The Transmitter takes the various positions of control inputs / channels from the user and encodes them in some format. These encoded inputs are then concatenated or encoded together into a packet or frame. The frame is then transmitted or broadcast via radio frequency on particular frequency or channel.
The transmitter may perform some kind of processing on the control input to generate an output channel; some channels may be totally synthesized from a combination of inputs.
The Receiver listens for transmissions on a particular frequency/channel. When a transmission is received it then decodes the frame from the transmitter. Each channel in the frame is then decoded. For each channel the receiver has, it outputs a servo signal to position an actuator to match whatever the user of the radio selected.
Receivers may perform some type of processing to the channels. Special Processing may also occur on lack of signal, or reception of an invalid signal from the transmitter. This can be used to put the airframe in some sort of fail-safe configuration to prevent interference or range problems from causing a potentially unfortunate situation.
The Servo is an actuator used to generate the position of a channel from an electrical signal output by the receiver. Indeed, any device generating the appropriate format may control a servo.
Some servos may perform further processing on the positioning signal to adapt it for a particular situation.
The Electronic Speed Control is another device which inputs a servo signal and generates a physical output. In the case of the ESC, it outputs voltage / current to control an electric motor in electric powered models.
Even if you aren't interested in PPM radios, it is important to know about how this signal works. It shows up directly or indirectly in servo control signals, and the design of the entire system. Until I have a better place here is a article on RC transmissions.
DSM is the "new kid" on the radio control block. It uses the 2.4GHz frequency range used by WiFi and portable phones. Instead of fixed frequency, the xmitter and receiver are bound to each other; they basically remember each other's addresses. Each time the transmitter is turned on, it finds an empty 2.4GHz channel to use, and stars occupying it. If it can't find an empty channel, it waits for one to become open. When the receiver is turned on, it scans the channels looking for its transmitter. Once it is located it enables itself. If it can't find its transmitter, it doesn't do anything. With this going on, one doesn't have to worry about channel confusion.
Depending on the manufacturer this can have different names; here are a few:
As with PCM there is little or no compatability between manufacturers. Fortunately traditional servos work well with these radios, as do the newer digital and hi-rate servos.
The frequency range used is the 2.4 GHz band, which runs from 2.400 to 2.4835 and contains 79 1MHz frequencies (or channels). (XXX find a reference, some articles imply more channels)
Radios (transmitters and receivers) used for airborne, aka air or heli radios are a bit more complex than the ordinary DSM radio. This is, presumably for two reasons; the speeds and energy of airborne models, and the ranges they operate at. Instead of a single 2.4 GHz channel, the transmitter actually finds and uses two channels. When the receiver is initialized it looks for the transmitter's signal on two channels. Once it finds both signals, only then does it arm itself for operation. Two channels you ask -- how does the transmitter and receiver xmit on or receive on two channels?? With two radios of course! With two antennas of course; the antennas are properly mounted 90 degrees to each other to eliminate radio signal polarity problems -- if the signal polarity is wrong for one antenna, it will be correct for the other!. That is what the DSM2 standard is all about -- two channels of DSM. Actually the receiver may have more than two radios, each with a separate antenna; some receivers allow for 1-3 extra outboard receivers. The receiver then examines the signals from all the radios and then processes the signals to generate correct operation in the face of various difficulties.
What kinds of difficulties? Here are some of the few more obvious ones I've read about:
The disadvantages with DSM radios are few, but they are something to be aware of:
Other than that, the upsides are quite nice:
There are two uses of the term channel in RC radios:
Channels are the number of servos which can be run by a transmitter. But, then why do some 6 or 7 channel radios support 9 or 10 channel receivers? Well, that is because some aircraft or models need more than 1 servo to run a control function. For example, how about an airplane with 1 servo per aileron. Or split L/R flaps, with 1 servo on each side. Or heavy elevator forces, with 2 servos. The extra channels on these receives can output signals as appropriate from the "base" channels to provide for these extra servos.
One flight-use of an AUX channel in helicopters is often to control the gyro mode or gyro sensitivity. Another common use is dedicated RPM control signal. That means that 6-7 channels can be used just by flight control issues in a helicopter. This makes a 7 channel radio a bare-bones or a one extra function radio, compared with an airplane where the same radio can have 3 extra channels.
And then there are more control channels for primary flight. This brings the channel total up to 8 before any accessories. Perhaps that is why Helicopters can easily use 12 channel radios!
The servo takes the servo signal from the receiver and transforms it into a physical position. That physical position is used to move or drive something on the model -- a control surface, or power control, accessory, or whatever.
The servo signal is a combination of voltage and signal. Servos have a 3 wire connector which carries the following signals from the receiver.
Essentially the Signal is a type of PWM (pulse width modulated) control input. I say a type of PWM because it really isn't conventional PWM. To servo waits for a 0>1 transition, and then uses the duration of the 1 level pulse (until 1>0) to determine the servo position. There is a min and max pulse length which is then interpreted into a 0°<>180° servo position. The center of the throw (aka stick centered) is typically around 1.5ms, with a 0° of 1ms, and a 180° position of 2ms. To move the servo someplace in-between, an intermediate pulse width is interpolated. The servo then waits for the another pulse to reposition itself. The off (0) signal between control pulses is of a non-determinate width, it varies without any effect on the servo. If the gap is about 10-30ms everything will work OK. If longer, the servo can goto sleep, or may start acting a bit irregular when the next control input is received.
