|
Background
There is a lot of confusion surrounding PCM and exactly what its advantages and disadvantages are. PCM and PPM are both encoding methods used for FM signals, and PCM is not a totally different transmission method like AM. Therefore, the PCM signal is affected by the same interference as a PPM signal. The reason PCM works better isn't because it gets less interference, it is simply better at dealing with it.
There is a fundamental difference between PCM and PPM signals. A PCM signal is digital, and a PPM signal is analog. In a digital signal, there are two distinct states, on and off, and to mistake one for the other requires a significant amount of interference. In the analog signal, there are a continuous number of states representing the signal, and interference can cause each state to shift to an adjacent one. Therefore, a PCM receiver will work perfectly, without failsafe or dropped frames, under much higher interference levels than a PPM one.
The digital signal allows the PCM system to perform additional functions as well, because additional data can be added to the signal. Firstly, PCM radios send what is known as a "checksum", which adds redundancy to the system by allowing the receiver to check that the data it is receiving is not corrupt. In fact, the checksum is so effective that it is nearly impossible for a receiver to incorrectly accept invalid data. If the receiver determines that the data is invalid, it will hold the last position. Almost all PCM radio systems also transmit "failsafe" data, which is used to set servo positions when signal is deemed corrupt for more than about half a second.
For a great, technical, in-depth explanation and comparison of PPM and the various types of PCM, visit this site: http://www.aerodesign.de/peter/2000/PCM/PCM_PPM_eng.html
Failsafe
Failsafe is a function of most PCM systems that allows the system to move the servos to a predetermined position upon loss of signal. Usually, there is an option for each channel to either move to a preset position or remain at the last commanded position. Because of the nature of failsafe, it can never be disabled. The PCM system knows when the data it received is invalid, and simply cannot be made to act upon it.
When many people first hear of failsafe, they think its a system designed to save models. Although it can, if properly programmed, it is really a system designed to save people. An all too common failsafe approach is to attempt to put the model into an attitude where it can land safely without control. However, unless you're flying a noseheavy rudder only plane with ample dihedral, there's really no control input that will land the plane safely without input. The overwhelming majority of PCM receivers are used in models which are, at best, neutrally stable. There is no single control input that will save such a model from every attitude. Therefore, a different approach is necessary for proper failsafe setup.
The best failsafe setup for most models is to hold all channels except throttle, and set the throttle to a predefined idle point. This accomplishes two things: it puts the models in a predictable low power path, and it provides a possibility of recovering control when the engine is responsible for interference. For spectators, the easiest models to avoid are the ones that are moving slowly and in a consistent path. Nothing is harder to avoid than an erratic model flying full bore. PCM can accomplish this by reducing throttle and maintaining all other control inputs at their last position. Furthermore, when the throttle is reduced, any engine related interference is likely to disappear, and control can be regained. Especially in gas models and helicopters, engines are responsible for vibration or electrical noise which accounts for a large portion of radio interference incidents.
It must be noted that this "best" failsafe setup is not the default for any known PCM radio. It has to be set explicitly by the user. Therefore, if you are a pilot using PCM, I suggest that the first thing you do when you set up a model is set up failsafe. You won't have time to set it up later when you need it.
Range Testing
Range testing of a PCM system is different from range testing a PPM system, because PCM systems don't exhibit any "glitch" upon loss of signal. When a PCM receiver nears the extent of its range, it first stops dropping "frames", or servo position commands. If enough frames are dropped in a row, the radio goes into failsafe until signal is regained. At no time do the servos jitter or bounce like PPM systems.
When range testing PCM radios, most pilots set failsafe to move throttle to idle, and walk the transmitter away from the model. When the throttle moves to idle, failsafe has been enabled, and the end of control range is assumed. This method fails to consider that, at some point before failsafe is activated, frames are being lost, and control is lost for small periods. Sometimes, frames are lost as early as 20-30 feet, even though the distance where failsafe enables seems normal.
If you wish to determine whether or not the PCM radio is getting the analog of PPM glitches, a different range testing method is necessary. To do this, one control needs to moved smoothly back and forth during the range check. The stick needs to be moved as quickly as possible without going faster than the servo can physically move. If, when watching the model, the surface hesitates, the system has received a glitch. If hesitations occur closer than 150 feet or so, there is a radio problem that needs to be addressed.
This range testing method addresses one of the most common criticisms of PCM radios; that there is no indication that signal is about to be lost. There is an indication, you just have to look differently to find it.
