By continuing to use this site, you agree to our use of cookies. Find out more
Forum sponsored by:
Forum sponsored by CML

Single servo ailerons

All Topics | Latest Posts

Search for:  in Thread Title in  
Martin Harris15/08/2019 18:13:28
avatar
8741 forum posts
214 photos
Posted by Jon - Laser Engines on 15/08/2019 17:01:14:

Unless the servo fails at anything other than zero deflection, then its game over :D

...which is a pretty good reason for having a second servo!

Bob Cotsford15/08/2019 18:35:53
avatar
7931 forum posts
436 photos
Posted by Martin Harris on 15/08/2019 18:13:28:
Posted by Jon - Laser Engines on 15/08/2019 17:01:14:

Unless the servo fails at anything other than zero deflection, then its game over :D

...which is a pretty good reason for having a second servo!

Didn't help my WotsWot when it failed at less than 50' after take-off, maybe a better flyer could have saved it but I couldn't!

Chris Freeman 316/08/2019 07:38:00
235 forum posts
305 photos

In the 70's and 80's servo' s were very expensive, had little power and many more components to them than todays servo's. The receivers were also nat as todays ones and extension leads and y leads were not used as these caused interference. The transmitters did not have any programming so all the instalations were manually set up, if the controls were reversed you had to use a reverse servo or change to to opposite side of the servo if possible. If you needed more movement you had to use a longer servo arm of move in on the control horn on the control surface.

Many people now use multiple servo's and the servos are of a better quality and power and the radio's make it far easier to sep up the aircraft. I have had very few servo failues in my aircraft and most could be traced back to vibration, poor set up or stiff controls. I often move a control surface via the pushrod and wonder how the poor servo copes with having to do this. Power source can also contribute to failures so on my larger electrics that use more than 4 cells or 5 servos then I will have a seperate power source for the radio and not the ESC / BEC

Jon - Laser Engines16/08/2019 08:24:33
4775 forum posts
179 photos
Posted by Bob Cotsford on 15/08/2019 18:35:53:
Posted by Martin Harris on 15/08/2019 18:13:28:
Posted by Jon - Laser Engines on 15/08/2019 17:01:14:

Unless the servo fails at anything other than zero deflection, then its game over :D

...which is a pretty good reason for having a second servo!

Didn't help my WotsWot when it failed at less than 50' after take-off, maybe a better flyer could have saved it but I couldn't!

That was the point i was trying to make. Its not a criticism of you Bob, but even with a spare aileron servo in place the model still crashed. Depending on the situation it might have been recoverable, in which case the accident falls into the 99% pilot error category, but even if it was the shock of something actually going wrong means that even very good pilots may freeze and/or make bad choices when presented with an expected problem.

Also, just to really go for it what would have happened if the model only used the one good servo? more servos means more points of failure and lower reliability.

I am playing devils advocate here a little. Per surface servos on ailerons are common and so much easier than a 2 foot torque rod out to the ailerons at the end of a 40 inch wing panel. This is especially true when there is a flap in the way and the whole installation would be far more secure. There is a theoretical safety margin should one servo fail, but failure is more likely due to increased complexity.

To get back to the op however, using twin servos will not help him one little bit.

Martin Harris16/08/2019 10:34:33
avatar
8741 forum posts
214 photos

Surely Bob was saying that one (presumably) aileron servo failed at 50' and the model crashed due to the failure being the root cause? While steering on rudder is a possibility and a useful skill to practice (saved a model that for complex but inexcusable reasons I flew with reversed ailerons - but then it was a Nova trainer with plenty of dihedral) most models will have insufficient yaw/roll coupling to compensate for any meaningful aileron deflection due to a failed servo.

Had Bob been flying with a single servo, it would have been half as likely to have failed on that flight (arguably slightly more than half due to it having been worked harder over its life in that model) but that failure which must have locked the aileron in a deflected position would have been almost inevitably fatal to the model in the single servo scenario. Had Bob been slightly luckier, he might have had a little more time to orient with the instantaneous trim changes experienced and got his model under control.

I've quoted 3 examples from my direct experience where models have been saved by the second servo continuing to function after a primary control surface failure - "crash investigation" is often unable to conclude whether a servo failure was the cause or result of a crash but these incidents were obviously diagnosable. On these results, 2 servos are statistically safer although of course the sample size is ridiculously insufficient for proper conclusions to be drawn.

We have no chance of eliminating all failures or pilot errors leading to the loss of our models (if Airbus or Boeing can't make failure or pilot error proof aircraft what chance do we have?) but easily built in redundancy (especially where it can coincidentally improve efficiency and handling) seems to be a win/win situation.

Actually, my statistical analysis has improved in support of dual servos as I've just recalled another example with my fairly large Mahers Pacer where slightly odd handling was traced to an intermittent fault in an aileron servo which was found after landing for investigation. It failed completely during testing!

