This is the first, and I hope not the last issue of a new column for the EZone, "The Inside Story". My name is Graham Stabler, and I will be acting as a kind of make shift editor, persuader, and compiler for the column. The column came about after longing for the days of Wayne's Indoor World. I asked Dave Lilley, the EZone's editor, if we would see something like it again. Todd Long had begun his Light Stuff column, but had to stop due to other commitments. As Dave explained to me, this was the basic problem. It is incredibly difficult to find someone to take on the burden of a monthly column. I considered the task, but could see the day when my personal stock of knowledge and material had dried up coming all too soon. Then there is the time involved. I am a PhD student, and writing up will loom all too soon. It turned out Dave had been thinking about an indoor column himself, and a request for a writer appeared in that month's control tower.
Now, if you are not confused, you should be. Why is this crackpot writing a column he cannot write and does not have time to write? Well, I am writing this column, but not on my own. I managed to fool, err, persuade some likely modelers to do most (they might not know that yet) of the work. There are about 12 of us. Some are famous and some are infamous, but all sharing an interest in micro/indoor models. We have experts in fixed wing, helicopters, MAVs (micro air vehicles), free flight, and everything in-between on the books. Therefore, we should be able to produce something worth reading while not having nervous breakdowns.
The basic format of the column will be to have a small amount of waffle at the beginning, a large and meaty feature article by one of our writers, followed by a selection of smaller mini-articles showing tips, techniques, and mini reviews. Then there will be the usual email input displaying reader's projects and ideas.
In this issue, I will be writing about the basic theory and function of magnetic actuators. These are much more common now, and many of you will use them, so please forgive me if I am preaching to the converted. A mini-article by Ralph Bradley on a handy way out mounting Lithium polymer batteries follows.
Magnetic Actuator Basics
By Graham Stabler
Put simply, a magnetic actuator is what you use when a servo is too heavy or too expensive. (That's done it. I have written the first line of the article.) I wanted to talk about the history of the magnetic actuator, but to be honest, I was a mere twinkle in my fathers eye at the time. Suffice it to say that the magnetic actuator was one method of swinging a rudder from side to side in the days of steam radio, before the advent of the interweb and proportional servos. In fact, looking through the average modeling magazine just a couple of years back, you would have been lucky to see a magnetic actuator apart from in vintage columns, but now they are reappearing and rightly so.
Why would anyone want to use a magnetic actuator? Would anyone want to recreate one's glory days? After all, who would want to reject the beautifully smooth, proportional control that modern servos can deliver? Well, the fundamental thing that makes them so attractive to the micro/indoor modeler is their weight, or rather lack of it. "Standard" micro servos can be bought from the local model shop down to around five grams, and can be made even lighter with a few modifications. These little marvels are light and very economical, but impose a fundamental limit on the minimum weight of models that may be built using them. Then there are the more specialist servos of WES-Technik, which tip the scales at as little as 2.4g. These are true masterpieces, but are not cheap, and cost is a major factor for some modelers (especially if from Yorkshire like me). The magnetic actuator on the other hand can cost as little as a piece of wire and a magnet, and can weigh as little as 10ths of a gram. With the aid of some modern (and some not so) electronics, they can also provide proportional bi-directional control. If that sound good, then read on!
What Are They Not?
They are not servos! Although we think of a servo as little back box with an arm that moves when we tell it to, the name really implies motor control with feedback. This is exactly how they work. A variable resister (potentiometer) provides a voltage to the control circuitry that varies with the position of the output arm. The servo's motor is energized until the voltage returned by the resister corresponds to the movement "requested" by the pilot, at which point it stops. If a load is applied to the output arm, it will tend to rotate causing the feedback voltage to change. This instantly causes the motor to energize, keeping the control arm in the correct position. This is all done by clever electronics so we can take it for granted.
In a magnetic actuator, however, the situation is different in that there is no feedback. As we shall see, the torque is merely proportional to the transmitter stick's movement.
So, What Are They?
One thing all magnetic actuators share is a coil. A coil is simply a piece of insulated wire wrapped around a former repeatedly.
This diagram shows a typical actuator coil made from copper wire with enamel insulation, sometimes known as magnet wire. Wire such as this can be seen in any motor and is usually golden brown in color. The coil shown here is "bobbinless". This is achieved by gluing the turns of the coil together, forming a solid block. This is quite a common technique for magnetic actuators in models and in industry.
