Thread Tools
Mar 04, 2019, 09:22 AM
e^(iπ) + 1 = 0
Miami Mike's Avatar
Quote:
Originally Posted by locoworks
never used all the options so dumb questions time. what does the symbol in L2 mean, and what is the purpose of the [ ] in L5 and L6?
The ~ symbol is called a tilde. You can type a tilde by holding shift and pressing the key that's below esc and above tab. It's supposed to mean "about equal to" but after some experimenting with OpenTX it's not clear to me precisely what a~x does. It appears to behave differently with different V1 parameters, behaving the same as = in some cases and exhibiting an asymmetrical range in some other cases. It might be worthwhile to investigate exactly how a~x works.

The | symbol used to indicate absolute value when used before and after a variable is called a pipe. You can type a pipe by holding shift and pressing the key that's below backspace and above enter.
Sign up now
to remove ads between posts
Mar 04, 2019, 05:01 PM
Registered User
Atomic Skull's Avatar
Quote:
Originally Posted by l shems
Hi all,

Since the latest opentx update, Lua is allowed to write a curve to the model, it has become possible to write a helicopter wizard. Only problem is the Heli mixer page. No Lua access exists to change that page, so no general approach can be taken to get ALL Heli types supported.

Looking into it in some more detail however, these mixes are not so complex for ccpm 90, 120, and 140 swash types. Actually, they can be done with just a few mix lines.
Please don't forget 135 degree swashplates.

Also 140 degree swashplates and 135 degree swashplates are not exactly 140 or 135 degrees they just round off the name for convenience.

The difference between 140 degrees and 135 degrees is that 140 degree mixing eliminates interactions on the elevator (with a 120 degree swashplate the elevator servo has to move further than the aileron servos and lags behind them because of this) while 135 degrees eliminates elevator interactions AND elevator/aileron cross coupling as well.

A cyclic ring function to prevent mixed elevator/aileron inputs from exceeding the maximum value of pure elevator and aileron would be useful as well.
Last edited by Atomic Skull; Mar 04, 2019 at 05:11 PM.
Mar 04, 2019, 06:24 PM
Have Fun and Just Fly!
l shems's Avatar
Quote:
Originally Posted by Atomic Skull
Please don't forget 135 degree swashplates.



Also 140 degree swashplates and 135 degree swashplates are not exactly 140 or 135 degrees they just round off the name for convenience.



The difference between 140 degrees and 135 degrees is that 140 degree mixing eliminates interactions on the elevator (with a 120 degree swashplate the elevator servo has to move further than the aileron servos and lags behind them because of this) while 135 degrees eliminates elevator interactions AND elevator/aileron cross coupling as well.



A cyclic ring function to prevent mixed elevator/aileron inputs from exceeding the maximum value of pure elevator and aileron would be useful as well.
Well, those angles can be covered easily. It is the other thing I don't get. This ringy thingy.

justfly solutions
Mar 04, 2019, 10:48 PM
Registered User
boomerwing's Avatar

Suggesting My Tutorial


Quote:
Originally Posted by locoworks
never used all the options so dumb questions time. what does the symbol in L2 mean, and what is the purpose of the [ ] in L5 and L6? also what is the difference between x and b in formula terms when the equation is the same except the letters?
You might find my Logical Switch Tutorial helpful. Have a look.
https://www.rcgroups.com/forums/show...witch-Tutorial
boomerwing
Mar 05, 2019, 12:55 AM
Registered User
Atomic Skull's Avatar
Quote:
Originally Posted by l shems
Well, those angles can be covered easily. It is the other thing I don't get. This ringy thingy.

justfly solutions
Back in the day, before computer radios you'd stick a "cyclic ring" in the gimbal which restricted the stick to a circular aperture. The reason for this is because aileron and elevator cyclic are additive and if you allow the full range of the pots on the angles of the cyclic stick it is possible to have more cyclic pitch on the angles than on the up/down and left/right axis. These days this is done electronically. Most flybarless controllers have this feature but a few don't and you also need it if you are setting up a flybarred helicopter. Essentially it draws an imaginary circle around the gimbal which touches the maximum limits of the pure up/down and left/right axis of the sticks and ignores any input that falls outside of it and ensures that the cyclic has the same maximum amount of cyclic pitch all the way around.


The problem with not having this is that you can get much more cyclic pitch on the corners which can cause mechanical binding in the head (this is especially true with flybarred helicopters with all of the links and levers in the rotor head) if you put the stick too far into the corner as well as causing it to be capable of higher roll rates on the angles as well. If you look at the analog stick on most game controllers you will notice that they have a circular aperture for the stick for the same reason.
Mar 05, 2019, 03:12 AM
Registered User
Quote:
Originally Posted by boomerwing
You might find my Logical Switch Tutorial helpful. Have a look.
https://www.rcgroups.com/forums/show...witch-Tutorial
boomerwing
thank you, yes it was very helpful.
Mar 05, 2019, 03:58 AM
Have Fun and Just Fly!
l shems's Avatar
Quote:
Originally Posted by Atomic Skull
Back in the day, before computer radios you'd stick a "cyclic ring" in the gimbal which restricted the stick to a circular aperture. The reason for this is because aileron and elevator cyclic are additive and if you allow the full range of the pots on the angles of the cyclic stick it is possible to have more cyclic pitch on the angles than on the up/down and left/right axis. These days this is done electronically. Most flybarless controllers have this feature but a few don't and you also need it if you are setting up a flybarred helicopter. Essentially it draws an imaginary circle around the gimbal which touches the maximum limits of the pure up/down and left/right axis of the sticks and ignores any input that falls outside of it and ensures that the cyclic has the same maximum amount of cyclic pitch all the way around.


