Exponential??????

[Bill Brown 3]

Hi guys & gals,

could any of you learned greater than i people please explain to me in leymans terms how exponential setting of your t.x. assis`ts with the control function?

I have read somewhere that within our field of use it "softens" the midrange controls, if this is the case i would assume that it would not be of advantage to the likes of the aerobatic or similar fliers who would demand an instant response on the controls, (possibly i could be wrong on this) researching the word gives an awful lot of mathematical jargon which does not enlighten me but makes me think it is another way of describing the theory of relativity.

Me , being of simple demands, would just like to know if it would enhance my flying experience and if so, in what way?.

Please don`t be shy, feel free to give me your angle on this.

I thank you in anticipation of your response.

Confused dummy

[Former Member]

[This posting has been removed]

[Dave 8]

Hello Bill,

Your own description is accurate enough. Some people like instant response regardless of what model they fly be it something tame or wild. Some flyer's use it, some don't. It is a matter of personal preference. I personally like a bit of expo to tame my twitchy fingers although I did learn without expo. As Percy suggests, try it yourself. No harm will come to your plane and it will not do anything weird. Just gain a bit of height and play around with the values until you come up with something you like.

[David Ashby - Moderator]

Bill, pop the word in the search box and you'll find some threads that should help

[cymaz]

This is quite good

[GrahamC]

If you do try it on your favourite model remember that the softening around the centre effect is achieved with minus expo on Futaba and (I think Hitec) positive expo on Spekrum.

One way to observe the effect clearly is to set one of your flying surfaces to 100% and -100% expo and you will be able to see what it is doing as you move the control. Just don't forget to resort it to a sensible setting before you go flying!

[Bob Bertram]

Cymaz,

As a newbie to all this and having had a Spektrum DX6i delivered yesterday, I found that article really good. Thanks for posting the link.

Bob

[cymaz]

Bob..........my halo is slipping

[Andy48]

If you've got a flight sim package which uses your own transmitter, that is the place to experiment with expo.

[Bill Brown 3]

Many thanks all, especialy cymaz for the link, dead interesting and have saved it in my docs. Will definitely have to have a play.

Happy flying all! (may all your prangs be from low altitude)

Bill

[Frank Skilbeck]

Bill I think for conventional aeroibatic planes you are probably correct, but for the 3D type models which have very large (i.e. >45 Deg) movements for the prop handing manouvers the controls then become hypersensitive when flying normally, so expo helps here.

I have a couple of warbirds as well that like a lot of elevator after touch down to keep the tail down, but without expo they would be very sensitive in fast forard flight.

I usually start these days with 30% expo and then reduce it if I think the model is not responsive enough.

I must admit having learnt in the days of basic radios (with left and right servos) I'm all for all the electronic help I can get

[Tom Sharp]

Definately a help for us shakey elderly pilots.

[Biggles' Elder Brother - Moderator]

I think the main benefit of expo lies in almost giving you the best of both worlds. At lot of the time you want very small control movements, if the response was linear (meaning equal stick movement gave you equal amounts of control movement throughout the range) then those small movements you need in most situations would be very tiny indeed at the stick level. For example, if you are flying a slow roll you want very small small corrections - and that's going to be difficult to do if we are talking about movements that amount to "breathing on the stick" almost! Ironically with a very aerobatic model this is often even more true as they can be very twitchy.

But sometimes you do want big control movements - a snap roll, a flick roll, a stall turn for example. They all call for big control movements to be put in very rapidly.

So we have a contradiction - sometimes we want small movements of the controls to match up to stick movements that are not so small, but at the same time we want "full scale deflection" to be quite large! Expo delivers exactly that. It "softens" the centre so that it is "desensitised" and allows managable degrees of stick movement for small, fine, control inputs. At the same time, as the stick movement ends the end of its travel, the "gearing goes up" and we get the big throws we want as well. One of the very few examples in life of having your cake and eating it!

Will it help you if you are learning? Probably yes. It will tend to make your flying smoother and help you "feel" the controls in "normal flight small movement conditions" rather better. As Dave says above - give it a go.

BEB

[Bill Brown 3]

Thanks again all, the fog is clearing.