What makes this simple component useful? The answer is magnetism. When an electrical current is passed through a wire, a weak magnetic field is formed around it. When the wire is then coiled, the field from each wire adds up to produce a stronger field and the coil becomes a magnet. In fact, it becomes an electromagnet. One end (hole) of the coil becomes a north pole, and the other a south pole. Which becomes north/south depends solely on the way the current is flowing through the coil.
This diagram shows the cross-section of the coil and the lines of magnetic flux due to the current flowing through the wire. You may have seen this pattern at school using a bar magnet and iron filings. The points to note are the positions of the poles (they could be reversed) and that the highest density of flux lines is in the core of the coil. This is where the magnetic field is strongest.
To understand how this is used to control a model, we must use a well know fact about magnets. Opposites attract. A magnet is now placed within the coil and allowed to pivot through its middle by some means. Don't get hung up by the shape of the magnet. The manufacturers can put the poles where they like. They are normally where you expect them, such as on the flat faces, but not always.
When a current is passed through the coil, the opposite poles attract. The north pole of the magnet is attracted to the south pole of the coil, and likewise, the south pole of the magnet is attracted to the north pole of the coil. As the magnet can only pivot, it rotates within the coil. If the direction of the current is now reversed, the poles of the coil flip, (i.e. north becomes south) and the magnet rotates the other way. By simply attaching a small arm to the magnet, this force can be used to control a rudder or elevator with ease.
It should be noted that magnetic actuators do not have to consist of magnets in coils, but they will include a magnet (generally) and a coil, so the basic principles are the same.
Most people are surprised to find that these actuators allow proportional control, but it's true. Honest! First, here's the science bit. The torque exerted by the magnetic field on the magnet is directly dependent on the strength of the magnet, the number of turns on the coil, and the current through the coil. So to increase or decrease the torque, we need to either get a better or worse magnet, wind more or less turns, or increase or decrease the current. The only factor that can be varied readily in flight is the current. Therefore, by varying the current proportionally, we can vary the torque proportionally, and by changing the current's direction, we can change the direction of the torque. Hence, with the correct circuitry, we can have proportional and bi-directional torque control.
This is where actuators really differ from servos. With servo, a given stick movement produces a given control surface movement independent (within reason) of the torque applied to the surface. For example, moving the stick to full throw at high speed (high load on servo), should give the same control surface movement as full stick while cruising (low load on servo). An actuator, on the other hand, gives a given torque for a given stick position, and this might not be enough to hold a certain control surface deflection. This may sound like bad news, but in practice, it makes little difference, as we fly by what the aircraft is doing, and not by how far the control surfaces are moving. In effect, our eyes are the feedback mechanism. Obviously, modelers hoping to fly micro pattern planes might have different ideas, but most will probably be pleasantly surprised.
Imagine an actuator installed to control the rudder of a model. When the model is stationary, the force required to move the rudder should be very low (assuming the hinging is free). This means that if you apply even a small current, the rudder will keep moving until the magnet catches the coil, the hinge reaches its limit, or until a purpose made mechanical limit kicks in. Therefore, when you demonstrate the proportional control of your actuators to the gathered onlookers, what they see is anything but proportional. However, if you then start the motor, producing some prop wash, the proportional nature is revealed. The airflow over the rudder tends to center it. Now a small current will only move the rudder a small distance.
You may ask, "If there is no centering force, don't the control surfaces essentially flap about in flight?" Well, the answer is yes and no. Although the airflow over them does tend to center them, some people report that the lack of centering force causes fluttering, a lack of control precision, and in some cases, unwanted aerobatics. The need for centering is really dictated by the model and the actuator. A stable high wing or biplane model will generally not care a jot about the effects of a free hinge, and of course, a light hinge requires only a light actuator. Therefore, ultra-lights or ultra-smalls are perhaps best off without centering. More unstable models, such as "low wingers", can really benefit with some centering. For example, most of the low wing fighters I have flown with un-centered actuator control do have a habit of turning in on right turns. According to Peter Frostick, one of our writers, part of the reason for this is the rudder wants to take the line of least resistance on the turn, so drooping under gravity produces even more right rudder. This can induce a scary dive, but it can be counteracted with careful left rudder once the turn has begun. I mentioned unwanted aerobatics; I have seen an extension of this idea recently on my 10" wingspan pager motor powered Spitfire. In the turns, it would indeed start to spiral in, so I countered this with left rudder. The problem in this case is that once gravity had been overcome the plane, the rudder would act as an elevator and cause the model to climb to an inverted pose, before either finishing the roll or crashing. The high power to weight ratio is also a probable problem, as is the general trim. This is my guess as to what is happening, but I would welcome other suggestions.