The problem with not having this is that you can get much more cyclic pitch on the corners which can cause mechanical binding in the head (this is especially true with flybarred helicopters with all of the links and levers in the rotor head) if you put the stick too far into the corner as well as causing it to be capable of higher roll rates on the angles as well. If you look at the analog stick on most game controllers you will notice that they have a circular aperture for the stick for the same reason.
Ok, that makes sense, and clarifies the difference I found between the opentx Heli page outputs and the arithmetic mixes I created.

Thanks again.

justfly solutions
Mar 05, 2019, 10:27 AM
e^(iπ) + 1 = 0
Miami Mike's Avatar
That sounds like an interesting problem. There's an ancient geometry problem known as Squaring the Circle, but this could aptly be called "Circling the Square" .

Atomic Skull's description doesn't sound ideal if it ignores any input that falls outside of the circle. I wonder if there's a possible solution that doesn't involve dead areas of stick travel at the corners and instead returns a result more like the way aileron and elevator control works.
Mar 05, 2019, 12:03 PM
Have Fun and Just Fly!
l shems's Avatar
Quote:
Originally Posted by Miami Mike
That sounds like an interesting problem. There's an ancient geometry problem known as Squaring the Circle, but this could aptly be called "Circling the Square" .



Atomic Skull's description doesn't sound ideal if it ignores any input that falls outside of the circle. I wonder if there's a possible solution that doesn't involve dead areas of stick travel at the corners and instead returns a result more like the way aileron and elevator control works.
I think it's a matter of preference.

You can either start limiting the collective in the corners, so you still increase roll at the cost of collective. I think that is done in the Heli page.

Or you just make it dead area and maintain the collective. Actually, you can't overdrive I think if the servos are properly limited on the outputs, and just pushing them into overdrive will also have the same effect as decreasing collective pitch (in respect to the pitch given on the sticks.)

I think most important is that by overdriving in the corners you cannot DECREASE roll by the limiting effects of the servo endpoints.

Nice problem to analyse.

Any comments/ideas on this?

justfly solutions
Mar 05, 2019, 12:03 PM
Registered User
it would be like trying to set elevons so at full pitch and roll the surfaces just came to their travel limits. this would mean cutting down throws to allow this, maybe this would not get enough pitch or roll alone though, dead areas isn't an issue with elevons, one surface just stops moving but you get a differential in positions.
Mar 05, 2019, 12:11 PM
Have Fun and Just Fly!
l shems's Avatar
Problem with heli's would be that by increasing collective pitch during a roll, you could decrease roll rate. I think that would not be desired.

justfly solutions
Mar 05, 2019, 12:23 PM
e^(iπ) + 1 = 0
Miami Mike's Avatar
Quote:
Originally Posted by locoworks
it would be like trying to set elevons so at full pitch and roll the surfaces just came to their travel limits. this would mean cutting down throws to allow this, maybe this would not get enough pitch or roll alone though, dead areas isn't an issue with elevons, one surface just stops moving but you get a differential in positions.
It's funny that you should mention that! I've noticed a striking similarity between this helicopter cyclic matter and the elevon setup I use for my foamie slope gliders, which allows you to get the amount of elevator travel you need while automatically allocating the rest of the available elevon travel to roll control.

... and all without any clipping.
Mar 05, 2019, 01:14 PM
e^(iπ) + 1 = 0
Miami Mike's Avatar
Quote:
Originally Posted by l shems
Nice problem to analyse.

Any comments/ideas on this?
To map the points within a square to the points within a circle, how about this?

Translate the X and Y values to polar coordinates. The angle would be the arctangent of Y / X (with code to avoid an error in the case of X = 0). The R value (the distance of the stick from center) won't be needed. The new X value would then be the cosine of the angle times the stick's X value and the new Y value would be the sine of the angle times the stick's Y value.
Mar 05, 2019, 03:15 PM
Have Fun and Just Fly!
l shems's Avatar
Quote:
Originally Posted by Miami Mike
To map the points within a square to the points within a circle, how about this?



Translate the X and Y values to polar coordinates. The angle would be the arctangent of Y / X (with code to avoid an error in the case of X = 0). The R value (the distance of the stick from center) won't be needed. The new X value would then be the cosine of the angle times the stick's X value and the new Y value would be the sine of the angle times the stick's Y value.
And now on opentx mixer lines without Lua

justfly solutions
Mar 06, 2019, 02:39 PM
e^(iπ) + 1 = 0
Miami Mike's Avatar
Quote:
Originally Posted by l shems
And now on opentx mixer lines without Lua
It's way too complex for that.

Here's a telemetry script for the X9D+ that demonstrates mapping the square area of stick travel to a circular area without clipping. The script is just a demo and not intended to have any practical use.



In practice you'd probably write a mix script instead.


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion FrSky Horus X12S - OpenTX 2.2 Discussion Raydee Radios 113 Sep 30, 2019 12:16 PM
Discussion Pixhawk 2 (un)official discussion thread Rusty105 Multirotor Drone Electronics 211 Aug 15, 2019 05:54 PM
Discussion EXTREME RC AUSTRALIA: Official product announcements & discussion thread Extreme_RC Australia 2765 Jul 13, 2019 05:27 AM
Discussion Official Drone Registration Discussion Thread **Discussion Here** bansheerider Model Aircraft & Drone Advocacy 4320 Dec 07, 2017 01:53 AM
Yippee! Official Portal 2 Discussion Thread Miami Mike Chit Chat 9 May 02, 2011 08:19 AM