Servos
Earlier in previous articles we had covered batteries, switch harnesses, chargers and touched base on the reasons for upgrading them. Now, the actual electrical current consumers will be investigated following the same concerns. Hopefully you will find some additional but related information on this topic useful.

Let us also consider servos as they are often the most desired onboard electronic upgrade. It can be stated that all modern servos will work with all radio systems if the plugs are wired correctly for polarity and signal. We have to deal with three conductors which are the positive, the negative, and the signal wire. It is possible in some cases to own brand X receiver and use brand Y servos by changing the pin locations in the plastic plugs. The connector pin locking clips are either part of the metal pins or part of the plastic connector. Sometimes the pins can be inserted into a different brand plastic plug, but not always. Additionally at other times, the index notch can be cut off the plug allowing insertion into another receiver. The key here to be cautioned on, is that you have potentially disabled a safety feature designed to prevent servo reverse polarity. If the center wire is the same polarity in both cases you are pretty safe that components will not be smoked if and when things are plugged in backwards, or unknowingly used with the other relative but different system. For the perfectionists correct plastic plugs and pins can be installed. Other people will use short servo adapter wiring harnesses, but at the risk of more mechanical connection points. As many may notice the pins can be gold or silver in color. Just because the pins are silver colored does not mean that they are silver plated. Silver offers the lowest resistance to electrical flow while gold plated pins have a slightly higher resistance. Gold however offers a higher corrosion resistance and over the long term will retain a better connection. This is why you see gold used in high end audio and video connections. It should go without saying that if you do not have the technical skills, you should stick with factory supplied servos of your particular brand or use prefabricated harness adapters.

The polarity coding for the four major brands, Futaba, JR, Airtronics and Multiplex are as follows. Futaba /JR center pin is positive and Airtronics/Multiplex center pin is negative. I did not specifically mention wire color coding since various pigmented wire harnesses are available from different manufacturers. Normally either red or orange is used for the positive conductor though. For the other connector pin locations to mean anything in relationship to the center pin the connector needs the index to be used as a reference. The diagram #1 below should most certainly give an accurate representation. Airtronics older male and female plugs have numbers molded into them which makes matters easier. Signal, negative and positive as in 1,2,3.

Much as appliances in our homes can be plugged in anywhere else in North America so should our servos. I hope one day that all the radio and servo manufacturers might cooperate together for the common good of the hobby to use a standardized wiring polarity and connector type.

Now that we know how to make any servo work for us, which one should we pick? Rather than start a war by suggesting my personal choices I’ll give some guide lines to think about. Good quality standard ball bearing servos can surpass the performance of top end servos used by expert flyers a decade ago. Somehow they managed full maneuvers and maintained a satisfactory level of reliability using 60 to 70 in/oz of torque. This brings up the popular question of killing the fly with a 140in/oz sludge hammer. The standard quality ball bearing servos have speeds of about .20 sec at 60 degree rotation which is fine for the cyclic and collective on a mechanical CCPM control system. The swashplate and rotor head basics have not changed much in ten years. What has changed is the tail rotor stabilization methods and flying styles.

Another way to look at the cyclic servo torque requirements is to evaluate the servo torque and rotational travel in relationship to the end result at the blade grips. The cyclic servo travels at more than 90 degrees while the blade grips rotate cyclically about +/-6 degrees each for a total pitch change of 12 degrees stop to stop. We can see an advantage here even without considering that the flybar on the average actually reduces the servo loading requirements. The ratio of 120:12 degrees or 10:1 suggests some pretty impressive torque available at the blade grips. The merit of using full cyclic ATV values and shorter servo arms becomes apparent…..something that the ECCPM system cannot do. The effects of servo dead band should have you pondering about exactly what tolerance figure is really necessary!

Modern gyros need fast servos to perform the best, but will work ok on lower priced ball bearing types. In most cases and if you are starting out there is no need to spend a hundred dollars on a super fast tail servo since you will not advantage its speed. By the time you do, rest assured there will be a better product to buy! High end piezo gyros matched to a fast and accurate servo will allow backwards flight at higher speeds lessening the piloting workload. When you do decide that it is time for a tail rotor servo upgrade, speed and accuracy are more important than high torque. This is because the T/R control has light loading requirements. All gyros and governors will work with any system with the correct number of channel requirements, and much like the servos need only the plugs altered.