The following are a few methods of centering that may be used.
Mechanical: Some modelers have added small strips of thin acetate/piano wire to either side of the fin that extend over the rudder. The rudder has to bend these strips to move, and is hence centered by them. Others use hinges made from elastic band. These create a centering force, are nice hinges, and easy to install.
Magnetic: Placing a small magnet or piece of iron/steel near the actuator can create a centering force. For proportional bi-directional actuators, this should be placed carefully so as not to bias the control surface one-way or the other. Some actuators are placed in pairs so that the magnets within them attract producing a centering force, but in such a way that the two actuators work independently.
In bang-bang and galloping ghost systems, a magnet or mechanical spring is often used to bias the rudder in a certain direction. Pressing on the transmitter button causes the actuator to move against the spring. By carefully pulsing the button or by letting electronics do it for you, the model can be made to turn left or right. A constant pulsing of the correct rate of course can keep the model flying straight.
Gravity: A lightly hinged elevator will essentially droop down, but once on the move, the airflow over its surface should push it level. Once in flight, we have a situation just like that of the rudder, but in this case, there is a difference. The actuator is not only working against the airflow, but also the weight of the control surface. In many cases, this does not matter, as the actuator will have plenty of torque. However, in the case of small actuators or large elevators, it makes sense to mass balance the surface. Extending an arm of some sort from the elevator forwards of the hinge line where a weight may be added to balance the elevator can do this. It should then hang level and all torque applied to it be against the airflow. Mass balancing might also be an answer to my Spitfire problem, if added to the rudder, but I broke the prop before I could test that theory.
There are several ways to produce a "rotating magnet in coil" actuator, but it all comes down to the method used to create the pivot that allows the magnet to rotate within the coil.
Firstly, we will consider the most obvious method, which is to fix some form of axle onto/through the magnet that can then be mounted in small bushes within the coil. This method can be seen on actuators made for the, not currently available, MicroMag system and on those made by Bob Selman and Gary Jones (shown later).
Another method is to build the hinge outside of the coil and have an arm that holds the magnet in position within the coil. Actuators of this kind are not currently available commercially, but offer a very easy way for modelers to convert tail-mounted actuators to remote ones or indeed make their own remote actuators from scratch. With a coil, a magnet, a few pieces of balsa, and some copper wire, an actuator is easily built.
The advantage of this type of actuator is that it may be mounted away from the control surface using either a lightweight push rod or a pull-pull system. This is very useful for scale subjects as well as for producing models that are not tail heavy. They tend to be slightly heavier than some other actuators due to the weight of pivots/plastic assemblies, and the push rods of course, but are certainly very desirable for many applications. The use of control horns also allows the torque to be traded for movement and visa versa.
These remote actuators are made by Bob Selman and Gary Jones. You can clearly see the magnet within the coil. In this case, an arm is attached to the magnet that allows pushrods as shown here or in pull-pull configurations. The actuators are mounted on a carbon rod at 90 degrees to each other. The magnets attract, creating a centering force.
This actuator by Nick Leichty actually has two coils acting as one, but the same principles of attraction and repulsion apply. This example is his "Std" actuator and weighs 0.28g.
This homemade remote actuator by Peter Frostick has the hinging external to the coil. The axle and magnet holder are made from one piece of Beryllium/Copper wire (strong and non-magnetic). In this case, the actuator operates pull only against a spring elevator. This actuator was coined the TabMag by Eric Hook in the UK some years ago. It was built as a simpler (to build) alternative to the in-coil hinged actuators. Although, I should mention there are modelers building these too. (Perhaps I will cover more on that in a later article.)
Control Surface Mounted
These include the famous BIRD (Built In Rudder Device) actuator made by Fritz Mueller and distributed by Cloud9 RC (link 3). In this case, the coil is mounted directly to the fin/stab and the magnet to the rudder/elevator by a small arm. The actual hinging of the rudder is what allows the magnet to rotate within the coil. It is therefore important that both the magnet and coil be centered on the hinge line. Otherwise, the magnet will translate rather than rotate when the surface moves. We want to keep the magnet where the field is strongest, so it is best if it rotates in the center of the coil. The original BIRD is supplied ready mounted on a balsa fin/stab and hinged to a balsa rudder/elevator. The complete assembly is then glued into the model. This is very handy, because most of the work is done for you, but can be inconvenient if you are building from another material, such as foam. It is possible to create your own version, but care is required in the hinging. It must be accurate to get the most out of the coil and magnet.