Edited By Martin Harris on 16/08/2019 10:35:12

Nigel R16/08/2019 11:32:18
avatar
2985 forum posts
471 photos

re Single / Dual servo failures;

It is not a straightforward "half as likely" proposition.

The big boys do fault tree analysis and put a probability on each outcome. I'm going to do a contrived example.

A single servo is straightforward:

1 it fails at neutral

2 it fails with control applied

3 it works!

in scenario 1, you can probably fly down on rudder
in scenario 2, you are going splat
in scenario 3 everything is good

Let's assign some exxagerated and made up chances of each.

A 1% chance of each type of failure. And thus 98% chance of everything working.

For dual servos there are now 9 combinations of outcome, and if we keep the same probability of individual servo failure, the chances work out thus:


1 left fails at neutral, right works - 0.98%
2 left fails with control applied, right works 0.98%
3 right fails at neutral, left works - 0.98%
4 right fails with control applied, left works - 0.98%
5 left fails at neutral, right fails at neutral - 0.01%
6 left fails at neutral, right fails applied - 0.01%
7 left fails applied, right fails at neutral - 0.01%
8 left fails applied, right fails applied - 0.01%
9 everything works! - 96.04%

and you may note, the probability of everything working is decreased a bit.

looking at those outcomes from a redundancy point of view:

scenarios 1, 3 and 5 can probably fly down on rudder; probability now 1.99%
scenarios 2, 4, 6, 7 and 8 are a bust; probability now 1.99%
scenario 9 everything is good 96.04%


To conclude:

It's now less likely that everything will work ok - although not by much.
The chance of an outcome where you can fly down on rudder is a bit better.

However. The big drawback is that the chance of an outcome where you get a catastrophic fail and can't fly down is also now twice as likely.

Now, my made up chances are not real. But they servo to show how the probability of success or failure is altered.

Still liking dual servos on the grounds of redundancy?

Edited By Nigel R on 16/08/2019 11:33:49

Simon Chaddock16/08/2019 12:03:15
avatar
5419 forum posts
2835 photos

I bet the OP is beginning to wish he had never asked. wink 2

Nigel R16/08/2019 12:06:00
avatar
2985 forum posts
471 photos

we're definitely guilty of thread drift here!

Martin Harris16/08/2019 12:27:37
avatar
8741 forum posts
214 photos
Posted by Nigel R on 16/08/2019 11:32:18:

Still liking dual servos on the grounds of redundancy?

Yes.

Your logic assumes that with a deflected control you will almost certainly crash. Having experienced a detached servo connection during a roll leaving an aileron at full deflection I know that it is possible to land under control with significant aileron deflection on a typical sports model and although I regard myself as reasonably competent I am by no means the best pilot in the club. I ended up holding opposite aileron and a large elevator input to maintain level flight.

Although your logic seems to work on first sight with your exaggerated figures, the real word chances of a double failure within a time frame that doesn't allow a landing are so small as to be disregarded. Your analysis does not allow for any differentiation between a double failure at some point within the 7-10 minute window of a typical flight and a virtually instantaneous double failure.

Nigel R16/08/2019 13:24:15
avatar
2985 forum posts
471 photos
Posted by Martin Harris on 16/08/2019 12:27:37:

Your logic assumes that with a deflected control you will almost certainly crash.

It was a stated assumption for the sake of a simple example.

The 'logic' itself makes no assumptions of anything.

I did state it was "a contrived example".

Posted by Martin Harris on 16/08/2019 12:27:37:

Having experienced a detached servo connection during a roll leaving an aileron at full deflection I know that it is possible to land under control with significant aileron deflection on a typical sports model and although I regard myself as reasonably competent I am by no means the best pilot in the club. I ended up holding opposite aileron and a large elevator input to maintain level flight.

Well that's one possibility. I didn't split 'fail deflected but manageable' and 'fail deflected, unmanageable'. Feel free to improve and extend!

Posted by Martin Harris on 16/08/2019 12:27:37:

Although your logic seems to work on first sight with your exaggerated figures, the real word chances of a double failure within a time frame that doesn't allow a landing are so small as to be disregarded. Your analysis does not allow for any differentiation between a double failure at some point within the 7-10 minute window of a typical flight and a virtually instantaneous double failure.

It's not my logic, I didn't invent it, it is just how mathematical probability works. Really the point of the example is to show that sometimes the resultant chance of an outcome is not what you might expect at first glance. i.e. the argument is not as cut and dried as "two good, one not so good".

If you wanted to turn my simplistic example into something closer to the real world - please do - add some outcomes, assign some more realistic probabilities, re-do the math and post the result.

Martin Harris16/08/2019 13:34:52
avatar
8741 forum posts
214 photos

No thanks Nigel - I'm happy with my empirical logic and life's too short!

Jon - Laser Engines16/08/2019 13:36:43
4775 forum posts
179 photos
Posted by Simon Chaddock on 16/08/2019 12:03:15:

I bet the OP is beginning to wish he had never asked. wink 2

lol too true.