Servo throws vary; a typical servo may have a throw from about 120° to about 180°. The standard may be about 120° of throw, with the 1ms .. 2ms pulse width. Using shorter or longer pulse widths can command more throw from servos which are capable of it. Once the servo starts trying to drive itself past its physical limits, that is all it can do! I mention 180 degrees a couple of times here because that is the first I read about. Later the 120° limitation was introduced by another source.
The typical servo generates position by converting that pulse width into a position info. The servo electronics then drive the motor in whatever direction is required to make the servo position itself to that position. The servo's control circuitry then does its best to keep the servo at that position. To determine its output shaft position the typical servo has a potentiometer which tracks the circular 180° motion of the servo output shaft. That potentiometer encodes the position so the servo control board can determine what signal (voltage, current) to feed to the motor to try and maintain the desired output position. To construct linear push/pull outputs from the servo, the circular motion of the output shaft is converted into linear motion by control horns attached to the output shaft. Some servos are directly linear and use a linear encoding device (aka a linear pot) to provide the feedback info.
The servo is actually quite an intelligent device. If you provide some form of legitimate looking input signal every 50ms or so, it will drive itself to the desired position and hold itself there. The servo can be quite energy efficient -- if no force is required to hold the position, it won't draw power to drive the motor. If a force is trying to displace the output, it will use only the power needed to maintain that position. If the force is greater than what the servo can provide, well it will slip until an equilibrium of power versus force is met. This efficiency of the servo can produce odd effects on battery life -- a servo with a higher output load will consume a lot more power than one with a light load. You might have great battery life with a servo and use, but then the same servo can cause quite a bit reduction in battery life on the next sortie -- if it now has to handle heavier loads. For example, a flight of nice and easy sport flying versus one of intense hi-speed aerobatics!
There are several types or variations of servo; many of them are not exclusive, there are various features which may be combined to make a servo appropriate for some use. Others, such as size are more straight-forward :)
Well, that is a misleading term, because there are too many kinds of digital servos. To simplify things I'll break them into two classes; high-update rate servos, and programmable servos. Digital servos have the normal electrical inputs and mechanical outputs just like a normal servo. What differentiates them is how they treat positioning and shaft position update rates.
I believe this is what most people refer to when they are talking about Digital Servos. Earlier I mentioned that if the servo update rate (time between pulses) is too long, the servo sorta goes to sleep. That is because the servo is more or less awakened by the pulse to start doing stuff. Once the pulse is over and the positioning is satisfied, the servo shuts down (aka sleeps) and doesn't do anything until the next position pulse is received. What that means is that during that "sleep" delay the servo output shaft could be displaced; the servo wouldn't do anything to try and correct that until the next position pulse came along to wake it up again.
With a Digital Servo the servo has a CPU and becomes an active entity. Regardless of the frequency of the control inputs, it is always working on maintaining -- with a high update rate -- the position last transmitted to it. This means that any undesired motion of the output shaft will be immediately noticed and power provided to the motor to reposition the output shaft correctly. This is really great, especially for high-demand applications, such as helicopter controls -- either the cyclic or collective servos on the swash-plate for the main rotor, or the tail rotor collective pitch.
There is a downside to this however. I mentioned that the servo is a relatively efficient device. Well, these digital servos can be providing power to the motor basically all the time trying to do a really great job of holding position. Plus it has a CPU that is running, etc. All this means that a digital servo can be quite power-hungry, even in light-load situations. Say for example vibration (helicopters have quite a bit of that) displaces the output slightly (but with no real force). The servo will kick in and immediately drive the output back where it should be. And it kees on doing this continuously -- the motor is running all the time, the servo is drawing power, and it consumes a lot more energy. Of course, you do get really fantastic control response!
Unlike a normal servo, this digital servo is really smart and can have its own CPU and, memory. You can talk to it and program behaviors into the servo. For example, provide a custom map of pulse width to output position. You want exponential in the servo -- you got it! Seems foolish? Not really. What if you have a device which has control limits which can't be exceeded -- regardless of the signal the transmitter sends? Well, program those limits into the servo, and voila they won't be exceeded. Or, say you have control linkages which have some non-linear aspect to them (very typical). But you really need a linear output, a 1 to 1 map of input position to output. Just program the servo (basically exponential in the servo, but more complex) to make the mapping of input to physical output provide a 1:1 ratio at the control surface movement. Or whatever else the servo can do for you!
Certainly the radio can do some of these functions. But mixing (or a careless user) might create odd effects that would cause problems and break things. Or, what if you have intelligence in the bird? For example an auto-pilot, or CCPM mixer, or ... ? Those things can't talk to the radio, they need to control things on the bird ... and now the servo can take care of limits and actuation nicely.
When a servo operates continuously it really isn't doing positioning any longer. Instead, it is performing velocity and direction control. For example an ESC is a type of Continuous Servo. So is a servo modified for continuous operation.