A second method is to mount the coil flush to the fin/stab, but still with the center of the coil and magnet inline with the hinge line. The magnet is connected to the rudder/elevator by a curved metal wire (none magnetic preferably) so the magnet rotates within the coil using the control surface's hinging as a pivot again. This configuration and the BIRD are equivalent. The coil and magnet have simply been rotated by 90 degrees, and a curved arm has been added to keep the magnet and control surface connected.
There is not much to choose from between the two basic configurations. The BIRD can be tricky to set up (if home built), but is sturdy and precise once done. It is also possible to build the rudder actuator into the rear of the fuselage, which can be desirable. The inline method may create less drag, as it displays less area to the airflow, and can be tweaked and re-tweaked with ease. However, it is probably a little harder to hide on a scale model. Additionally, the adjustability can also lead to misalignment after crashes, mishaps, etc.
The tail-mounted actuators are probably the lightest magnetic actuator configuration of all. They add no weight other than the coil, magnet, and a drop of glue, as they reuse the control surface's hinging. The downside is that they can result in a tail-heavy model, and they are not as easy to transfer from one model to the next. They also do not benefit a scale model's appearance.
The above shows the two basic configurations of a control surface mounted actuator. The one you choose will mainly depend on application and personal preference.
This is a BIRD actuator as supplied. The narrower of the two pieces of balsa is attached to the model while the larger is used as the rudder. Obviously, it may be trimmed to match the model.
Here is an example of a homemade "BIRD" built by Ralph Bradley. It displays excellent craftsmanship, precise hinging, and extremely low weight. These have been literally built into the rudder. The coil is glued to the fin/stabilizer, and the magnet is connected directly to the rudder/elevator.
This is an example of an in-line configuration made by Joachim Bergmeyer. It is an excellent example, as it not only has oval coils, but also has flat magnets. The geometry may be different, but the principle is the same. The magnet's poles (ends of the magnet, not the large flats) are attracted to the opposite poles of the coil.
Here is an installation of the coils supplied with the Dynamics Unlimited RFFS-100 receiver. It uses the inline configuration. The magnet in the coil is actually two magnets stuck together. They are held within the coil by a small piece of wire attached to the control surface.
There are countless ways of making an actuator, but most involve a single coil and a magnet. However, a few others are out there. Some act like a solenoid, and some have two magnets on arms attracted to the coil, but not residing within it. If you have any unusual actuators, drop me a line, or better yet, a picture.
A Few Practical Facts
We will hopefully cover the practicalities of actuators in the future but I thought I should mention one or two things now.
1) The magnets used are usually Neodymium-Iron-Boron rare earth magnets; these are the strongest magnets available and can be bought from many places including electronic surplus houses.
2) The coils are generally wound to be a reasonably close fit around the magnet; this maximizes the number of turns for a given resistance and makes for a more effective and smaller actuator.
3) The thickness of the wire is also important, as thicker wire requires more turns for a given resistance. This, as we know, creates more force, but in turn weighs more. The coils are generally rated by resistance simply because this dictates the current flow through the coil and the choice of actuator driver.
4) In the in-line configurations, the arm that holds the magnet is often made from copper wire, but any non-ferrous material is fine. Joachim has used thin ply for his. For the BIRD, the same is true. Balsa or thin ply is often a good choice for foam models.
A basic magnetic actuator consists of a coil and a magnet. It may be built into the control surface or used like a servo with a pushrod or pull-pull cords. Built in actuators offer the lightest potential solution, but can cause CG problems. Remote actuators can be placed at the CG and are easily concealed on scale models. Centering can be produced in a number of ways, such as mechanically, magnetically, or by the use of gravity. This can be beneficial for some models, but is wasteful of actuator force. Above all, the magnetic actuator provides a cost effective method of controlling a model aircraft at minimal weight and personally, and I love them.
