How much Expo and Why?

[Bob Cotsford]

Jon, even with my Taranis stick springs on max I still have issues with feeling centre position, or rather maintaining centre on one axis while moving the other. Same with my FF9 and 8. Give it a few years and you might find the same! I still have an old Horizon tx with it's Slec sticks, the tension on those is quite incredible yet back in the day they seemed normal. One day I'll have to see about obtaining some of the stronger springs that Futaba sell which apparently fit Taranis sticks - one day when I get around to it.

[Peter Christy]

Posted by Jon Harper - Laser Engines on 28/11/2016 12:24:46:

You can tighten your springs guys. Most radios are adjustable these days and this should help find the middle

Well I've got mine tightened as far as they will go, and they still feel feeble!

People laugh when I hand my old "red box" Sprengbrook around, but I always knew where "centre" was, and had no problem with aileron/elevator interaction!

There's a lot to be said for old transmitters!

--

Pete

[Dave Bran]

What a lot of old.........................................pilots......................

I usually shorten the sticks to minimum and run with the lightest spring that gives centre when released. I don't need the stick to tell me where I want it, I tell the stick!! If I want a true circular movement and interaction the last thing I want is the stick springs forcing "squared" action (Remember when many Heli pilots took the springs off completely??)

Joshing around aside, the length of sticks depends on hand size, thumbs or pinch, distance from edge of case and other case dimensions. My Development Horus arrived perfect for me (though loads complained and adjusted more tension), the Prod Horus has arrived with springs too strong for my liking (Can be adjusted without case removal though).

[Jon H]

Posted by john stones 1 on 28/11/2016 12:32:50:

I wind my sticks all the way down and tighten the springs up, it feels awful when there's no positive feel for the center, long sticks are not superior......hands/thumbs come in differing sizes

John

Long sticks in relation to your fingers. Hows that?

But its interesting you guys all like super stiff sticks. I slackened off the ones on my futaba 8j as they were rock solid. I like to be able to feel the centre but i dont want a notchy feeling as i go over it as it makes stick inputs a touch abrupt.

[Dave Bran]

Posted by Jon Harper - Laser Engines on 28/11/2016 13:26:34:

Posted by john stones 1 on 28/11/2016 12:32:50:

I wind my sticks all the way down and tighten the springs up, it feels awful when there's no positive feel for the center, long sticks are not superior......hands/thumbs come in differing sizes

John

Long sticks in relation to your fingers. Hows that?

But its interesting you guys all like super stiff sticks. I slackened off the ones on my futaba 8j as they were rock solid. I like to be able to feel the centre but i dont want a notchy feeling as i go over it as it makes stick inputs a touch abrupt.

EXACTLY!!

[iqon]

None ......never found the need......thats what fingers thumbs are for.....

[Cuban8]

Light sticks for me as well. I also slide a short length of thin silicon tube over the throttle detent brass strip, so I get a totally smooth action not unlike a quality hi-fi volume slider. Much nicer for heli collective.

[john stones 1 - Moderator]

Yep no ratchet on throttle for me, took it off for helis and prefer it, some of you fellas must have wrong size hands

[Jon H]

I dull my throttle ratchet as much as i can. The old futaba 6ex had a ratchet so severe i almost felt the need to declutch every time i wanted more power!

[Biggles' Elder Brother - Moderator]

Yeap, I removed the throtle ratchet some time ago - when flying MR's. I wouldn't put it back now. But I do add a piece of heatshrink to the temsion spring - this increases the the friction of the contact and just stiffens it slightly - which I prefer as, firstly, I then know that if I put the throttle stick in a given position it will stay there! And secondly I get little or no "cross talk" with the rudder - useful when doing cross wind landings!!

BEB

[Don Fry]

Well thank you gentlemen for my continued education. I made posts above as to which way negative and positive exponential went. Am an ex Futaba user, was used to programming negative numbers in, since about 1995.

About 2 or 3 years ago I migrated over of Spektrum, as models were scrapped, broken or sold. I continued to put a smallish amount of negative exponential in. Never crossed my mind that the two brands go the other way.