[Ed Anderson]

 

Dual rates and exponential allow you to change how responsive the plane is to your stick movements. If you have them set-up on a switch, you can make these changes while the plane is in flight. This might be useful as you move from take-off to normal flight. Perhaps an instructor has a trainer plane she would like share between new pilots and more experienced pilots. It would be convenient to be able to change the plane's behavior depending on the pilot without having to move the linkages.

 

 

DUAL RATES

 

 

Of the two, dual rates has been around longer and is simpler to understand.

Dual rates are based on changing how much a surface can move. Let's use

rudder set-up to illustrate this.

 

 

If your instructions say to set 1" of throw left and right, that would be

the recommended surface movement at full stick movement. When you move the

stick 1/4 of the way, you would get 1/4" of rudder movement. At 1/2 stick

you would get 1/2" of rudder movement. You get a direct, proportional and

linear relationship between stick movement and surface movement. At 100%

stick movement you get 100% of the maximum surface movement that you have

set. In this case 100% stick equals 1 inch.

 

 

With dual rates we can change to a second maximum at the flip of a switch.

Let's assume you have the standard throw set as the high rate. Then, using

the procedures outlined in your manual, you set low

rate at 50%. At this setting, when you move the stick all the way over you

will only get 1/2" of surface movement. However stick movement and surface

movement remain proportional. So at 1/2 stick movement we will get 1/2 of

the 1/2 inch maximum or 1/4 inch of surface movement. Your rudder movements

remain directly proportional but are now based on a smaller maximum.

 

 

We can say that control and response are both proportional and linear. That

is, all the way through the stick movement the rudder will move with us in a

linier fashion. If we move the stick 20% we gets 20% rudder. Move the stick

62% and the rudder will move 62% rudder movement. If we plotted a graph with

stick movement on one axis and rudder movement on the other, the graph would

have all points along a straight line at a 45 degree.

 

 

How does this effect the handling of the plane?

 

 

Continuing the example above, we have high rate, at full stick movement

equals 1" and low rate set at 1/2" maximum rudder movement.

 

 

On low rate, for each small movement of the stick, we get less movement of

the tail surface. So, on low rates the plane will be less responsive to the

same amount of stick movement. This may make it easier to fly as we can

make smaller adjustments when we move the stick. We have finer grain

control.

 

 

On high, we get more movement of the rudder for each unit of movement of the

stick. We get a faster response from the plane for the same stick movement.

If you have ever worked with a precision tool or instrument, this is like

having course adjustment and fine adjustment.

 

 

As new flyers often have a tendency to over control the plane, it is not

uncommon to set-up the plane with smaller throws so that the pilot is less

likely to get in trouble by over controlling the plane. Later when she gains

confidence and the right feel for control, surface movements can be

increased to make the plane more responsive. Originally this had to be done

on the plane. Many RTF planes come set-up this way. They are set for mild

response for initial flights. Then the manual explains how to increase the

rates as the pilot gains experience. Some RTF planes now include a dual

rate style control on their radios.

 

 

With dual rates on the radio, this can be done at the radio rather than

working on the plane itself. This is much more convenient. Dual rates can

even allow the instructor to take control, flip to high rates and pull the

plane out of a tough situation that the student could not handle. Dual

rates can be very helpful during training.

 

 

Of course we can always have it the other way where the low setting is the

"standard" recommended by the instructions and a high setting might be our

aerobatic setting or our 3D setting where we want 1.5" of deflection at full

stick. This allows us to take the plane from mild to wild at the flip of a

switch. 

 

Edited By Ed Anderson on 05/03/

Edited By Ed Anderson on 05/03/2013 00:43:54

[Ed Anderson]

 

EXPONENTIAL

 

 

Exponential changes the relationship between stick movement and surface

movement. When using exponential, stick movement and surface movement will

no longer be linear. What does that mean?

 

 

Exponential is going to allow us to shift some of the rudder response so

that we get a different amount in the early part of the stick movement as

compared to the later part. Let's stay with the rudder example above.

 

 

At 100% stick movement we would still get 100% surface movement, but at 50%

stick movement we might only get 25% rudder movement. This would be like

having low rates on the first half of the stick travel and high rates on the

second half of the stick travel. That would give us a "softer" response

around the center of the stick area, and a faster response toward the end.

 

 

How is this beneficial?

 

This gives us finer control when we are making those

typical small adjustments to the plane when we are cruising around, just

like low rates. However if we suddenly want a big surface movement to get

out of trouble, to respond to a gust of wind or to perform that big stunt,

we

still have the big surface movements we need without having to manually

switch to high rates. One of the criticisms of using a low rate for

"flyability,

is that it limits the pilot's ability to get out of trouble when you are on

low rate.

