General R/C Engineering - Always Under Construction
This page covers the engineering aspects of radio control and is primarily focussed on electronics and the human factors issues related to the design of radio control transmitters. It is also an opportunity for me to have my say publicly about what I think of the engineering "efforts" so far over the past 30 years without actually getting arrested !
Firstly, let's get down to the basics. ALL of the radio control systems on the market at present are domestic consumer grade electronics. They are designed and made like TV sets - I think that says it all.
When I fly a pattern ship I am in control of several thousand hard earned dollars worth of equipment splitting the air at over 200kph and weighing in at 5kg. This is a LETHAL weapon and in keeping with its status it is carrying 10 million dollars of public liability insurance. Oh, if only the insurance companies really knew what garbage is being used to control these missiles I really don't think you could afford the insurance.
If you are the type of modeller who builds ratty models on the basis that they are only going to crash anyway, or you buy the cheapest and nastiest equipment because these are only toys and don't deserve any more, then you can save yourself a lot of connect time by not bothering to read further. If you are professional in your approach to everything you do and you build models on the assumption that they are to give you 10 years of service, then I just may have something of interest to say to you.
I will cover some very controversial issues in this page and look forward to lots of mail and discussion about the topics. This is my web page and I can pick any size soapbox I want !!! Seriously though, I am a professional engineer and the stuff which I will cover here is based on sound engineering principles. There's no particular order to it and no fancy formatting, just a sequence of discussions on topics that I have been thinking carefully about for many, many years.
Transmitter mode
I feel safe from any lynching parties up here on my mountain so I am going to launch into one of the most controversial subjects in radio control, the mode 1 / mode 2 argument. Really it is not an argument at all. The control of an aircraft in flight requires control over the three axes of rotation of the aircraft, which are the longitudinal axis (roll), the transverse axis (pitch) and the normal axis (yaw). Due to the design of an aircraft and its inherent stability (static and/or dynamic), it is only necessary to have uninterrupted control over the pitch and roll axes. The yaw control is used primarily to balance the secondary effects of the other controls except during some specialized aerobatic manoeuvres.
Fundamental rule No. 1. If you do not have complete control over the pitch and roll axes then you are not in control of the flight path of the airplane.
If you had nothing else to do except fly the airplane then mode 1 would be ok. Real flying however, requires the manipulation of other controls such as trims and switches. With a mode 1 transmitter this requires relinquishing at least one of the primary control axes, or at the very best having only limited control over it. With a mode 2 transmitter, the left hand is always free to roam and operate the less important controls while the right hand maintains full control over the pitch and roll axes.
There are many arguments against mode 2 some of which I will address here. In my experience it is only mode 1 flyers who make issue of the mode debate, somehow attempting to justify the inefficient means by which they control their aircraft. Mode 2 flyers just seem to get on with flying.
Argument 1.
The human hand is not designed to pull a stick straight back without introducing some sideways movement. True. JR's solution was to produce gimbals which could be rotated so the fore and aft stick motion was aligned with the direction of thumb movement and this solution worked well. A better solution is to look at the whole issue of controlling the sticks with the thumbs. What we are really doing is controlling a high speed aircraft with our thumbs and at the same time holding the transmitter, using the same hands which are providing the fine control. This simply does not make any sense. The correct solution is to support the transmitter from some other part of the body, say in a tray, and manipulate the sticks using the thumb and forefinger, preferable with long sticks. The problem of hand geometry is immediately eliminated. The stick is pushed forward with the thumb and pulled back with the forefinger and it is pushed or pulled from side to side using both thumb and forefinger. Fine, linear control with no mixing is easy to achieve with a little practice.
I suggest that it is about time radio flyers opened their eyes and looked around them at the real world. The facts speak for themselves. Almost all of the top flyers in the world today fly mode 2 with long sticks and a transmitter tray. The reason we are trapped in the mode 1 camp is purely a result of commercial pressures and greed. In the old days when nobody could afford a radio with more than 2 channels mode 1 was all there was. Naturally, two channels, two sticks. The dealers pushed the radios on modellers and into the training systems in the clubs, then they pushed more sophisticated radios which of course had to be mode 1 because that's what their client base had been trained on. Unfortunately, the dealers continue to push the myths about the mode debate in order to maintain continued sales of radios.