It solves one puzzle. I have a small leccy 3D model, half plan and half own design. Powerful and twitchy, and still trimming it. I have it at about 50 to 60% negative exponential. No wonder it is twitchy. Only fly it when I am feeling real good, and calm.

Have just spent an hour removing exponential, positive, mostly IRO 15 a 20 %, from the fleet. I suspect I will have an eye opener next time out.

Thank you again.

[Gary Manuel]

A trick I use to make sure I get expo the right way is to dial loads of expo in (maybe 50%) and see that it is going the right way. In this case the control surface should barely move when the stick is moved around centre, but then move dramatically towards the outside. This confirms that you have got it working the right way. Now you can reduce the expo to a reasonable figure (10%, 20% or whatever), without going back through zero. You now know that you have the right amount AND it's the right way round.

The same process can be applied when adding a mix. Add loads of mix, to see if it is going the right way, then reduce it to whatever you need.

[Rich too]

Posted by Biggles' Elder Brother - Moderator on 28/11/2016 11:17:06:

Well first of all another voice to say Justin is right! Whether +ve or -Ve expo "softens" the centre depends on what brand of radio you use. Fubtaba is -Ve, Spektrum is +ve, Taranis is +ve, JR is +ve (I think?).

Second - this is a very personal thing - I do not believe you can really say there is a definative "rates are better employed than expo" or vice-versa - or that people in general "don't understand" expo! In my experience I've seen models well set up from a "primarily rates" position and equally I've seen models well set up fropm a primarily expo position - both can work well. Its very much a matter of preference and secondly (to a lesser degree in my view) it's a matter of model characteristics. I feel that because I believe it is possible to get a good set uo on any model using either rates or expo as your primary tool, but its probably also true to say that with certain models one route may be easier than the other.

There is nothing at all to stop you uisng both of course!

Personally I use expo primarily - I tend not to use rates. I use expo on elevator and ailerons - I very rarely use it ion rudder, I don't seem to feel the need. The amount I use varies enormously depending on what the model's flying caharacteristics are. For slower, more ponderous, models (biplanes the Cub etc) it tends in general to be a smaller amount - typically 10-15%. In a few cases it could be 0%! At the other end of the scale very fast flying models - deltas etc - I use more - maybe as much as 50% in very extreme cases.

I often have variable amounts of expo available via three position swithches - sort of "rates for expo" if you like! This can be used to address issues where you might want more expo in flight than you do say on take-off or landing.

BEB

I have to agree. For some reason I have always struggled to find the switches on my trans, either for killing the motor, or for retracts, and therefore I avoid using them if at all possible - I have never in 35 years used rate switches!

[Rich too]

Posted by Gary Manuel on 28/11/2016 18:09:25:

A trick I use to make sure I get expo the right way is to dial loads of expo in (maybe 50%) and see that it is going the right way. In this case the control surface should barely move when the stick is moved around centre, but then move dramatically towards the outside. This confirms that you have got it working the right way. Now you can reduce the expo to a reasonable figure (10%, 20% or whatever), without going back through zero. You now know that you have the right amount AND it's the right way round.

The same process can be applied when adding a mix. Add loads of mix, to see if it is going the right way, then reduce it to whatever you need.

Thanks Gary, that's pretty much what I had to do on my WW. As a matter of interest, why would you want the expo set the other way, for MORE movement around the centre?

[Geoff S]

Posted by Jon Harper - Laser Engines on 28/11/2016 12:24:46:

You can tighten your springs guys. Most radios are adjustable these days and this should help find the middle

I agree that long sticks are superior as well but not to the extent that i need one of those tray radios our European..err, friends? like to use. I feel that strapping on a radio like im about to go rock climbing is a little excessive.

I've been using either side pieces (Mux 3030) or a tray for years because I need to rest my hand on something and use finger/thumb on the sticks to get any sort of control delicacy (such as it is in my case ). However, the thought of rock climbing with something sticking out from my waist has never crossed my mind.

I used to be a minimalist, too. In fact I always thought that sailing wearing gloves to protect my hands was akin to having a bath with my socks on (which I once tried btw) but times change.

