Parts 5 & 6

Transmitter trays

Now, how you hold your transmitter and which stick mode you use can be as controversial as what style of flying you do. Before taking up helis I used to fly fixed wing mode 1 (that's with the throttle on the left stick) and operated the sticks 'thumbs on top' . I decided to swap to mode 2 (throttle right) for helicopters so that none of the undesirable fixed wing reflexes like slamming throttles shut would get in the way. To improve my control over the fine stick movements needed in the hover I started to operate the sticks 'finger and thumb' style. Since this makes it hard to keep a reliable grip on the transmitter I started to use a neck strap. I still was not happy with the security of my grasp on the transmitter as it didn't balance very well about the neck-strap fixing point. (Sweating and trembling hands had a bit to do with the problem too I guess) In desperation I gave a transmitter tray a try.

 

I got the thing mail-order and when it arrived I was horrified. It weighed in at about two pounds and had a deck area like a young aircraft carrier. It looked as if it normally got used to serve ice-cream and popcorn at the local cinema. I'd got some ribald comments off my flying buddy, Ian, just for suggesting a tray. I was cringing at the prospect of what he'd say when he actually saw the thing. (He dubbed it an '18 hour girdle') However, I had to admit the thing worked. If you have never used a tray before it takes some time to adjust to the thing. However, the transmitter is really well located and I'm convinced that, for me, it gives me better fine control in the hover and better access to idle-up and throttle hold switches than I had before. Having one time recently flown without a tray (I forgot to take it to the field!) I certainly don't plan to switch back.

I recently got into a 'conversation' on ModelNet with Peter O'Connor and It transpired that Pete manufactures transmitter trays in the USA under the 'Petal Manufacturing' trade name. I jumped at the opportunity to try a tray designed by a flyer rather than a confectionery vendor so I horse traded a simulator for a tray. I can only hope Pete is as happy with his half of the deal as I am with mine. The 'Petal Elite' tray is light, weighing in at just on a pound (450g) which saves on the neck muscles. It allows use of the buddy socket while the Transmitter is mounted on the tray - useful as it allows me to use the tray with the simulator. I can also charge the transmitter without removing it from the tray (which should save me arriving at the flying field without it!). Most important for me (and a feature not on my previous tray) I can now adjust and optimise the height of the hand rests ('Pro-Pads') relative to the transmitter.

The 'Petal Elite' and 'Sabre 6' trays as they arrive. If you're like me you spend time assembling things like this without reading the instructions and end up with a Dali sculpture...

Sorted! The Elite tray fits the Futaba FF7 like a glove. The 'Pro-Pad' hand rests are adjustable for height

Rear view. Buddy and charging sockets accessible with tranny still on the tray

Straightforward handle clamp holds tranny to the tray

If you would like more details of the Petal range of trays, Email Peter O'Connor at 76055.30@compuserve.com. (Editor's Note February 1998 - We can't guarantee that this still works...)

Tail rotor set-up.

In the last couple of issues I have looked briefly at the setting up of collective pitch and throttle. The set-up of the tail rotor is closely linked to this so let's look at what we are trying to get from the tail rotor.

The first function of the tail rotor (as anyone who has had a tail drive failure will tell you) is to counteract the torque reaction on the body of the helicopter from the main rotor. It's quite handy to get some idea how much thrust the tail rotor must produce to do this. Let's use the example I gave in December's article with a '30' sized machine hovering with a head speed of 1750 RPM. In this case about 300 watts (0.4 hp) is used in turning the main rotor. At 1750 RPM this takes a torque (turning effort) of about 1.6 Newton Metres (1.2 ft lbs). With the tail rotor some 0.7 metres away from the main shaft it needs to push with a force of about 2.3 Newtons (0.5 lbs force) to give the desired torque.

Main rotor tilt

As an aside, This side thrust would push the heli sideways (to the left for a clockwise rotation machine) unless counteracted by a tilt of the main rotor disk. With a typical '30' machine weighing in at about 2.75 kg the side thrust of the tail is about 8% of the weight of the helicopter and the main rotor will need to be tilted about 5 degrees from the horizontal to counter this. Incidentally, the attitude adopted by the body depends on many factors such as the vertical position of the tail rotor, the lateral and vertical C of G positions, the teeter stiffness of the head etc., but the tilt of the main rotor disk is not affected by these things.