With a continuous servo the pulse duration no longer maps to a position. Instead it maps to a speed and direction of operation. Full Reverse -- Stop -- Full Ahead, for example. Note that even though you want the servo to operate continuously, that some type of positioning may be desired for uniform speed control (guarantee a speed, not just hope for one). Or, actual positioning for the mentioned Stop command. For example, a servo driving a winch. You may really want it to act as a brake when it isn't moving, in which case some sort os positioning to hold things still may be desired. You may even be able to get free-wheeling in this type of servo by putting the servo to sleep so it stops any positioning.
Servo sizes somewhat reflect the scale of the model they will be installed in. So that means how large a servo will physically be, as well as how much control throw it will have -- a bigger servo generates a bigger throw -- well, a larger control horn generates a larger throw.
The force available is also an issue -- a larger model requires more force ... and a larger servo generates more force. And vice-versa for a smaller servo. So, you could find a micro servo performing a task in a giant 1/4 scale model -- something which doesn't require much control throw or force. But a smaller model could have a hefty giant sized servo in it to control some high-force control surface.
For A Helicopter the radio may need to have some special functions. With a helicopter there may need to be some mixing of channels. There are many ways of controlling a heli rotor system which the pilot doesn't see -- all the pilot wants is pitch and bank and AOA control. But a particular Helo could do that with 2 or 3 servos, in many different configs. But the radio needs to translate the pilot's stick inputs into appropriate servo actuation to do this. The helicopter may also have mechanical mixing of the inputs to deal with, and the radio needs to know about them so it can modify its output signals properly.
Most of the computer radios can handle helicopters w/out problems. Older radios may be helicopter-specific, since they need extra special functionality to control them. A Helo radio may also have special modes for stunts, 3d, flying, hover, etc, and the switches to actuate these modes.
The biggest difference between a helicopter radio an an airplane radio is that in the heli radio, the "throttle" stick motion actually controls two radio channels:
Other Heli radio differences
This is why Heli radios are complex compared to Airplane radios ... If you really notice .. the mixing has master channels and slave channels. In other words, the thing you are controlling, and the other thing which is being modified as a result of the control. The sum of these mixes (which all run at the same time) have some things being masters for some items, and slaves for others. Talk about having a smart radio which doesn't go into melt-down making this all work ...
Sometimes this is called revo mixing, or [anti]torque mixing. Is used to minimize the effect of power changes on helicopter heading control. This is mixing done by the radio to increase/decrease yaw (tail-rotor) input with throttle (increase/decrease) changes, so that torque effects of throttle are minimized. If you have a heading or rate gyro this is not needed; the gyro takes care of maintaining yaw control for you. Power may be the wrong term -- this is affected both by changes to throttle and by changes to collective pitch.
More collective pitch means more blade angle means more
drag, and a slower rotor speed.
So, with more pitch, more throttle.
With close to zero pitch, less throttle.
With negative pitch ... more throttle, at least for
aerobatics and inverted flight.
And this, is why heli radios have different flight modes,
because when you are hovering or in non-aerobatic
flight ... you don't want more engine power / rotor speed
with decreased pitch!
Is needed because as you command more cyclic pitch, you'll need more power to keep the rotor head speed constant. So both left and right aileron will result in increased throttle mix.
Is needed for the same reason as the L/R Cyclic mixing. More blade pitch, more drag, more power. But, it can be different than the L/R mixing due to the effects of airspeed (induced flow) on the rotor disk ...
This is needed because when you command more anti-torque it pulls engine power due to the increased pitch of the anti-torque rotor. When less anti-torque is commanded, less engine power is required (torque turns the helicopter), so need less power to keep the rotor speed constant.
This starts out being more straight-forward, just like setting forward/reverse servo relationships on a airplane. Then it gets more complex due to rotor-head geometry, servo position. And then you add in bell mixing and hiller mixing on the airframe. And then you add blade to head actuator angles and ... and .... it makes my head hurt until I can teach it. There are a couple different setups here for the various different heli geometries. Mostly this doesn't get too complex until you do the next ...
This is like the above, but worse. What happens here is that there is NO collective pitch servo. Instead, there are a bunch of servos directly connected to various portions of the swash plate (see geometry issues above). Collective Pitch is done by moving all of these (now) cyclic pitch servos the same amount at the same time. Cyclic pitch is done by moving the servos differentially. Remember the geometry issues mentioned above which make this a fun problem? It gets worse now, because servo arm geometry issues start getting involved here too.
OK, so we have this cool CCPM setup. Imagine you are making a collective pitch change, but the 3 CCPM servos are channels 1, 4, and 8. With the collective pitch change first the swash-plate will tilt one way .. then another way .. and finally back to level. Well, that sorta bites. So, the radio can re-order the channels to put the CCPM servos right next to each other in the on-the-air order. That means all the CCPM servos get their command almost right at the same time ... or maybe even AT the same time if the receiver waits for all 3 before moving any of the CCPM channels. Now the swash-plate doesn't wobble as collective pitch is changed. Same problem occurs with cyclic pitch changes too in this setup :)