Geoff

[Peter Christy]

I've been using either side pieces (Mux 3030) or a tray for years because I need to rest my hand on something and use finger/thumb on the sticks to get any sort of control delicacy ......

And this is one of the reasons I've switched to an old-fashioned "cuddle-box" (single stick) transmitter for some of my models. I don't need a tray, I don't even need a neck-strap. It has a nice long stick and strong springs, and it is surprisingly easy to hold on small amounts of aileron for slow rolls that go from horizon to horizon!

The only way to get a tranny like this now is to scour the vintage "for sale" columns. I've been told that Americans would kill for my Ace MicroPro single-stick - but I ain't parting with it......!

--

Pete

[Biggles' Elder Brother - Moderator]

Posted by Rich2 on 28/11/2016 19:35:58:

As a matter of interest, why would you want the expo set the other way, for MORE movement around the centre?

Yes if you were to use, say -ve expo, on a Spekky Tx you'd get a control that was more sensitive around the centre and became less sensitive towards full stick movement. I suppose its just possible to think of reasons for doing that - but I can't at the moment!

Exponential is actually a mathematical function - which can take +ve or -ve values as inputs. It can also be formualted slightly different ways - hence why a) you can use both and b) some manufacturers have it the other way round to others.

BEB

BEB

[buster prop]

Back to the OP which was about exponential. I have some of the PZ UMX models and before AS3X I found them difficult to fly in anything but a flat calm with no expo at all. I had the original micro P51 which was a twitchy little beast and almost as bad were the Corsair and T28 Trojan, all 3 much better with about 30% expo. I also have the newer AS3X micro P51 'Meg' and even that benefits from some expo. Normally, I might use 20% on elevators and ailerons for my power models and gliders. In my experience it certainly helps but I use the minimum needed to smooth things out if you know what I mean.

[Maurice Harvey]

What we are trying to do when we dial in exponential is to convert rotary movement into linear movement. This begs the question, why do we use rotary servos? Is there a company that still makes the old linear servo that was once the norm? After all, the indoor flyers are using them. All that is needed is to up the scale.

[stu knowles]

I have read that dialling in 30% expo is needed to give a linear response to a rotary output servo. It is undoubtedly true that a rotary servo will give greater linear movement per degree of stick movement around the centre that at the extremes of travel (subject to all linkages being set 'square' . 

I have no idea if the 30 % value carries any validity, someone with a maths degree can work it out and report back.

[Frank Skilbeck]

Posted by stu knowles on 28/11/2016 22:32:41:

I have read that dialling in 30% expo is needed to give a linear response to a rotary output servo. It is undoubtedly true that a rotary servo will give greater linear movement per degree of stick movement around the centre that at the extremes of travel (subject to all linkages being set 'square' .

I have no idea if the 30 % value carries any validity, someone with a maths degree can work it out and report back.

No really, if the servo and control arm are the same length then the rotary movement of both will be the same. Conversely a linear servo connected to a control horn on a hinged surface would have natural exponetial.

[Jon H]

Posted by buster prop on 28/11/2016 20:58:36:

Back to the OP which was about exponential. I have some of the PZ UMX models and before AS3X I found them difficult to fly in anything but a flat calm with no expo at all. I had the original micro P51 which was a twitchy little beast and almost as bad were the Corsair and T28 Trojan, all 3 much better with about 30% expo. I also have the newer AS3X micro P51 'Meg' and even that benefits from some expo. Normally, I might use 20% on elevators and ailerons for my power models and gliders. In my experience it certainly helps but I use the minimum needed to smooth things out if you know what I mean.

these little models may need expo as they are super sensitive but also the little servos they have are nowhere near as accurate or incremental as a bigger servo.

That said, I have a umx su26 and some of the small blade helis and never used expo on any of them but I did calm down the rates

[Peter Jenkins]

Posted by stu knowles on 28/11/2016 22:32:41:

I have read that dialling in 30% expo is needed to give a linear response to a rotary output servo. It is undoubtedly true that a rotary servo will give greater linear movement per degree of stick movement around the centre that at the extremes of travel (subject to all linkages being set 'square' .

