A Programmable Dynamic Attitude-Aware Motor Speed Control
for Electric-Powered Control Line Aircraft
|There is currently a big shift from internal combustion engines to electric motors for powering model vehicles of all sorts - airplanes, helicopters, boats, and cars - and of all control modes - autonomous (free flight), radio control, and control-line. The state of motor and battery technology has passed the point where the weight and thrust available with electric power meets or exceeds that of engines for most applications. Costs are pretty much at parity as well when you compare engine vs. motor and fuel vs. battery acquisition and cost of ownership over the life of the power system.|
All sorts of useful electronic peripheral equipment has been developed for use with electric motor power: programmable electronic speed controllers, motor cutoffs based on altitude and/or elapsed time for free flight, motor timer/speed controls for control line, and even engine noise generators to give life-like sound to otherwise eerily quiet war birds and commercial transports, to name a few. These devices had made the switch to electric power nearly seamless for most flyers. There is probably little demand for a spray bottle of burnt fuel residue for coating the model after a session, but I personally would like an of Eau du Running Fox 35 air freshener just for the sake of nostalgia.
As a life-long control line model airplane flyer with only sometimes moderate aerobatic success, I have followed the writings and videos of modelers who have mastered the arts of building, building, flying, adjusting, and even repairing contest-grade stunt machines. I remember being in awe the first time I heard a perfectly-adjusted engine break into a high-speed 2-cycle mode when the model pulled upward from horizontal, settled back into a richer 4-cycle mode on a downward path and back to horizontal, and then repeated the sequence maneuver after maneuver. The practice is commonplace today, but I have to wonder who was the guy who first decided to exploit the phenomenon to make a more perfect flight?
The 2- vs. 4-cycle alternation is due, per my understanding, to the gravitational effect on the fuel flow rate to the carburetor when the engine and fuel tank are in various positions. In a nose-high attitude the fuel in the tank and in the fuel line are lower with respect to the carburetor than when in a horizontal position, so more suction (vacuum) is needed in the line to provide the same rate of fuel flow. For a given engine speed (RPM) the suction is constant, so the greater weight of fuel in the line does not flow as quickly, resulting in a leaner (faster) engine run. As the engine increases in RPM the suction will increase accordingly, providing more suction to counter the weight of the fuel. Depending on where the fuel mixture is set, that could cause the engine to oscillate between 2-cycle and 4-cycle operation in that transition area. However, because everything is happening so fast, there is no time for that to occur. Besides, all maneuvers are actually flown in an increasingly upward arc - even the triangle when viewed in the 3-dimensional path it is flying on the surface of a hemisphere - so the fuel weight is continually increasing as the angle of flight (not necessarily angle of attack) is continually increasing. As the model transitions from vertical back to horizontal at the top of the maneuver, the weight of the fuel is returned to normal so the engine kicks back into the 2- and 4-cycle threshold realm. Then, as the nose of the model goes down, the fuel mixture is richened additionally and a slow-running engine is the result. The whole process helps keep the airspeed of the airplane more constant throughout the maneuvers.
So, with the advent of electric propulsion that easily provides sufficient thrust for the entire stunt aerobatic pattern, the control line community is now living with a constant thrust output at all flight attitudes from the power plant. I have not read in, for instance, Bob Hunt's column in Model Aviation whether this is a problem. Masters like Bob can no doubt adapt to any situation, but then again electric-powered models have not been running away with all the top prize positions. Maybe the absence of automatic variable thrust is one of the reasons. I should write to Mr. Hunt and ask him, but maybe someone will read this article and ask him for me. My question is why hasn't someone designed such a motor speed control yet that detects the models attitude and adjust the motor speed accordingly?
I'm throwing this idea out to companies like Winged Shadow Systems, who make some ingenuous peripheral products like the How High altimeter and the Thermal Scout thermal detector, and the Sky Limit altitude/time limit motor cutoffs. Surely those guys can design and affordably market a dynamic, attitude-aware motor control for electric-powered control line airplanes. I provide here a basic outline of the concept, what I title "A Programmable Dynamic Attitude-Aware Motor Speed Control for Electric-Powered Aircraft©." While its indented initial application is for control line aircraft, it is possible to extend the usage to free flight and other modes of flight.
The first thing that would be needed is a device capable of sensing the airplane's attitude; i.e., whether it is climbing, diving, or in level flight. That information would suffice to command two or three three unique motor speeds, which could easily be programmed by the user in terms of percentage of full speed or as an absolute number. The controller would convert the speed data into the pulse train format required by readily available electronic speed controls (ESCs). A simple mercury-filled or electrolytic tilt sensor can do the job for providing two or three discrete motor speeds. If you want a continuously variable speed controller that can respond to all attitudes throughout a 360° rotation (rather than just two or three), a more sophisticated sensing device is required (an integrated circuit, often based on MEMS). Any attitude sensor will need to have a response time fast enough to keep up with the model's flying speed and be able to operate properly even under conditions other than an acceleration of 1 G toward Earth's center (the direction we normally experience).
A good place to start looking for motions and position sensing gear is the robotics websites. Robotics hobbyists have found all sorts of nifty parts for sensing and controlling their platforms. The radio controlled multicopters (quad, penta, hexa, etc.) that are extremely popular now are chock full of low-cost, high accuracy sensors for detecting tilt and acceleration on all three axis. ROHM Semiconductor is well-known for its sensors and happens to manufacture the RPI-1040 ($2.29 from Digi-Key, half that in large quantities), a 4-direction detector that measures 3.1 x 3.1 x 0.8 mm, weighs next to nothing, and has a 10 µs response time. Because the flight path is always on the surface of a 3-dimensional hemisphere, two orthogonally-mounted tilt sensors would likely be required with a microprocessor algorithm that receives an orientation fix at power-up with the model sitting on the ground with wings level. The conceptual block diagram below outline my idea.
Please contact me if you are interested in pursuing the design and testing; I do not have the time or equipment necessary to tackle it on my own.
Here is the text of the above conceptual drawing:
Programmable Dynamic Attitude-Aware Motor Speed Control for Electric-Powered Control Line Aircraft©
Kirt Blattenberger -- Copyright December 30, 2012
This concept for a programmable dynamic attitude-aware motor speed control for electric-powered aircraft uses one or more sensors to detect and report the aircraft's 3-dimensional attitude to a microcontroller, which subsequently calculates an appropriate rotational speed for the motor (rotations per minute, e.g.) and sends an appropriately formatted pulse train to the external electronic speed control (ESC), although an on-board ESC may be incorporated. Motor speed settings may take the form of an absolute number or a percentage between 0 and 100%. The user programming interface may consist of a set of onboard switches and/or an off-board serial electrical interface and/or parallel electrical interface. An ability to store and recall one or more programs should be provided without requiring an external power source to be connected. Although any response to the tilt sensors may be programmed, the primary intention is to detect whether the aircraft is in horizontal flight or whether it is climbing or diving, and then command the motor’s speed accordingly in order to effect a flight speed throughout a maneuver that is more consistent than would be with the motor running at a constant speed. The process would attempt to emulate an internal combustion engine's ability to switch between 2-cycle and 4-cycle modes (thereby changing engine RPMs) when assuming various attitudes, although other uses are possible.
Posted December 30, 2012