If you have the luxury of affording the fastest and most powerful servos controlling the main rotor the actual feel of the machine will change. While it will respond quicker to initial cyclic commands, the overall maximum cyclic rate will still be dictated by the flybar, blades and helicopter setup. In other words once the servo reaches its commanded position the rest is up to the rotor system. Where one can advantage the higher servo torque best would be on the collective system during 3-D styles. This is because of the additional mass of the collective controls, the mechanical travels involved and the lack of any flybar factor. The flybar tends to reduce or aid the cyclic servo loads. During hard 3-D maneuvers it can be seen to lag behind the rotor disk removing cyclic blade pitch. During less aggressive inputs it will actually lead the main rotor thus assisting cyclic blade control. Model pilots adjust to the servo speeds just like they adjust to a different helicopter setup. Where a high end servo has the theoretical advantage is when stopped. This is because a normal servo is weakest when stopped. It also draws the highest power when most heavily loaded under this condition. Higher powered servos will drain the battery sooner for this reason when starting, stopping and accelerating. The question is do we add a bigger battery pack, or quick charge / swap packs often to avoid the weight trade off.

It is possible to buy an expensive servo with high torque, super fast speed with metal gears only to find minor slop in the gear train. While very rare it is plausible for metal to metal gears to cause glitches due to lack of an adequate insulating lubricant. The plastic gear trains are immune to this and stay tighter longer. Certainly if a high torque servo is allowed to solidly bind, the high torque motor has the possibility to strip the gears or over heat to burn out.

Speed and torque can be given at 4.8Volt or 6 Volt. It only stands to reason that the higher voltage will give more torque and speed. If this is not stated in the documentation when close comparisons are made, then how are we to know which is faster? Some servos quote dead band for centering accuracy but most don’t so this leaves things open to guess work unfortunately. In the past some servo specs had reference to a stalled current consumption……a worst case scenario.

There are three motors of two types that I presently know of inside our servos. The coreless and the standard 3 and 5 pole motor. The five pole motor naturally offers better performance over the three pole. Several advanced motor designs are certainly possible, but the most important choice for servo design is between iron-core and coreless motors. In the basic iron-core motor, wire is wrapped around a piece of iron, and electricity (through the brush and commutator) turns the wire and iron mass (armature). In the case of the coreless motors, a hollow armature coil, usually made of copper, is situated over a stationary magnet system. When energized, only the coil moves accomplishing the mechanical work through the shaft that is attached to it. The elimination of the rotating iron mass offers the advantage of fast dynamic response, high power efficiencies, and a low current draw. To look at it in a practical way, the coreless servo is more accurate and faster. Because the main difference between the two is that the coreless armature or rotor has no iron core, causing it to be lighter, it can also be substantially smaller. Remember the stationary magnet resides where the core would normally be inside the hollow armature. Physically, the magnet assembly has a hole in the center for the rotor shaft to pass through. See photo #3.

Digital servos and servos with a high refresh rate are the newest breed. This is also highly reflective of their price. Gyros with 2700 compatibility are able to update the servo faster offering better yaw control or stability compensation. The frame rate from the receiver to the gyro still remains the same standard lower rate. Due to the faster gyro update information sent to the servo from the gyro, the motor needs to be quick and geared correctly to fully advantage the design. This will cause it to consume more electrical current.

The newest digital servos offer superior dead band and stationary holding power. In addition, higher torque is available at a slow servo movement but at the cost of an increased electrical current drain from the battery. A very few of these can be programmed internally for speed and travel independent of the radio transmitter menus. Be advised that if your controls bind at extremes with any of these digital servos, a very high current flow will surely reduce the battery energy level quickly.

Sometimes we mechanically check for binding on the bench and think that everything is fine. The swashplate tilt deflection is below maximum and the collective is also checked to be fine at both extremes. We have even gone so far as to successfully check the cyclic by boxing all four corners. What may have been missed is a possible rotating control system limitation if the rotor system was not checked properly and completely at various points of rotation. By holding the cyclic control at full fore/aft and roll positions (boxed), then turning off the helicopter electronics for a rotational check may indicate a control system throw limitation which is exceeded. As the rotor is turned slowly through 360 degrees the deflected servo or servos will be forced back towards neutral. Find the binding and correct the matter rather than needlessly burning battery energy. It gives a whole new meaning to the term “turn and burn” doesn’t it!