Simple Magnetic Mount For Small Lithium Polymer Cells
By Ralph Bradley
I have been using small magnets to mount the wings and battery packs on my micro models for the past three or four years. The magnets are such a convenient way to mount these items and work so reliably that I have given up using the other popular mounting methods. The magnetic battery mount is also an ideal choice for profile models.
I recently completed a new profile design for my RFFS-100 system, and I wanted to incorporate my usual technique of mounting the magnets flush with the fuselage side in 1/8" diameter holes. I use gold plated rare earth magnets that measure 1/8" in diameter by 1/16" thick and weigh a tenth of a gram each. These magnets (item #48) can be purchased on-line from a small company called Forcefield. Two of the 1/8" magnets are enough to hold a three-cell 50mAh NiCad, so I was sure that since the new 140mAh LiPoly cell that was to be used for the design is about one-third the weight of the NiCad pack, there would be no problem with two magnets. If fact, it might only require one.
When I received the first of the new cells, I discovered there was going to be a small problem after all. Because the LiPoly cell is encased in a plastic coated aluminum envelope rather than a can, the magnets would not stick well enough to hold the battery in place, as the magnet was only attracted to the nickel electrode inside the cell. I found the magnets grip was further reduced after I added the connector and shrink sleeve to the cell. I thought about going back to using a small square of Velcro, but I really wanted to find a way to make the magnets work. I decided to make some coffee and see if I could think of some clever, and more importantly, a light method of employing the magnets.
As luck would have it, I needed to open a fresh can of coffee and while doing so, and the answer was literally right under my nose! Magnets love tin cans, and the coffee can lid was a nice thin, plated piece of tin. All I had to do was cut a small strip from the coffee can lid and attach it to the battery with a matching strip of servo tape. This was very simple, and it only added an additional tenth of a gram. I used 1/4" wide servo tape (doubled sided foam tape), and cut the tin slightly wider than the tape. A light sanding with some 240-grit emery cloth did a nice job of cleaning up the edges.
On my model, I had some wire running along the battery mounting area, so the tin strip and servo tape provided a nice stand off to keep the battery from touching the fine wires running down the side of the fuselage. You can reduce the size of the tin strip slightly, but I chose to make it the full length of the battery. I discovered that by mounting the magnets close together (about 3/8" apart), I was able to slide the battery fore and aft a bit too fine-tune the CG.
|The Lithium Polymer cell complete with tin strip and the two tiny 1/8" magnets installed flush with the fuselage|
On my next profile model, I plan to mount the RFFS-100 receiver unit in the same manner. With a little planning, I should be able to mount the receiver and battery with the same pair of magnets with the receiver on one side, and the battery located on the opposite side.
This is the end, until next month at least. Please send me your feedback, pictures, or requests for topics to be covered to my email. (Click on my name at the top of this article to send me mail.) I hope that we will have even more stuff for you in the coming months when we get into the swing of things, but your input and ideas are needed, from a picture to a whole article.
In fact, rather than be so vague, let me ask for your "super hints" little things that make building easier or models fly better. Don't worry about diagrams. Just an explanation or a picture would be fine.
Until next time,
|Nov 01, 2006, 06:53 PM|
nice and very interesting article about DIY actuators.
But how to conect it to the receiver?
Can I do it to a normal like "Penta 5"?
Some need to built electronic?
perhaps it was published withhin an other inside story but I could not found somthing by search the forum.
|Nov 14, 2006, 03:54 PM|
Joined Oct 2006
Awesome article! I've just started a stupidly ambitous project of a sci-fi flier with a 10" WS... These things are just what I needed
Same question tho': How do I hook it into the Rx?
Many, many thanks!
|May 29, 2008, 06:58 PM|
Joined Aug 2007
Same question tho': How do I hook it into the Rx?
No answer to this in a couple years?
Actuator is useless without a controlling device (IE: receiver)
|Thread||Thread Starter||Forum||Replies||Last Post|
|Article Inside Story - June 2003||Graham Stabler||Electric Flight Events||7||Jun 27, 2003 08:38 AM|
|Inside Story - February 2003||dave_lilley||Indoor and Micro Models||10||Mar 17, 2003 04:54 PM|
|Inside Story - February 2003||dave_lilley||Site Chat||0||Feb 23, 2003 08:35 PM|
|Inside Story - January 2003||dave_lilley||Indoor and Micro Models||15||Jan 11, 2003 08:34 AM|
|Inside Story - January 2003||dave_lilley||Site Chat||0||Jan 07, 2003 12:34 AM|