 

 

Let's look at that aerobatic or 3D pilot we mentioned above. He has BIG

surfaces and BIG throws set which makes the plane very responsive to small

inputs. If he were to set exponential rather than dual rates, then he could

have a very soft center to the stick. He could make fine adjustments when

needed, but get big response when he needed it and there would be no need

to flip a switch during the flight. Cool?

 

 

Let's try some examples that involve numbers. The numbers I am going to use

may not map directly to your transmitter as different manufacturers have

different interpretation of exponential and what the numbers mean, but the

overall impact on flying is the same. They just express it differently.

 

Let's say that under standard set-up conditions exponential will be

expressed as zero. This means we have the same linear response we have

always had. Now, if I put in -50% exponential, that might mean that for a

50% movement in the stick I only want to get 1/4 surface movement but

when I move the stick to 100% I want full 100% surface movement. An

input into the set-up menu of +50% might mean that for the first half of the

stick movement I want more of the total surface movement. This would

make the center area very responsive while leaving find grain control at

the ends of the stick movement. I am not sure where this would be used,

but that is how it would work.

 

 

It is important to note that exponential does not imply a sudden change in

rate. Rather it is a smooth change in rate. So the further we move the

stick, the faster we get more stick movement. If we were to plot the percent

stick movement to percent surface movement we would not get a straight

line as we normally get. We would get a curved line indicating that the

further we move the stick the less linear the relationship between the stick

and the surface.

 

This is one of those things you are just going to have to try to fully

understand. At first it seems it would make it difficult to predict how the

plane will behave depending on how much you move the stick. However in fact

most people tend to fly more by input/response rather than where the stick

is in its travel. You move the stick and watch the plane. After a while you

develop a good understanding of how the plane will respond to a given stick

movement, but you know that it will be influenced by wind, air speed, and

other factors.

 

 

I typically set up a switch with about 35% exponential so that I have a

softer response around the middle if I want it. That gives me gradually

faster response as I move toward the extremes of stick movement. On my

radio I have dual rates and exponential available and I can use them

together. I can also set them by surface.

Edited By Ed Anderson on 05/03/2013 00:41:32

[Bill Brown 3]

Ed, Thanks very much for taking the time to explain in so much depth, your method of explanation is very easy to follow, now i think i could also explain exponential if i were asked.

Best regards

Bill

[Steve Dorling]

Small point - depending on how your mechanical links are set up - you're probably (almost undoubtedly...) flying with expo anyway (before you start with electronic tinkering)!

[Ian Jones]

Yes that's right Steve. Assuming the full range of the servo arm's movement is being used then the starting point is the equivalent of having a degree of positive (on Futaba) expo in. There is more movement around the neutral stick position. 1/2 the stick movement does not = 1/2 the surface movement. Half the stick movement does = 1/2 the servo arm movement. To get the stick movement and the control surface movement to proportionaly match each other some exponential adjustement is required.

I have explained this in detail in other threads but in short it is because the point at which the control rod is attached to the servo output travels in an arc. By the time the arc reaches the end of it's range it is moving the attachment point sideways as well as in the direction of the control rod so less of the servo movement is contributing to the control surface movement.

Adding or subtracting expononential changes the multiplier by which the position of the servo arm is calculated relative to the amount of stick input that is made.

So, a small amount of expo can be described as softening the stick movement around the nuetral position but this description is misleading if you want to understand what is actually happening which with a relatively small amount of expo adjustment is that the the control surface movement is being made to be more proportional to the amount of stick movement. Larger amounts of exponential adjustment take this effect to more extreme degrees eventually resulting in little movement around the neutral point and lots more at the end of the stick movement.

Probably the best way of getting to understand this is to use a spare memory slot on your Tx and dial in 100% expo on a control surface then slowly move the stick - you will soon see the effects of expo adjustment. On Spektrum TX's you can only do this on the memory the rx is bound to so if you do this test on a "live" memory make sure you have put it back where you started afterwards!

Edited By Ian Jones on 01/06/2013 10:33:26

[Piper Cub]

Posted by Steve Dorling on 01/06/2013 06:53:31:

Small point - depending on how your mechanical links are set up - you're probably (almost undoubtedly...) flying with expo anyway (before you start with electronic tinkering)!