Tail trim in the hover

Coming back to the trimming of the tail, A typical '30' size tail of say 210mm diameter turning at 8750 RPM (5 times the main rotor speed) will generate the required thrust with around 7 degrees of pitch (depending to a degree on how much the fin masks the tail rotor). So this gives us a starting point for the tail trim in the hover. One complication about the hover tail trim is that it is RPM dependant. If the hovering head speed is reduced say by 10% (from 1750 to 1575 rpm) the power being used to drive the main rotor is reduced because of a reduction in the profile drag power (that used simply to push the blades through the air) but the Induced power (used throwing air downwards) stays unchanged. So the torque needed to turn the main rotor falls by about 11% while the fall in the tail rotor speed reduces its thrust by 19% giving about an 8% shortfall. To make this up the tail needs to have a higher pitch angle.

The moral of this is get the collective pitch/throttle set-up right first then start working on the tail rotor trim. One useful aspect of this is the way it can be used to confirm your suspicions about a fall off in engine performance. If, with a previously well set up machine, the motor sounds sick and the nose is trying to swing left (right for an anticlockwise rotor) then its a fair bet the revs are down.

Revolution Mixing

Next let's see what we need as we depart from the hover. At full power let's assume we have our pitch/throttle sorted so the revs are close to those we get in the hover. The extra power is then being transmitted to the main rotor as an increased torque. Let's assume our example '30' has a motor capable of giving 700 watts (0.94 hp) to the main rotor while having enough left over to drive the tail and the gear losses. With this amount of power we will, at 1750 rpm have a torque at the head of 3.8 Newton metres. The tail pitch will need to be increased to about 13 degrees to balance this.

In 'normal' pitch range the tail rotor pitch needs to reduce as the throttle is closed (and the collective pitch is lowered). A fairly obvious point here is that when the throttle is closed to tickover (or at least to the point at which the clutch disengages) the engine will be applying no torque to the main rotor so the tail rotor pitch should be zero.

There are several ways of dealing with these changes of tail trim with throttle position. Even simple heli radios have some form of automatic tail compensation. The simplest just provide mixing between collective pitch and tail rotor pitch (often called Revolution Mixing). Again, even the basic transmitters allow for two rates of mix; one below half stick and the other above half stick. The traditional method of adjusting these mixes is to indulge in meteoric climbs and descents adjusting the 'up' mixing rate to keep the heli straight in the climb and setting the 'down' mixing rate to keep it straight in the descents.

Throttle-Tail rotor mixing

All this sounds very reasonable but there is a fairly fundamental limitation of the collective to tail rotor mixing principle which becomes apparent when we consider inverted flight. If we have a 'V' shaped throttle curve so that upright hovering can be achieved with about +4 degrees collective and 50% throttle when the stick is at about the 3/4 stick position while inverted hovering is achieved with -4 degrees collective and 50% throttle at about the 1/4 stick position. The problem for our revo mix is that for both upright and inverted hovering the tail rotor pitch needs to be the same (or very nearly so). This is something a simple revolution mix can't give. I think there is another problem with revo mixing, at least in the way its implemented on the transmitters I've looked at. They use the mid-stick position as the neutral point for the mix. This causes a lot of flyers to get a fixation about getting the hover point at mid stick. This in turn causes them to set up much bigger collective pitch changes (maybe 10 degrees) for the bottom half of the stick movement than for the top (maybe only 4 degrees).

Another curious result of the mid stick reference point is that changing the collective pitch at the hover point will change the collective throw in the top half of the stick movement. This in turn changes the amount of collective mixed into the tail as you push the stick to the top. So, guess what, you now have a different tail rotor pitch at the full power position, and you get a change in the tail trim at full power when you haven't been playing with that part of the set-up at all! Complex transmitters such as the Futaba 9ZHP, JR PCM10SXH and the like can be programmed to have different revolution mixes for each idle up state, so you can, with effort, engineer what you want.

