Almost since I flew my first control line model (TD-4) in 1960 I began to understand the need for a better way to teach someone how to fly a control line model. I certainly did not "get it" on my first few flights. (thank god for rubber bands and grass) In fact I did not really learn how to fly in a true controlled manner till I was 9 years old. (1963).

As time wore on and I began to wear out I could see that control line was in trouble. That was a long time ago. At that time I had no idea as to how you could make a really easy to fly trainer that would not crash. One day a few years ago while watching Pat Mackenzie fly a foam F3P model in a gym it came to me. If you were going to build a crash proof trainer you had to tell the pilot of the crashing model to straighten his hand out and level the plane. Of course you would never have enough time to do this so I needed a better solution. Why not just seize control from the pilot before things get really out of hand. There was just one long problem, the pushrod. The pushrod had to go. What to replace it with you ask? Why of course, a servo. And why a servo? Because if the bellcrank and the control surface are not connected mechanically I could seize control and redirect the model back to safety regardless of where the pilot's hand was. And thus was born the idea of control line fly-by-wire.

I have a reasonable idea of how to determine whether the model is right side up or upside down and climbing or diving. I think I can make the model take off under it's own control, climb to say ten feet and then let you "play" inside the hemisphere (or control how much of the hemisphere you can play in) then level off and land.

Having the original post of pictures of a bellcrank moved to an area not connected to electric flight without so much as a heads up from the moderator is rude. I put significant effort into documenting the parts to be able to share with all modelers. Now that effort has been undone.

I must say that the negativity towards the idea and the veiled implication that I might be cheating if I used such a model and put in a very good flight has left me questioning whether any further effort in this regard is warranted. You are certainly entitled to your opinions but not to disparaging my character. Judges and fellow competitors I may fly in front of or with may read these posts and form an opinion not based in fact. It is too late to tell me that that was not the intent as the damage has been done.

If Kim Doherty puts in a great flight with a fly-by-wire model (which I am quite capable of doing without this system) then I must be cheating.


Incredibly disappointed,

Kim.

Maybe the moderators are not as skilled as Sparky and could not move just the last part of the thread to the debate-zone.
I would like to see the technical info returned to this section. Sparky can you help?

-Chris

Quote from: Kim Doherty on August 28, 2013, 04:29:17 PM
I have a reasonable idea of how to determine whether the model is right side up or upside down and climbing or diving. I think I can make the model take off under it's own control, climb to say ten feet and then let you "play" inside the hemisphere (or control how much of the hemisphere you can play in) then level off and land.

So do it Kim, we will certainly applaud you. The biggest trouble here is that you take all too personal. It is not about YOU, not about "cheating Kim", it is and always was about RULES for stunt. The problem is that RULES cannot prevent anyone using FBW from cheating. And it is clear that anyone having well working FBW system vill see how to push it further. Now you described your autopilot, Doug wrote about gyro, I wrote about quadcopter robotics, it is eavrything doable with existing technology. We simply 1/ are not able define what is and what is not cheating and 2/ cannot inspect and verify if that defined cheating is or is not used. That is reason why I think the controlls IN OUR ACTUAL STUNT should be mechanic.

As I wrote - I will be happy and I will enjoy some "open class" with all allowed ("all" because we cannot define what exactly), but I am sure that it will be constest of 3 or 4 persons on planet, and I simply want my old friends fly with me longer.

Well said Igor.

As for moving the thread - the original thread had nothing to do with electric power for models.  It had to do with control systems ...

as does this one.

Still Kim's original post is of interest to many of us that are interested in electric powered control line. Let us not begrudge him a chance to show innovations, that he is working on, to those of us interested in them. Putting the thread into the debate section does just that. Persons most interested in his activities are mainly those of us that frequent the "Gettin all AMP'ed up" forum. I personally do not pay any attention to the Debate forum nor do I usually view the forums on construction, finishing etc.