Argument 2.
The brain cannot control two functions accurately together on the one stick. Don't laugh, this is a serious argument and I here it almost every time the mode debate is raised. Excuse me while I pick myself up from the floor, ha ha. Who are these people trying to kid. From a physiological perspective, the brain is actually more capable of controlling two functions on one stick because the functions are not independent, rather they are very closely inter-related. The major control processing burden therefore falls to only one hemisphere of the brain and the left/right internal communications pathways are not needed. I am not about to labour this issue and the reader can find much information from the large number of reference texts available. Suffice to say modern fighters do NOT have two sticks, one for pitch and one for roll.
The throttle trim
The most useless and least used control on the modern transmitter is the throttle trim. Sure, everybody you ask says they use it all the time, and the pattern flyers will tell you it's essential and they can't do without it. What a load of utter hogwash. A transmitter with a computer encoder must devote one of its most precious resources to the throttle trim, an analog to digital encoder channel, and for what ? So that it can be used as nothing more than a glorified stop switch, because that is all most flyers use the throttle trim for.
Firstly, from half throttle onwards you can forget the trim because it has no measurable effect, a consequence of the design of the carburettors used on model aero engines. So the throttle trim only effects the lower range of throttle operation. There is no reason to use it in flight. If a different throttle setting is required then the throttle stick should be moved, not the trim. In fact a trim really only makes any technical sense when it is associated with a control which returns to centre since its function is to set-up any offset for the centre position. In the case of the throttle there is no centre position and no return to centre so the concept of a trim is meaningless. The primary use of the throttle trim is to enable a high and low idle setting to be established, normally with the low idle setting actually shutting down the engine. This function should be assigned to a switch, two position for mortals who just need high and low idle, or three position for pattern jocks who think they need three idle positions. Switch resources on a computer encoder are much less precious than analog resources.
So what do we do with what was the throttle trim. Obvious, it becomes the mixture control and remember, forward for rich, back for lean, just like the bigguns. Truly, I met a guy who claimed the correct set-up was forward for lean. His reasoning, wait for it, forward means go faster !! He's right, lean it out it will go faster. Until it quits or seizes, then it will go very slowly indeed. Unless you're a bulldozer driver, stick with the standards. Push it forward to give it more juice. Of course not everyone wants mixture control. Well that's their funeral. Mixture control should be used on all model aero engines and it should be coupled to the throttle through software which measures throttle position and speed and direction of throttle stick movement. These inputs should be used by the software to compute the optimum mixture setting at that instant with the mixture input setting the desired deviation from the computed value. The Eagle Air computer encoder does this and we consider this the correct solution.
Control trims and other controls
The standard layout of controls and trims on radio control transmitters is a human factors engineers worst nightmare come true. Just because the first transmitter ever made had the trims and controls in a certain position does not mean that such positions are correct or that the next 10 million transmitters should be built that way. This discussion will cover the correct trim and control locations for a mode 2 transmitter.
The concept of mode 2 is maintenance of complete control over the primary pitch and roll axes using the right hand while the free left hand manipulates the ancillary controls, which includes the trims. Note that the correct operation of trims presupposes mode 2 operation. In full size aircraft operation, the appropriate control force is held and the trim adjusted to null out the control force. In the case of models the procedure is the same except that the control force being held is the simulated force generated in the pilot's brain, a consequence of learning to fly radio controlled airplanes. Of course, not everyone does it this way, but we are only concerned here with flyers with the professional attitude and the will to operate airplanes properly.
Assuming the above, then the traditional location of the trims just doesn't cut it in the real world. The cross over of the transmitter by the left hand to adjust the pitch and roll trims is awkward, risks moving other controls such as flap levers, and interferes with the operation of the primary control stick. Furthermore, modern Japanese radios have so many humps and bumps and simulated plastic cap screws that it is nearly impossible during the crossover to even identify what is the trim lever without having to look down at the transmitter. Note this carefully. If you are not looking at your model then you are out of control, by definition.....