Myths
If it wasn't for PCM failsafe, I could have saved the model.
If your model enters PCM lockout, there is already enough interference to render a PPM receiver unusable. Furthermore, the erratic glitches of a PPM receiver are just as likely to crash a model as a PCM failsafe's "hold". The fundamental problem is that there was enough radio interference to keep the radio from functioning properly, and no receiver, especially PPM, will help when that occurs.
It would be safer if I could turn off my radio's failsafe function.
PCM's failsafe function is intrinsic to the nature of PCM. The fact that PCM is a digital signal with signal integrity checking means that the receiver knows when the signal is getting interference. If the radio were to act upon these errant signals, the resulting glitches could be several times worse than the same glitches in PPM, although slightly less frequent. If a signal is getting interference, the best thing to do is ignore it, and hold the servos in a predictable position in hopes that the signal returns.
I don't like PCM because I can't tell if there is interference or not.
The fact is, most of the time, you can tell that there is interference. PCM masks interference for three reasons: signal quality, failsafe, and checksum validation. Because the PCM signal is digital, it is less susceptible to interference, and interference that results in glitches in PPM will have no effect whatsoever on PCM. Failsafe activation is a sure sign that the interference is high enough that a PPM receiver would have long ago been rendered useless.
Of more concern is the last reason for interference masking: checksum validation. When a checksum validation fails, the frame is discarded, and the servos hold their position until the next valid frame. If enough frames are discarded, the radio goes into failsafe mode. However, some people contest that there is no way to know glitches are occurring until the radio enters failsafe. This is untrue, however. On the ground, you can test for dropped frames by moving the surfaces slowly during a range test and watching for hesitation. In the air, the model will feel unusual when frames are being dropped. The feeling is akin to servos pausing or getting stuck briefly. This might not be noticeable to novice pilots, but I believe most experienced pilots would be able to recognize a receiver that was dropping frames.
If I enable failsafe properly, the plane will land with minimal damage.
Most of our models don't fly very well in free flight. If you happen to be flying a model that can fly hands off for long periods, then perhaps this logic is valid. However, if you're flying a helicopter or a typical aerobatic plane, there's really no control inputs that will unconditionally save a model. Trying to save a model should not be the goal of your failsafe setup, the goal should be to save anyone who happens to be in that model's path upon loss of control.
PPM is more responsive and more precise than PCM.
This is a remnant from older PCM systems, which were inferior to the systems we use today. Modern PCM systems have resolution higher than most servos. At worst, modern PCM systems are as responsive as their PPM equivalents. In some cases, such as Futaba's PCM 1024, the responsiveness is even greater than PPM systems.
If PCM goes into lockout, it takes forever for it to come back.
A PCM receiver takes roughly 3 times as long to regain control after signal is restored. This equates to about 60 milliseconds. 60 milliseconds is faster than any normal human can react to the control being restored, and its also faster than the servos can respond to control being restored. So, it may take longer, but the delay won't be noticeable. People think there is delay for signal acquisition because the receiver generally takes a second to respond when it is first turned on. The delay you notice when you turn on your radio is not the same delay you'll notice when you regain signal in flight, because the radio has to go through initialization routines when it first turns on.
Safety
PCM has a number of safety advantages over PPM, all of them relating to the predictability of PCM's loss of signal behavior. When a PCM receiver loses signal, or gets one with significant interference, it will either hold servo positions or move them to predetermined "safe" places. This is in contrast with PPM, which could do anything, from holding the servos to moving them to the extents of their travel. A few examples might clarify my point.
First, consider someone hovering a helicopter near themself. If radio control is lost with a PCM receiver, the helicopter will drop throttle abruptly, indicating that signal is lost and something is wrong. This will give the pilot time to make the decision to get away. In the case of a PPM receiver in the same situation, it is quite likely that the model could give the inputs necessary for the helicopter to chase the pilot, a potentially dangerous situation. This isn't a purely hypothetical situation, either, as just about anybody flying helicopters for a few years can tell you.
Next, consider a pylon racer flying the course at full speed. If this plane was subject to interference with a PCM radio, the throttle would cut and the model would continue on its path, which is predominantly away from spectators. In the case of PPM, the throttle would not be cut, the plane may become erratic, and there's a chance it would head for the crowd. Again, this is not a purly hypothetical situation, as I have seen it myself.
In summary, a predictable response on loss of control is always preferred to an unpredictable one. Ideally, we would eliminate loss of control altogether, but because we can't, PCM is a great safety benefit.
| 