However I prefer a simpler solution and that is to employ throttle to tail rotor mixing. After all, its the engine that's producing the torque you're trying to cancel out! On a Futaba Field Force 7 for example I inhibit the revo mixing and set up the 'P-MIX' free mixer to mix the throttle channel into the tail rotor channel. As with revo there are separate mixing rates for the upper and lower halves of the throttle movement. Having established the tail rotor trim for the hover I then adjust the mixing rate for 0 to 50% throttle so that the tail pitch is reduced to zero as the throttle is closed to tickover. I then adjust the mix for 50 to 100% throttle openings to maintain acceptable tail trim in full power forward flight (I'm not much given to doing full bore vertical climbs so I don't see the need to trim for them.) Now, since the throttle is in the same position for inverted manoeuvres as it is for the equivalent upright manoeuvre the throttle to tail rotor mix gives the same tail rotor pitch in both cases which is what we need. It may not be right but at least it will be wrong in a way that's consistent between inverted and upright flight!

There is an even simpler solution; the one adopted by Bob Johnston (I know, I'm starting to sound like Bob's publicity manager). He uses no form of tail compensation but instead does it all by pure pilot skill. (Sickening isn't it!)


Editor's Footnote: It's rumoured that Colin is working on the next interface for his simulator, which involves no transmitter at all, simply a phono plug interface directly to his brain. This means he will be able to `think' a manouevre and the model will perform that same manouevre. I understand there are practical difficulties with one's mind wandering during the flight - and it has not yet been decided where the phono socket will be fitted.....

 

Part 6 (Originally published March 1996)

 

The Tail Rotor Gyro

Last time I was considering the set-up of the tail rotor, and in particular, the various automatic methods for adjusting the tail rotor pitch to compensate for main rotor torque variations that occur between hover and descent and between hover and climb. Perhaps the most important idea to come away with is the impossibility of the 'perfect' tail compensation arrangement and I think its fair to say most set-ups rely fairly heavily on the performance of the tail rotor gyro to iron out the remaining imperfections so its worth looking at the basics of the gyro.

The basics of the tail rotor gyro

Without a gyro, the 'natural' handling characteristics of a helicopter tail rotor is (as anyone who learned in the days before gyros will tell you) rather unpleasant. This is because the natural damping of the helicopter in yaw is small so controlling it without the help of a gyro is a bit like balancing a marble on a sheet of glass. The purpose of the gyro is to provide artificial damping of the yaw motion of the helicopter so its more like balancing a marble on a sheet of glass coated with a thick layer of oil.

To generate this effect the gyro needs to incorporate some method of measuring the rate of yaw of the helicopter. This may now be done by a conventional mechanical gyroscope or by the use of a solid state rotation sensor. It could also be done by an optical method called a 'ring laser' though these are so expensive that I don't think they have ever been applied to a model application.

From Figure 1 we can see how the gyro system provides the desired damping of the helicopter yaw. Lets assume for a moment that the system in figure 1 is in trim with the tail rotor thrust just balancing the main rotor torque and the yaw rate at zero. If say a gust of wind disturbs this balance and the helicopter begins to yaw. The yaw rate sensor detects the turn and produces an output proportional to the yaw rate. This is passed via the gyro 'gain' control to the amplifier and on to the tail servo. Now the signal from the rotation sensor is in such a direction that this movement of the tail servo causes a change in the tail pitch that opposes the initial turn. So, if the gust initially causes the helicopter to yaw to the left the gyro system will apply some right tail rotor control to oppose the swing.

What if the direction of the rotation sensor signal is wrong? Well, in this case the response of the gyro to a small swing to the left is to put a bit of left tail rotor control in just to help it on its way! This of course increases the yaw rate that in turn increases the sensor signal that in turn puts in yet more tail control, etc. Very quickly the small swing builds up into a full blown pirouette. So, if you do get the gyro sense wrong it will very quickly let you know! A friend of mine went to great lengths to make sure he had the sense of the tail stick control and the gyro right on a new Concept 60 he was putting together. He checked it over at least five times before finally flying the model. He was therefore more than somewhat surprised when, just at the point of lift-off, the machine went round like a blue-bottle on pyrethrum. How did he get it wrong after all that care? Well, simple really; he put the tail blades on back to front and then used the direction the blades were facing when working out the sense of the tail control and gyro!