The correct location for the trims is on the left hand side of the transmitter. Pitch trim to the far left beside the throttle lever, roll trim below the throttle lever and yaw trim above the throttle lever. The mixture lever fills the remaining position to the right of the throttle lever, as per full size practice. In these positions the trims are readily accessible during flight and their location is easily found blind by reference to the edges of the transmitter case.
Carrying the mode 2 concept to its logical conclusion, then in order to be in control of the airplane at all times it is necessary to never relinquish control over the primary control stick, and thus there should not be any other controls on the right hand side of the transmitter other than the pitch roll stick. Thus we end up with a transmitter with all of its controls on the left hand side with the exception of one gimbal. This is correct and makes sense. The problem is the meter, aerial and power switch which use up a considerable amount of space in the centre of the transmitter. Well the fact is they are only there because that is where they were on the first ever transmitter and nobody ever thought about putting them anywhere else.
The meter and the power switch should be on the right hand side, preferably above the right hand stick. In fact the best place for both is behind a panel. Remember, the meter is not used in flight and if it is looked at then you are not in control of the airplane. All controls and indicators not actually used in flight should not be present within the pilot's working environment. In other words, if it is not going to be used in flight then it should be hidden behind a panel. I'm not talking about Battlestar Galactica clones designed to impress the shop window browsers. Designs like that are just toys. I'm talking about real transmitters designed to control lethal pieces of very expensive equipment and protect you from a full frontal assault by the legal profession.
With the meter and power switch gone, the transmitter centre section is available for controls so now we have two thirds of the transmitter available for ancillary controls in an uncluttered and easily accessible arrangement and the right hand one third of the case containing nothing but the primary pitch and roll control.
The best location for the aerial is vertical straight out of the top of the pilot's head. This is where engineering and the softer sciences overlap. If you implemented the ideal engineering solution for aerial placement you would probably have a lot of difficulty integrating socially into your club and at best you would be considered eccentric. Someone, sometime, somewhere, while you are not looking would screw a propeller to the end of your aerial !!!
Since the mode 2 transmitter is ideally operated from a tray then lateral balance is not a serious concern and so the aerial should be mounted out the top of the right hand side of the transmitter, angled so as to point vertically during normal operations. The vertical aerial orientation suits the radiation pattern for aircraft flying within an arc extending from roughly 30 degrees above the horizon to at least 60 degrees. If the airplane is below 30 degrees then it would be close in so range is not a problem. If it is above 60 degrees and you lose range then all you have to do is wait.
The ideal aerial does not interfere with the pilot's vision, unduly effect the balance of the transmitter or create unnecessary windage. If you can tolerate some of the technical limitations and oddities a small 1/8 wave centre loaded rod with an adjustable capacitive hat and a BNC connector as its base meet the requirements perfectly. These are really easy to make and tuning is a snip with a GDO. Just make sure you have the rf output stage retuned to suit this type of aerial.
The remaining controls will mainly be switches and buttons. Forget colour coding. If you have to look at the controls then you are a menace to everybody and have no business flying radio controlled airplanes. All controls should be identifiable by feel and should be located according to their function and frequency of use. For example the left snap roll button on the left and the right snap roll button on the right. This may seem obvious and it is but there are other less obvious arrangement options which require discussion.
Controls which set such things as depth of mixing should not even be present. These parameters are set up as software data inputs although switches to select certain preset values are acceptable. The retract switch should be positioned to enable operation with a spare left finger from the full throttle position. It should be close to the centreline of the yaw control since at full throttle only small amounts of yaw input will be used. The positioning should be optimised for the retraction cycle at full throttle because this is the most critical time when the airplane is at full power close to the ground. The extension cycle is normally done during a relatively leisurely phase of flight so control location is not as important. The idle up switch can be anywhere since it only operated on the ground. Snap rolls are best initiated with buttons and not switches. A switch should be used to pre-select the snap roll mode, either upright or inverted and two buttons initiate either left or right hand snap rolls. Analog controls such as flap levers should be stiff with a very solid ratchet to provide good user tactile feedback. The provision of finders such as on the Multiplex sets is a very good idea indeed.