More than likely, only on aileron though.

[chris basson]

Why only Aileron?

I'm slighlty confused!

CB

[cymaz]

You can use expo on all flying surfaces and throttle (depending on Tx model and make)

[Biggles' Elder Brother - Moderator]

OK, I've heard this argument the Ian makes (previous page) a number of times - but I'm afraid I don't buy it! In the past I've let it go - but I have some time on my hands tonight! So, if you're feeling up for a quick debate on kinematics here we go,...

Basically the argument that Ian invokes is that all controls on our models have an unbuilt non-linearity of movement because the servo output is a rotational arc not a pure translation. The theory goes something like this:

Consider the servo in its neutral position with the arm at right angles to the servo body. It starts to move through, say the first 10 degrees - this produces a certain amount of translational movement of the control rod - let's say 5mm for arguments sake. Now consider the point when the servo arm is well on through its rotational movement - say at 40 degrees to the servo body. If it now moves another 10 degrees the control rod will not move 5mm - it will move much less - say 2mm. This is because the horizontal component of that part of the arc is much smaller. The diagram below illustrates this:

The figure above shows the servo in its first 10 degrees (approx) of rotation thus producing a control rod translation of say 5mm. The figure below shows the same servo - again rotating by about 10 degrees - but this time end the end of its rotational travel and producing a much smaller translational output on the control rod.

So the argument says that as the amount of translational (ie "side to side" movement we get decreases as the servo throw increases the system is inherently non-linear. This much I agree with. In fact we could show that the function that defines the translational movement is of a cosine form and is indeed therefore non-linear.

But,...this only considers half the system!!! In an actual control system there is "the other end" - the end with the control surface attached. Here what happens is exactly the reverse of what happened at the servo. At the servo; rotational movement was converted into translational movement of the control rod via the servo arm. At the control surface; the translational movement of the control rod is converted back into rotational movement of the control surface by the control horn. And (this is the important bit!) that second conversion is non-linear in exactly the opposite sense to the that of the servo - therefore the non-linearity at the servo is exactly cancelled out by the opposite non-linearity of the control horn!

The system is an example of a classical mechanism known as a "four bar linkage". The figure below illustrates the point:

From this diagram I think its obvious that any rotation by the servo arm is matched by an exactly equal rotation of the control surface - ie 10 degrees servo rotation leads to10 degrees of control surface rotation; and this is true through out the servo's range - there is no inherent non-linearity in the motion.

Notice that if the control horn length was different from the servo arm length there would be a scaling effect - ie 10 degrees of servo rotation will map to different amount of control surface rotation - may be 6 degrees say. But this difference would be the same thought the motion - so the motion would still be linear, but over either a greater or smaller range that's all.

BEB

[Alwyn Gee]

Sorry BEB but you have argued yourself into a corner here. Your second diagram says it all. for the same amount of movement at the stick the servo arm moves forward a lesser amount therefor the control rod moves the same lesser amount and the surface will also move a lesser amount. As both the servo horn and the control horn are in the same plane they move equal amounts. So the non-linearality comes from the surface not moving as much for the same amount of stick movement. The article in the magazine very recently regarding expo makes it clear that by reducing the movement around center with expo we are in fact linearising an inherently non-linear control.

For your theory to work the control horn in your third figure would have to be angled towards the servo so that the movement increased as it came towards the centre or right angled position. That of course would then pose a problem in the opposite direction As you correctly say the only influence you can have by adjusting the rod in the various holes is to alter the overall mechanical travel of the control surface.

Thats my vision of leverage as I understand it.

Alwyn

Edited By Alwyn Gee on 30/06/2013 22:38:44

[Biggles' Elder Brother - Moderator]

Sorry Alwyn but I think you're wrong mate. Equal stick movements correspond to equal rotational movements of the servo - equal servo rotation equals equal control surface rotation - basic principle of a four bar linkage. So equal stick movements give equal control surface rotation - QED!

The non constant translation of the control rod is a complete red herring - as its counteracted at the other end!

BEB

PS I think the article you are refering to was written by me! It is talking about a different type of non-liearity entirely. It is describing the effect that the aerodynamic effects of equal control surface movements may not be linear themselves!

Edited By Biggles' Elder Brother - Moderator on 30/06/2013 22:46:03