Unfortunately, getting the sense of the gyro wrong isn't the only way to get things screwed up. Most beginners are looking for as much help as they can from the gyro. Since the higher the gyro 'gain' setting the more the signal from the gyro resists any yaw motion of the helicopter it follows that beginners are going to want to run as much gyro gain as possible so as to buy as much time as possible. However, if the gain of the gyro is wound up to full its more than likely the tail of the helicopter will go into violent oscillation once the machine leaves the ground. Why? Well the answer lies in the delays in the gyro - tail servo - tail rotor system. The first delay in this system comes simply from the yaw inertia of the helicopter itself. Simply, it takes a certain amount of time for the yawing of the body of the helicopter to respond to any change in tail rotor pitch (and hence thrust). Exactly how long depends on many things - how the weight on the helicopter is distributed along the body, the diameter of the tail rotor, the chord of the tail blades, the tail RPM, etc., etc. However the typical response time is of the order half a second (500 milliseconds). Next comes the delay of the rotation sensor in the gyro which may be anything from 20 to 100 milliseconds. Next comes the tail servo which may contribute a delay of between 50 and 200 milliseconds. There may well be other sources of delay - slop in control linkages, radio system frame rates, etc.

Figures 2 and 3 show the ideal and real cases for the gyro damping. In the ideal case the tail rotor pitch is arranged so that it always produces a force that directly opposes any oscillation of the tail. In this case oscillations are likely to die out very quickly. In the real case the delays in the gyro and servo mean that the actual tail rotor pitch changes occur later than they should and its possible for the tail to spend some of its time actually increasing the amplitude of the tail oscillation. If the gyro gain is high enough any disturbance can set the tail off into a set of self- sustaining oscillations that can be very violent. The normal solution is to turn down the gyro gain until the oscillations don't occur. However there are a number of things we can do to reduce the tendency to oscillation and allow us to gain a higher degree of damping from the gyro before risking these instabilities.

First in my list would be to eliminate any avoidable delays by getting rid of as much slop and 'give' in the tail pitch control linkage as possible. To my mind, this seems to be an area of helicopter design that usually gets less priority than it should. Often the tail servo is mounted at the front of the radio tray and the long tail control linkage goes through some idler arms, some dog-legs in light gauge piano wire, or some poorly located bell-crank (or sometimes all of these). Add to this some stiffness in pitch sliders and feathering bearings and the servo may have to move quite a bit before anything at all gets through to change the tail blade pitch. There is certainly a healthy market in rear servo mount and push-rod kits that are designed to overcome these shortcomings in the standard designs and one of these may be a worthwhile investment even for the beginner.

Tail servo arm length

Another effective delay reducing measure is simply to use a large radius arm on the tail servo. In fact I set up my tail linkage so that the full travel of the servo cannot be used without the linkage binding. This apparently dangerous arrangement relies on the simple fact that, in flight, the full authority of the tail control is not transmitted through to the servo because, as soon as a tail command from the stick starts to take effect the gyro senses the yaw and opposes the authority of the stick thus reducing the servo movement. The advantage of this arrangement is, for a given speed of servo, the tail pitch can be changed more quickly than with a shorter servo arm so minimising the delay introduced by the servo. Increasing the length of the tail servo arm increases the authority of both the gyro and the stick. As far as the available yaw rate is concerned the two effects are competing so, if nothing else is changed, the two pretty much cancel each other out. Remember that, by increasing the servo arm length you also effectively increase the gyro gain and as a consequence it may at first seem that the longer servo arm has made the tail stability worse. However, if you subsequently have to back the gyro gain control off somewhat to prevent oscillation the overall effect should be an improvement in the tail stability. Next time I'll be looking at some other gyro defects, their consequences and some likely areas for gyro improvement.

Until next month

Colin Mill

 

(Parts 5 & 6 Originally published February / March 1996)

Copyright Colin Mill and Lance Electronic Publishing 1996/7/8