeared the first time they
flew the test plane under a low thunder cloud. It seems that the normal
potential difference between ground and upper atmosphere *reverses*
under a cloud that is generating a high potential field of it's own.
First, the plane got a little unstable, as the field around the very
edge of the cloud cancelled out the normal field - no correction
input, no automatic control. Very soon, though, the plane was under
the cloud itself; with the field reversed, the damned thing flipped
over and flew nice and steadily - UPSIDE DOWN! As it left the cloud,
it flipped right back up and continued flying normally. This worked
alright for them, but they decide it needed a bit more work before
it was applied to a plane with passengers in it!
This little gizmo - they called it an "Electrostatic Autopilot" would
seem to be ideal for a flying platform. Think about it a while - it
just might work!
Dave db...@cup.portal.com
Mark Robert Thorson
stepper linked to the throttle.
One possibility would be to use pulse-jet engines, in which a computer
controlled when the spark was delivered to the combustion chamber.
By ramping the pulse rate up and down, you might achieve the same effect.
Because the suction of the exhaust gases draws the fuel in, it would
be self-throttling.
(On the other hand, this decision would probably turn your project from
"model flying saucer project" to "engine development project".)
Jeff Forbes
>Because RC engines are diesels, they must be controlled by a servo or
>stepper linked to the throttle.
4-stroke model airplane engines have been available for several years. The
are becoming more popular since they sound more like real airplane engines
and have a better power band.
Jeff
Mike Harris
(I believe). I bought mine about 5 years ago. Never quite finnished it before
I got into Schluter RC helicopters.
It has a central pod with a motor. Has a free spinning foam outer ring
(the ufo part). Has a vertical "tilt" flap which tilts it, and a "rudder"
flap which turns it. The rotating outer ring automatically stabilizes it -
no need for another gyro! Takes a '60 engine with three channel radio.
Sell you my kit for a song and a dance.
Mike Harris - KM4UL har...@dg-rtp.dg.com
Data General Corporation {world}!mcnc!rti!dg-rtp!harrism
Research Triangle Park, NC
Avatar
+stepper linked to the throttle.
And because of this they are very easy to control!
+
+One possibility would be to use pulse-jet engines, in which a computer
+controlled when the spark was delivered to the combustion chamber.
+By ramping the pulse rate up and down, you might achieve the same effect.
+Because the suction of the exhaust gases draws the fuel in, it would
+be self-throttling.
+
Nope. The only way to throttle a common pulse jet, like a Dynajet, is to
put a clamshell thrust deflector into the exhaust stream! You cannot vary
the amount of fuel being delivered; it will simply cough and die! The same
goes for choking the air delivery! Once the engine coughs, it's over! A
restart requires compressed air thru the "Flowjector's" nozzle and a steady
spark until ignition is achieved. Once the engine starts, it goes into a
resonant-burst combustion cycle, the frequency of which is between 180 and
240 Hz. This is frequency is determined by the geometry and size of the
engine, length of the tail pipe, the shape and number of the head's intake
petal-valves and a dozen other problematic variables. In short, it would
be easier to use the common 2-stroke model aeroengines, with a ducted fan
to garner the desired thrust....much easier!
(ps-any other model turbine/turbojet engine builders out there?!)
--
-Avatar-> (aka: Erik K. Sorgatz) KB6LUY +-------------------------+
Citicorp(+)TTI *----------> panic trap; type = N+1 *
3100 Ocean Park Blvd. Santa Monica, CA 90405 +-------------------------+
{csun,philabs,psivax,pyramid,quad1,rdlvax,retix}!ttidca!sorgatz **
David Svoboda
>Power.
>------
>1. Use RC variable speed engines (this could be difficult to control
> needing seperate servos etc.
More like impossible to control. While R/C engines can provide ample
power for a project such as this, modulating the power for control is
not practicable. The response time for pitch and roll control would
need to be nearly instantaneous (ask any R/C helicopter entusiast about
that) and you just cannot vary the power that quickly on an R/C engine.
>2. Use electric motors and supply the power through wires to the craft,
> as I doubt it could lift the power supply too (easy to alter the
> speed, but will there be enough power to lift the device ?)
Same argument. An electric motor that provides POWER is necessarily
either heavily armatured or has very high speed, either of which precludes
quick speed changes.
A possible solution to this might be a set of variable pitch propellers,
which would function on the same principal as the collective pitch on
a helicopter. Years ago these propellers were sold for R/C aerobatics
purposes, but were abandoned due to the hazards of blade separation on
high powered engines. My suggestion, if you could find any of these,
would be to use them on the lower powered electric engines. And BTW,
if you are running wires from the ground for power, why don't you just
have a digital hard link, and do away with the radio.
Or you could make some sort of "air valving" system, which would constrict
the propeller inlet or output by way of some sort of shutter system.
>Control.
>--------
>3. Build a complete R-C interface including onboard micro to control
> the servos and/or system in 5 below.
Yuck. You might just want to build the device I describe below.