I have no idea if the 30 % value carries any validity, someone with a maths degree can work it out and report back.

I read a trimming manual written by a very experienced Australian F3A pilot. He gave 35% expo (in the sense of reducing movement around neutral) as providing a linear output to the servo arm when compared with Tx stick movement. Most F3A set ups suggest around 10 deg of movement on elevator and ailerons and around 25 deg on the rudder and 30-35% expo seems to work well with these settings. You tend to move the sticks a long way in F3A to achieve the required smoothness, accuracy and manoeuvre size demanded of this discipline.

Some F3A pilots use expo to give them the fine control needed for manoeuvres with much higher movement to cater for spinning. I find excessive expo difficult to cope with and so I use 3 rates to cover a Spin condition, landing and normal aerobatic flying. It works for me. It is also a function of how accurately you aim to fly. Obviously, in F3A, accuracy is paramount but it also depends hugely on the design of the aircraft. Some designs have no cross coupling in knife edge while others will need some mixes to stop unwanted pitching or rolling. There is nothing to stop you controlling those by yourself but you must also be able to fly the manoeuvre that incorporated the KE accurately as well e.g a rolling loop.

However, the single most powerful method of taming a twitchy model is to move the CG forward till it become less sensitive - reducing control throws and using expo, if needed, also help but get the CG wrong and low throws will not cure twitchy aircraft.

I believe it was Doug Spreng (of Sprengbrook) who said "soft sticks sell transmitters, hard sticks win competitions" - perhaps his views are worth considering.

[Simon Chaddock]

"What we are trying to do when we dial in exponential is to convert rotary movement into linear movement."

Really? Don't control surfaces move around a hinge point. Surely that is also rotary motion!

[Gary Manuel]

Quite right Simon.

Let's also not forget that our transmitter joystick is also rotary.

What we are ultimately trying to do with expo is to get our stick inputs to result in an equivalent control surface movement. Let's assume that our joystick has a movement of 30 degrees from centre to the end stops in either direction. Let's also assume that we set up the end points of our model to give a deflection of 30 degrees from centre. We'll also assume that our servo arm and control surface horn are exactly the same length. This means that the servo is also moving through 30 degrees. 1 degree of movement on our joystick will now result in 1 degree of movement of our servo and control surface. For simplicity, let's assume that our joystick length and control surface length are both exactly the same (say 50mm). This now means that the outer edge of our control surface will move exactly the same amount as our joystick.

Imagine now that our joystick is directly moving our control surface (the radio and servo become irrelevant). In fact, we can imagine that our joystick IS our control surface.

It is important to realise that what we are interested in, for controlling our model is not how many degrees the control surface move, but how many millimetres it moves measured perpendicular to the surface. (A longer control surface on a given wing profile would require less angular movement than a smaller one to achieve a similar aerodynamic effect).

Minor maths alert! - the perpendicular distance that the control surface moves is given by x cos(a), where x is the distance the stick is physically moved and a is the resultant angle of movement. Around the centre of movement, this is cos(0) = 1. Towards the end of the movement (30 degrees) it is cos(30) = 0.866.

What this means is that the effectiveness of the control is reduced by a factor of 1 - 0.866 = 0.134 at its extremities compared with its centre. Exponential is required to make this become linear. The actual exponential figure used will depend on the parameters used by the radio manufacturer and may vary from one make to the other. One could easily see that a figure of 13.4% might just do the trick here?

It follows that ALL models require some amount of expo to make the control surface perpendicular movement become linear. It also follows that the greater the angle of control surface movement is, the more expo will be required. For example, extreme 3D control surfaces moving by 60 degrees, would be reduced by 1 - cos(60) = 0.5. Could this require 50% expo to correct?

I have not mentioned the "feel" of the model yet. This is very subjective and is why the use of expo is a personal preference. Modellers who have been brought up using non computerised radios will be used to flying without exponential and find it strange to fly a model with expo. Others will feel that the model is more twitchy in the centre and feel the need for expo. F3A / pattern flyers and others who are looking for "unnatural" increases in sensitivity around the centre of movement will crank up the expo figure to give them the smooth control around the centre that they are looking for.