So now on to the main reason for this article. Prior to this document we had gone over switch harness and connector pin limitations. These have a maximum rating and since it has now become possible to exceed them in certain cases, I thought it interesting to pursue the matter. I have flown 60 sized machines with 50-70 in/oz rated servos through loops, rolls, flips, autos etc and found them adequate. Since I need a common figure I picked 40 in/oz which is the same as a 5 pound weight suspended below from a ½ inch servo arm. You are probably still wondering what exactly is this guy up to now.
Using this common torque value by way of a spring scale or weight we can see exactly how much current is used to hold different servo types in position. We can also make a judgement on centering accuracy. Further to this the transmitter stick can be moved rapidly to one extreme and the peak servo current measured. Very quick short movements can be simulated using a switched channel to crudely indicate the kind of loads a gyro can impose upon a servo. For these tests to be broad spanning the spring scale or the weight hanging from the servo arm at a specified distance can be removed so that non-loaded values might be recorded. We must remember that as the arm rotates more, the torque loading will be reduced due to differential just like in the helicopter. The main thing is comparative consistency here. To check the centering accuracy, a dial test indicator is used with various techniques. Using the specified arm of .5 inch it can be stated that .009″ on the DTI equals approximately 1 degree. The spring scale is implemented to verify that the weighted or loaded dead band result would be double this figure when loaded in both directions. See photo# 4 & 5. Certainly full ATV must be used in all cases. This brings me to the question about digital programable servos with internal adjustments for speed and end points. To test these for electrical current consumption in the worst case, real world condition would require maximum internal speed and travel adjustment. Unfortunately I didn’t have any of these to test and cannot attest to their qualities.

After obtaining some data, we as modelers have to decide what to do with it. We have to consider that the tail rotor and throttle servos are lightly loaded. Since most people use different servo combinations this has to be factored in. To me this suggests reducing in some cases a (throttle) servo’s constant electrical load value. The tail will work more steadily than any other servo but it will “situationally” use more current since the servo motor is accelerated and decelerated at a high rate due to gyro compensation. What this also means is that while it can drain more battery energy than say a cyclic servo (higher average current flow) over the duration of a flight it might not have the highest peak current consumption. When the gyro is settled down or with the tail controls held at one extreme energy consumption will be lower. Running a higher frame rate servo with the appropriate gyro will use more power. Flying styles definitely have a big part to play here. It would be interesting for someone to fly a helicopter aggressively with a remote indication of current drain and voltage then compare this craze to that of a beginner. While this might be the best way to judge a particular situation, I feel that using a substantially larger battery, common sense, correct system wire sizing in combination with the proper ESV, will fully compensate for the newer servo technology and different flying styles.

We statically checked the electrical current consumption of a non-running 30 sized machine equipped with four standard 3 pole servos and one normal coreless servo. This was accomplished by moving all the sticks rapidly on the ground with indications of just under 1 amp. We tried the same thing with a 60 sized machine using four quality name brand regular coreless servos with one digital servo on the collective. The blades were not installed in both cases. The 60 machine drew 1.5 amps. It certainly causes a person to take notice.

The following servos were rated on the bench, Futaba148’s without bearing kits as a 3 pole type and Futaba 9202’s as normal coreless types. The Airtronics 94158 was used as a high torque conventional coreless servo. Finally two new digital servos from Futaba and two from JR were tested. These tests are not meant to rate one brand against another but simply to give a broader spectrum, or more basically, be indicative as a type rating. As you move up in price from and including the 9202 “type” expect damped pots (vibration isolation), coreless motors, dual bearings, “O” ring sealed cases and output shafts. These manufacturers can be categorically relied upon to offer quality products similar in cost. I initially checked all servos with small and large battery packs and found in some cases increased stalled current and holding power with the higher AH rated ones. The same indication resulted from using the larger gage wiring harnesses with the more powerful servos.

From the tests carried out in chart #6 it would seem possible for the radio system to consume well over 3 amps in certain configurations. If a poorly rigged control system with binding is factored in this will surely have an impact. It is possible that reports of 1200 ma/hr batteries lasting for only two fifteen minute flights may not actually be all that far fetched.

If you look at the conventional coreless 9202 and compare it to the newer digital type as exampled in the DS 8231 it can be seen that plastic gears are tighter but more importantly a much lower deadband with solid centering and holding power becomes obvious.