It is (presumably) also of interest to some 'glow only' members who never come here (electric power).  One can always post a short note with a pointer to the thread in this forum.

It is perhaps more suited to either the "Design" or "Engineering"  board.  Even the 'Rules' board would seem to be at least somewhat appropriate, given what the whole thing turned into.

I would agree that moving it to the "Debate" board may have not been helpful to it.  I don't go to that section much either.

Mike, Point taken.

Quote from: Mike Anderson on August 29, 2013, 07:19:15 AM
It is (presumably) also of interest to some 'glow only' members who never come here (electric power).  One can always post a short note with a pointer to the thread in this forum.

It is perhaps more suited to either the "Design" or "Engineering"  board.  Even the 'Rules' board would seem to be at least somewhat appropriate, given what the whole thing turned into.

I would agree that moving it to the "Debate" board may have not been helpful to it.  I don't go to that section much either.

New section created

Quote from: Kim Doherty on August 28, 2013, 04:29:17 PM
I must say that the negativity towards the idea and the veiled implication that I might be cheating if I used such a model and put in a very good flight has left me questioning whether any further effort in this regard is warranted. You are certainly entitled to your opinions but not to disparaging my character.

 I haven't seen anyone suggesting that you are cheating in any of these threads (nor the one before). And in fact, since there are no rules preventing any possible functionality of the system, it's literally not possible to cheat as it currently stands. Using a fully programmed autopilot is current LEGAL and there is nothing to prevent it in the rules.

  What everybody is concerned about is autopilots. As I mentioned when this first came up, *autopilots are going to be banned*. I don't think you are likely to disagree with the point we don't want an autopilot doing autonomous maneuvers while the pilot just stands there.

  The problem remains, as Igor correctly points out, and taking as a given that autopilots are going to be banned - *how* are you going to ban them? One way certain to work is to require mechanical connections. It's a crude response, it unnecessarily disallows your fly-by-wire system that most of us think is OK, but it certainly bans autopilots effectively.

 As previously, if you don't want your fly-by-wire system to get banned it would be worth your while to put efforts towards a rule that permits yours, and still clearly prohibits autopilots. I think that is more important to your cause than developing the fly-by-wire system further. I have been thinking about it for ~15 years now, and I still can't come up with anything better than "mechanical connections" but unless you do, your system is going to get banned, too.

  Brett

Quote from: Brett Buck on August 29, 2013, 04:46:11 PM

  What everybody is concerned about is autopilots. As I mentioned when this first came up, *autopilots are going to be banned*. I don't think you are likely to disagree with the point we don't want an autopilot doing autonomous maneuvers while the pilot just stands there.

  The problem remains, as Igor correctly points out, and taking as a given that autopilots are going to be banned - *how* are you going to ban them?

 I have been thinking about it for ~15 years now, and I still can't come up with anything better than "mechanical connections" but unless you do, your system is going to get banned, too.

  Brett

Brett,

There is no question, without reservation that I would not support any autonomous flight path control capability.

Maybe the first step is to define exactly how far we are prepared to go in the non mechanical linkage direction. For me, the only part remaining would be to experiment with yaw control since it is available through a mechanical linkage. So primary flight surface control without the aid of sensors and receiving primary input via a bell crank that is attached to the lines, yaw control through a gyro or other sensor and motor control without limits.

What I have trouble understanding is that you and all members of the AMA have for decades taken a person at their word that they were the builder of their own model. If they declare it so then it is. Why should this be any different?

As far as testing to determine if the model is using flight path control, simply tape the lead outs together, power the model up (without a prop) and move the nose in three dimensions. If the controls move it is using flight path control.

If the model passes technical inspection but is suspected of transgressing the rules I suggest that someone file a protest and a competent flier be appointed to fly the model and render an opinion. I have no problem letting any competent flier fly my model. If that is the price of admission then so be it.

I do not see the FAI rule being changed any time soon.


Kim.

