Although I have known about this forever it just now occurred to me that my quad may not be handling as well as it could be because of this. What I'm talking about is how it's more sensitive when you push directly towards an arm. For example in X mode a sideways flip is slower than a diagonal flip along the arm's axes (or the reverse if in plus mode). This also translates into the general handling where you have to use slightly less throw on the sticks when you're pushing directly towards an arm. You get used to it but it feels kind of funky. I noticed it after I had spent a whole bunch of time flying my heli then came back to the quad.
This got me thinking that maybe a swash ring would help here. I use er9x which has an electronic swash ring that I use with my heli's to prevent binding of the swash and I'm wondering how it would work with MultiWii. At first it seems straightforward but then I started thinking about the PID rate settings that some of us use. This has an exponential type effect on the PID in that it reduces the P the more stick throw you use. I'm not sure how a swash ring will interact with that because it will be reducing the stick throw in some directions and MultiWii will in turn be reducing the throw by an exponential amount if you're using a rate setting.
I think the ultimate goal would be to make the copter react at the same rate no matter which way the stick is pushed. So in my example above with an X mode quad I want it to flip at exactly the same rate whether I'm flipping towards the side or along a diagonal. I think this might make it handle better in general. I'm not sure if just the swash ring will work if you're using PID rates though, that might interact weirdly. It might require a non-circular swash ring to get it just right. Or maybe changes in MultiWii.
Anyone experimented with this?
Sensitivity on the arm's axes; use a swash ring?
Re: Sensitivity on the arm's axes; use a swash ring?
Did a quick test with 100% swash ring. Yeah, the exponential effect of PID rates adversely affects the roll rate when the stick is pushed to the corners. eg. it slows down a diagonal flip too much. However, first impression is it did make the quad feel and handle a lot better in general. Feels more natural and not so squirrelly.
Seems like maybe MultiWii itself could have some sort of swash ring type behaviour. It needs to have a non-circular shape (algorithm) based on any PID rate settings.
Seems like maybe MultiWii itself could have some sort of swash ring type behaviour. It needs to have a non-circular shape (algorithm) based on any PID rate settings.
Re: Sensitivity on the arm's axes; use a swash ring?
Yet, true. It has to do i think with asymmetry. On a octo, duodec
you will less feel it. On a quad its obvious, at full stick, sometimes only 2 motors work (+ & -) and the other extreme 4 motors (++ & --). So aerodynamics play a role too, slowing down 2 motors instead of 1 ...
manu

manu
Re: Sensitivity on the arm's axes; use a swash ring?
Hey alll , this is LittleMo.
Yes, it would be less noticeable the closer together the motors are. I'm surprised no one has tried to fix this before because it makes my quad handle so much better IMHO.
I'm currently looking at making the swash ring in er9x configurable as a Super Circle. Or maybe it should be in MultiWii, I'm not sure. I have some graphs showing what I'm talking about here:
http://9xforums.com/forum/viewtopic.php ... 59&p=38854

I'm currently looking at making the swash ring in er9x configurable as a Super Circle. Or maybe it should be in MultiWii, I'm not sure. I have some graphs showing what I'm talking about here:
http://9xforums.com/forum/viewtopic.php ... 59&p=38854