Care must be used when interpreting the 8700 data since this is a lower torque tail rotor servo with a respective higher accuracy based rather on speed and super fast updating. It was not designed to carry a big cyclic or collective load as in the tests. In other words if the loading values are reduced a much lower deadband and substantially better centering will be evidenced. What impressed me with this unit besides speed and smoothness, was the relatively low current draw especially when you consider what a good job it can do.

My overall impression of the digital servo type is very positive but my feeling is that they are not required in all cases by the average sport flyers on the cyclic, tail rotor or throttle. I do think they can be the “Cat’s Meow” when installed on the collective! You may want to look at the general cost rating in chart 6 based on the most expensive servo being 100 percent, then think about your flying requirements and financial status. Handling these high end servos physically while hooked into a radio system causes it to become very obvious of the superior performance, even when no fancy tools are used for validation. Where I think they will be best advantaged when used throughout a helicopter is with a fast and sensitive 3-D setup requiring the highest precision and control speed. Certainly piloting skills must be up to par with these very high specifications. They are none the less appreciated in much the same manner as a fine and accurate high end model helicopter. Sooner or later these will become as common place as regular coreless servos and with similar pricing hopefully. During all the tests these servos were far from being over tasked. Intense holding power, speed and the ability to very accurately return and follow stick movements are their virtues. All this is unfortunately at the expense of a higher battery drain. The JR DS 8231 seemed to be the exception for some reason. The digital servo should be favored by the demanding model aviator using ECCPM. The reason I say this is due to their virtues being ideally suited for, or causing a reduction in this control systems drawbacks. As far as I am concerned, piloting inputs at the sticks will never exceed the digital servo’s speed or centering design limitations. Any system control errors will be solely the result of servo arm geometric differential and reduced ATV induced resolution. This is obviously something that cannot be blamed on a servo.

Some of you may not be aware that Futaba offer a battery fail safe in some of their radio systems. The GV-1 governor also has such a feature built in. Further to this, Futaba supply an add on device called “Fail Safe” incorporating this battery fail safe feature along with the usual fail safe pulse monitoring feature. Basically what it does is to cycle the throttle to a low position selected by the user in the case of a battery having low voltage. This happens at a specific low voltage limit which may be exceeded unknowingly. I have had mine activate when in fact the battery still contained a good charge. The problem first appeared when I installed three high torque servos on the main rotor controls. When I replaced the switch harness and the battery lead with a larger size the problem went away. This is just something to keep in mind should you ever come across it.

While on the topic of the GV-1 engine governor, I’d like to quickly convey my satisfaction and pleasure in using this device over the long term. I have yet to find an equal in performance, reliability or features and I clearly refuse to go back to throttle curves with inferior cyclic/throttle mixing hassles required for quality aerobatics. I have another unit on order as I write this. I have used it on both glow and gassers with similar results. As they say, give credit where credit is due.

Enough about servos and things so on to receivers. Some receivers will work with a different transmitter brand. Other receivers will not work with different transmitters within a common brand. JR and Airtronics PPM are interchangeable but both will usually not work with Futaba. Some expensive radio transmitters have a selection to offer compatibility to get around this restriction. Various Airtronics PCM receivers are not fully compatible within this brand or with others for that matter. It is best to always perform ample research when you seek a chance to mix and match. Mentioned are but a few examples.

With all the recent advances it is hard to guess accurately where technology may lead. If I were a betting man I’d be thinking in the future direction of brushless motors and servos without pots (variable resistors). Most long term servo problems occur due to brush or pot failure. By removing physical wear due to frictional contact, extended servo life will be possible. Commutation can be carried out electronically by sensing the motors position using hall sensors or similar methods. The life of the motor could now be dictated by bearing life rather than brush wear or pot failure. The rotors of these motors are made up of permanent magnets with a stator coil assembly surrounding the rotor to provide the moving magnetic field which drives the motor. The rotating magnetic field is controlled and sensed electronically based on the rotors position in the absence of brushes and commutation. There is no reason that the same rotation positional sensing methods could not be applied to the servos output shaft location, similar in function to the present day pot. The life span of modern servos is hard to judge due to different operating conditions. What I can tell you is that one gentleman I know (Rod Byrd) has logged 100 moderately aggressive hours on 9202s installed in an X-Cell 60 without failure. He replaced these servos last year on principal only, since no hard data exists or is recommended on the matter. I happily look forward to future technological improvements.