Quote from: Kim Doherty on August 29, 2013, 08:00:09 PM
What I have trouble understanding is that you and all members of the AMA have for decades taken a person at their word that they were the builder of their own model. If they declare it so then it is. Why should this be any different

  Probably shouldn't be, but I can tell you there are people lining up to write that proposal. It was rejected before, maybe it will again, but I wouldn't count on it.

   You have a point about trusting the competitors. But, BOM is actually a rule, and you have to lie to break it.

    With this, there is no rule at all right now, so no rule to break, an autopilot is absolutely permitted as it stands today.  I presume that you would suggest "electronic coupling of the flight surfaces shall limited to causing the surfaces to move in proportion to the bellcrank deflection" or some such. That might be a viable rules proposal. THEN, your theory about trusting the people to not cheat would be one possible outcome.

    But you need to make the argument and convince people it is workable. Starting by stating that people are trying to conspire against you is not likely to bring people to you point of view.

    Brett
   

Ok so debate aside, Kims system is not cheating, Brett makes some good points also.
What constitutes a "auto pilot"? Is it the part of the system that "Knows" when the airplane is not level and autonomously corrects it? I don't think that any system could sense and correct a non level flight situation, sure level flight is achieved at around 4-6 feet, but what happens when the model is at 45 degrees up from center at the top of a simple loop? You certainly don't want the system to correct non level flight at the top of the loop.
Is the auto pilot the part that adds a little more or less out board flap in a maneuver? Is it the part that adds a little more or less throttle?
Is it the part of the system that has a little slop in the flap horn (I've tried this) to maybe soften the control at neutral in theory to help eliminate hunting.
How do you "prove" that you have no auto pilot, how do you "test" a system to prove it is not autonomous?
Brett, if there is a rule banning autonomous control systems, then yes it would be like the "bom" rule.
Thoughts?

Quote from: MikeCoulombe on August 30, 2013, 07:11:20 PM
How do you "prove" that you have no auto pilot, how do you "test" a system to prove it is not autonomous?
Brett, if there is a rule banning autonomous control systems, then yes it would be like the "bom" rule.
Thoughts?

  Kim's test (hold the leadouts and move the airplane around and see if it reacts) sounds OK. Of course you could have the timer enable the system at motor start.

    I don't think, the way the stars are aligning, that it will be like BOM, because I think the overwhelming likelihood is that anything but mechanical connections are going to get disallowed. Not because of any sort of vendetta, but because no one can think of anything better that is sure to work. I don't necessarily advocate that, but that's what I expect to happen.

    BTW, what we mean by "autopilot" could mean almost anything from weak stability enhancement, to full-on automatic flight. Typically we use it to make the control input a "rate command" system where the bellcrank angle represents an target angular rate, which is then achieved by a combination of feedback sensors. I am not entirely convinced that this is a bad thing, but I bet I am in the minority. It's a very small step from there, to generating the rate commands to perform a perfect maneuver automatically when triggered. Everybody agrees that this is not acceptable.

     Brett

I think everyone is making a mountain out of a mole hill. It is a huge jump "HUGE"from this to any type of pre programmed flight system. The only thing I would be concerned with is stabilizing Gyro's which is doable from here.

Berkeley had a model back in the 1940's or so that had mechanical flight stabilization. If you are going to have a rule to eliminate electronic stabilization then you also need to include mechanical stabilization as well. It had a pendulum to augment the user control inputs. It was the Berkeley "Bug" and the control system was called "Autotrol."

-Chris

I think the real issue here is the fact that some would move to suppress innovation and creativity in a hobby that should welcome  and embrace new technologies.
Stunt has always been about innovation.

If we move to stifle the exploration of new innovations and technologies, then we might as well sign stunts death certificate now!

Go ahead, propose your little witch hunt rules that suppress free and independent thought.

I can hardly wait to see you prove your control system is all "mechanical" when it is enclosed in the airframe!