Based on the above reasoning, the resulting transmitter design would be very, very clean. A flat plate front panel with no unnecessary bumps to provide the pilot with aliases and the only things visible being the controls actually used in flight. Graphic display, data entry keypad, power switch and meter if any hidden behind a panel. In keeping with this concept, the identification of controls by feel should be supplemented by an audible identification responding solely to a touch on the control. Maybe a sexy Texas female voice is going a bit overboard, but certainly some whoops and bells such as heard on the flight deck of modern aircraft would be appropriate for some of the more critical ancillary controls.
Transmitter complexity
Modern computer controlled transmitters are very sophisticated pieces of equipment indeed. As an engineer the very first question I ask is "why ?". If I can't get a satisfactory answer to this basic question then I find it necessary to re-evaluate the entire transmitter concept. As it is, there isn't a satisfactory answer. Let's have a decent "down to earth" look at the r/c transmitter from an engineering perspective (not a salesman's view).
What is the basic function of the transmitter ? The transmitter's function is to gather user input from various controls to produce information which is modulated onto an rf carrier and transmitted to the equipment being controlled. Simple, if it did nothing else then it would be fulfilling it's function. This is all very obvious, so why have most of the manufacturers simply "lost the plot".
There are many aspects of controlling "real" models, such as aeroplanes and boats, which require some transformation of the user input prior to being assembled into the information package to be transmitted. Such transformations will require some control over the way in which they are to occur to account for short term changes in the operating conditions of the model.
In the modern transmitter, the actual transformations and assembling of the information package is best handled by a microprocessor, and in fact this is just what microprocessors are best at doing. The control over these transformations is managed by a user interface which is essentially just a means by which the user can plug numbers into the processes. From now on I will use the correct terms which are as follows. The processes which determine the transformations are called the programs, and specifically for r/c transmitters, the model descriptions. The input from the controls and the user interface is called the data. It's very simple really. We take data and feed it into a program and information comes out the other end which we then send to the model. What happens to it then is not of concern for this discussion.
So what is the basic problem with modern transmitters ? Simple. The users have been given ever more ability to input data with a consequent increase in the complexity of the user interface but have NEVER been given the ability to alter the programs, only select programs from a palette of options. With manufacturers developing increasingly complex programs the complexity of data entry has risen to the point where the user needs to, in many cases, buy specialised books in order to learn how to use the transmitter. What is even worse is that many of the features of the programs are redundant for particular models and are never used, but the complicated operations still remain. These transmitters (all of them) are advertised as "programmable" by the manufacturers. They are not programmable and it is impossible to alter the programs. The only thing that the user can alter is the data which the programs use.
So I ask myself has the implementation of modern technology in r/c transmitters really improved my lifestyle, or at worst made it possible for me to control r/c aeroplanes better than before. The answer is no. The transmitters are so complex to operate that I need to buy books on how to use them and they still do not do some of the things that I really need to do whilst providing features that I am unlikely ever to use. I am trapped into model descriptions (programs) written by engineers who flew their first r/c aeroplane after they started work for the manufacturer. Without 30 years of experience they rely on input from "consultants", people who have grown up on the very technology they are consulting on.
What is the solution ? First we realise that the place for complex and sophisticated user interfaces is at home within an environment where time is not a primary concern and reference material is readily available. Second, we make available a means by which the processing of data can be altered to suit the application thus reducing the amount of data needed to only that which is absolutely necessary for the specific application. Thirdly, we arrange that the user defines the data requirements of the programs specific to the application and so eliminate the requirement to learn how to operate the transmitter (you do not have to learn what you yourself defined). This solution will yield a very simple transmitter which is able to outperform ANY transmitter currently available. It will not look pretty, in fact it will be very plain. The beauty and elegance of good engineering design is that a simple and functional exterior hides a sophisticated interior, and in the solution which Eagle Air has come up with, the sophisticated interior is mostly software contained within simple, almost trivial, hardware.
To accomplish the above it is necessary to design a transmitter which can be programmed by the user. The other important engineering goal is that it is cheap to manufacture. Using the very latest technology this is possible. The following design criteria need to be met;
The microcontroller based encoder developed by Eagle Air meets all of the above criteria and details of this encoder and its operation are available on this web site.
Copyright © 1995-1999. Eagle Air Australia Pty.Ltd. All Rights Reserved.