>4. Use standard 4 function RC. (How do I then control the total
> power to the engines if I use 1 above without going to the
> much more expensive >4 channel RC systems ?)
You need three channels. Pitch, roll, and lift. You would want to
use one servo per power unit (assuming power modulation is used by way
of cyclic pitch or air valving) and use an electronic mixer to mix
the control inputs to the proper control outputs. The functionality of
this device would be as follows:
Cyclic stick fwd: Increase (Both)rear thrust, decrease (Both) front
Cyclic stick rght: Increase (Both) left thr, decrease right
Lift stick fwd: Increase all trust
Of course the inverse of each of these.
These electronic mixers have been around for about 15 years, used to
control flying wings (elevons) and V-tail aircraft (rudder-vators). I
leave it to your ingenuity to design it. If you need help, refer to
R/C Modeler magazine, mid-70's, construction article for Cristy Mixer.
>5. Build a 2 axis stabilization system to keep the craft level in the
> x and z axis. Some thoughts on this are: use a gyroscope (maybe
> heavy and overcomplicated). Use 4 height measuring devices, use
> an upside down joystick with a weight attached. This device would
> have to be light to keep the all-up weight low.
> The device could only be used if I use 3 above.
Gyro is possible. Many rate gyros are available which plug directly
into one of the channels on a radio. These are used for tail rotor
stabilization on R/C helicopters. These are just for stabilization,
however, and the craft must still be flown manually.
Height measuring: Maynard Hill, holder of many R/C duration, distance,
etc. records, uses a homebrew electrostatic wing leveler. I once saw
it for sale in a model magazine somewhere. As I remember, certain
atmospheric conditions upset it. (But who flys in thunderstorms,
anyway? :-)
Topsy-turvy joystick: Oops. I forsee a spiral crash as the pendulum
swings round and round, thinking it's perfectly level...
From: m...@cup.portal.com (Mark Robert Thorson)
>Because RC engines are diesels, they must be controlled by a servo or
>stepper linked to the throttle.
Semi-diesels. Glow plug.
>One possibility would be to use pulse-jet engines, in which a computer
>controlled when the spark was delivered to the combustion chamber.
>By ramping the pulse rate up and down, you might achieve the same effect.
Well, my dynajet needs no spark, once it's running. The exhaust tube
turns cherry red and keeps combustion free running. Also, can you say
LOUD? I'd hate to think of four of them. Expensive too, and not too
powerful. Neat curiosity though.
From: for...@aries.scs.uiuc.edu (Jeff Forbes)
>4-stroke model airplane engines have been available for several years. The
>are becoming more popular since they sound more like real airplane engines
>and have a better power band.
Less HP/displacement and HP/weight, though. That's why you don't see
4-stroke helicopter models.
From: har...@aquila.rtp.dg.com (Mike Harris)
>Look into a model called the "TURBOPLANE". Produced by an Austrian company
That's TURBOPLAN, isn't it?
>Sell you my kit for a song and a dance.
If no other offers, please call me. :-)
From: sor...@ttidca.TTI.COM ( Avatar)
>be easier to use the common 2-stroke model aeroengines, with a ducted fan
>to garner the desired thrust....much easier!
Big can of worms. And reeeaaal expensive.
When you buy a radio, be sure to get one on an aircraft frequency. Also,
narrow band for 1991.
Good luck!
Dave Svoboda, Motorola CID, RTSG, 1510 W Shure Dr., Arlington Heights, IL
uucp => {uunet|mcdchg|gatech|att}!motcid!svoboda 60004
internet => motcid!svo...@chg.mcd.mot.com Ph (708) 632-2406
Don't listen to me, I'm just a puppet of individuality.
Darshan V. Bhatia
>
>NO NO NO! Whether two or four stroke, model airplane/car/boat/etc engines ARE
>NOT DIESEL, they are 'glow ignition' NOT 'compression ignition'. That little
>filament that lights up when you connect the battery *stays* hot when the
>engine is running and provides the source of ignition. The metal in the
>filament is selected for its ability to catalyze the combustion. Sorry about
>the digression, this isn't really a s.e topic; but I saw the need to dispell
>some false info...
>
>Walt K.
--------------------------------------------------------------------------------
I beg to differ. While most RC engines are glow plug, I've had model airplanes
that were indeed COMPRESSION-IGNITION. These had NO glow plugs. 'FROG' in
England made these engines and 'AURORA' in India also made the same. I don't
know of other companies but I'm willing to bet they were made elsewhere too.
These engines ran on fuel that had the following composition:
a) 50% ETHER
b) 25% KEROSENE
c) 25% CASTOR OIL (or SAE 50 oil)
I could never get them to run with SAE-50 oil. Castor oil was the only thing
that I could use successfully. They used to have a little screw that went into
the combustion chamber which allowed one to change the compression ratio.
-darshan
Walter A. Koziarz
>In article <29...@cup.portal.com> m...@cup.portal.com (Mark Robert Thorson) writes:
>>Because RC engines are diesels, they must be controlled by a servo or
>
NO NO NO! Whether two or four stroke, model airplane/car/boat/etc engines ARE