Quote from: Robert  Storick on August 31, 2013, 01:48:04 AM
I think everyone is making a mountain out of a mole hill. It is a huge jump "HUGE"from this to any type of pre programmed flight system.

  No, it isn't. It's actually quite simple once you have a processor in the loop, then it's just a matter of software. Plenty of people already know how to do it.

    Brett

Quote from: MikeCoulombe on August 31, 2013, 11:50:36 AM

Go ahead, propose your little witch hunt rules that suppress free and independent thought.

   That's not helpful, nor likely to win any converts to your position.

    Brett

Quote from: MikeCoulombe on August 31, 2013, 11:50:36 AM

Go ahead, propose your little witch hunt rules that suppress free and independent thought.

You are right Brett I won't win any converts to my opinion with such volatile rants.
I apologize to the forum and retract the statement.
Sorry.

I just feel that the all out ban on anything but direct mechanical control will jeopardize the future of stunt.
We are presenting our passion for precision aerobatics to a new generation, a digital generation who I believe want to be able to endlessly "tweak" things to make them do what they want them to do.
I'm sure that many new flyers would embrace the technology.

Quote from: Brett Buck on August 31, 2013, 03:44:44 PM
  No, it isn't. It's actually quite simple once you have a processor in the loop, then it's just a matter of software. Plenty of people already know how to do it.

That, I think, is a stretch.  To do an autopilot that could beat one of the world-class guys would take a lot more than just the ability to string a few lines of code together.

Quote from: Tim Wescott on August 31, 2013, 05:44:37 PM
That, I think, is a stretch.  To do an autopilot that could beat one of the world-class guys would take a lot more than just the ability to string a few lines of code together.

  Well, you would certainly have to do some math, but it's simple enough.

    Brett

Well I think it would take more than just a processor in the loop. You need appropriate sensors to detect altitude, speed, and directions. However there are flight computers for quadcopters and such with the necessary hardware, and at reasonable cost.

-Chris

Quote from: Chris Sarnowski on August 31, 2013, 08:36:29 PM
Well I think it would take more than just a processor in the loop. You need appropriate sensors to detect altitude, speed, and directions. However there are flight computers for quadcopters and such with the necessary hardware, and at reasonable cost.

  Absolutely! To first approximation, you only need gyros. Make the deflection of the bellcrank represent the commanded rate, command the servo to a value proportional to the bellcrank deflection, and then add to that the difference between the rate command and the gyro rate. If the rate deviates from the desired rate, the gyro corrects it. That's a reasonable, simple autopilot design, that would make the airplane much easier to fly because you no longer have to pay much attention to whether or not your control input did what you expected.

    Now, replace the bellcrank input to this system with something that, when triggered, generates a fixed pitch rate command, and integrates it until you get to 90 "steradian degrees", then sets the commanded rate back to zero. Trigger with "hard-over" bellcrank position, below hard-over, it works like it did in the previous example. Now you have a perfectly programmed square corner, repeat 4 times, perfect square, correct for drift by changing trigger points.

   That could be done with a very minimal processor contained in the postulated "black box".

   If you want it better, then start adding accelerometers to create a "strapdown" inertial reference, then you won't have to correct for drift due to wind any more and can program complete maneuvers.

    All this sort of processing is quite simple and people have been designing systems like this for more than 50 years for satellite control and rocket guidance. The Atlas missile guidance system autopilot worked pretty much exactly like the simple system above, the guidance law output was a body angular rate command that was implemented with a feed-forward (the bellcrank angle generating the control deflection) and gyro rate feedback to correct errors in the open-loop feedforward. The guidance was external on that one, based on riding radar beams early on.  Which brings up the other half of this that no one has mentioned - our  soon-to-be-permitted 2.4 ghz radio links. Which could provide the external reference to make the thing entirely automatic.

   It would take some work but all of this is well within the state of the art.

    Brett

So the tech exists, banning it won't make it go away.

Personally I can hardly wait for a fully programmable over the counter control system.