I hated probability at school so getting into the nitty gritty is of no interest to me at all.

The point i was trying to raise earlier is that to claim dual servos on ailerons is for redundancy is a flawed argument if you dont double up everything else (someone else said this too).

No matter what you do there are still single point failures in the system and the biggest of those is the pilot. With that in mind, and considering the reliability of well maintained equipment vs the reliability of the average pilot the entire redundancy argument is almost a moot point.

Nigel R16/08/2019 13:47:01
avatar
2985 forum posts
471 photos

Indeed it is Martin smiley

Ultimately, servos are pretty reliable beasts, particularly if you avoid the chaff.

I have models with multiple servos in the wing and I'm perfectly happy to fly them!

Making our installs mechanically good is far more important than arguing the toss over the tiny chances of different types of failure of our varying setups - we're just shooting the breeze here smiley

As a thing to note, the only servo failures I have ever experienced involve undersized servos and big surfaces. Mechanical mistakes on my part, in other words, during the install. Or, pilot error leading to a slight donk with subsequent damage that later caused complete failure.

RC Plane Flyer16/08/2019 14:15:51
624 forum posts
22 photos

I am reading all the responces thank you all. Appreciate I did have a control set up slightly higher than should be without my usual expo set up for my flying routines.

And without going off on another tangent is more up than down a better set up with a single or two servo set up ?

I have over 20 plus years building and flying IC models and tamed them all but this one is not to my liking as yet

Thanks again for all your inputs

Nigel R16/08/2019 14:27:59
avatar
2985 forum posts
471 photos

"is more up than down a better set up with a single or two servo set up "

are you referring to the acrowot? if so I'd aim for equal movement in both directions

you can do differentially easily with either a single servo or dual servos

for a single servo, use a disc, and a pair of holes offset forward (low winger) or backward (high winger)

for dual servos, you can offset the servo arm, or program it in the tx

gangster16/08/2019 14:29:51
avatar
941 forum posts
16 photos

RC Plane Flyer Sorry you are having issues setting up this plane, some are just like that...but I believe you still need to stick to the basics.. The Acrowot dates back to the time when people were flying basic four channel radios and it has always has had a reputation for doing that All computers, bells and whistles and fancy words will not solve the issue if it basically isnt right Nothing wrong with a single servo, we have used them for years, put something half decent like a Futaba 3001 or hitec equivalent in and it will be reliable.. With a good basic set up get it flying perfectly then you can gild the lily if you wish by playing with your tx settings If it wont fly a base set up eg like a good old Futaba M series then it will never fly with the most expensive TX in the world.

Martin Harris16/08/2019 14:32:05
avatar
8741 forum posts
214 photos
Posted by Jon - Laser Engines on 16/08/2019 13:36:43:

The point i was trying to raise earlier is that to claim dual servos on ailerons is for redundancy is a flawed argument if you dont double up everything else (someone else said this too).

No matter what you do there are still single point failures in the system and the biggest of those is the pilot. With that in mind, and considering the reliability of well maintained equipment vs the reliability of the average pilot the entire redundancy argument is almost a moot point.

Moot point? Four models I can think of would probably not exist today if the servo that failed was the only one fitted. Despite the convoluted and almost certainly incomplete statistical analysis modifying the conclusion by a few percentage points, the likelihood is that this would have been the case in two of them.

Yes, even a well thought out model aircraft installation with redundancy will have numerous single points of failure but having redundancy on primary flight controls operated by complex electronic/mechanical systems which can't be easily checked internally and can fail instantaneously gives you a fighting chance if one fails. How do you "well maintain" your servos? Functional checks will pick up gradual failures and some intermittent faults but even replacing them on a service life basis introduces the possibility of early life component failure...

I don't use independently controlled surfaces on all my models but where I can do so practically, I would always opt to do so.

Nigel R16/08/2019 14:37:33
avatar
2985 forum posts
471 photos
Posted by Martin Harris on 16/08/2019 14:32:05:

Despite the convoluted

It really wasn't.

Martin Harris16/08/2019 14:40:50
avatar
8741 forum posts
214 photos

Maybe not but it was almost certainly incomplete.

Nigel R16/08/2019 14:44:49
avatar
2985 forum posts
471 photos

It was a contrived example.

It is far from complete.

All Topics | Latest Posts

Please login to post a reply.

Magazine Locator

Want the latest issue of RCM&E? Use our magazine locator link to find your nearest stockist!

Find RCM&E! 

Support Our Partners
CML
Addlestone Models
Gliders Distribution
Wings & Wheels 2019
Pepe Aircraft
Cambridge Gliding Club
Slec
electricwingman 2017
Advertise With Us
Sarik
Latest "For Sale" Ads
New Poll - Sticky situations...
Q: How often - when using superglue - do you end up with it on your fingers?

 Every time
 Occasionally
 Sometimes
 Rarely
 Never
 Wear rubber gloves

Latest Reviews
Digital Back Issues

RCM&E Digital Back Issues

Contact us

Contact us