Curve bug ?

Post Reply
willhac
Posts: 89
Joined: Tue Feb 19, 2013 11:25 am
Country: France

Curve bug ?

Post by willhac » Sat Aug 08, 2015 11:21 am

I noticed something odd while setting up this model : the end position of the servo is moving where I think it shouldn't.
See channel 2 or 4 (ailerons) on my model, when I apply a curve -100 -50 0 50 100 on the 2nd mix, the servo ends are the same with !AIL or AIL. (normal)
But if the curve is -100 -75 -50 25 100, the servo end position moves a little bit, like if changing point 2 had an influence on point 1 !
My version is r818 for 2561.

Guillaume
Attachments
F-86.eepm
(1.43 KiB) Downloaded 69 times

User avatar
jhsa
Posts: 18022
Joined: Tue Dec 27, 2011 5:13 pm
Country: Germany

Re: Curve bug ?

Post by jhsa » Sat Aug 08, 2015 12:07 pm

Humm, I also don't understand that.
I think the end points should not change as the first value of the curve is -100% and the last is +100% anyway.
A curve of "-100 -75 -50 25 100" should have the exact same position at the end of the travel as a curve of " -100 -50 0 50 100", as the end values are exactly the same.

I also played with the dual rate and changed the values to 100% and the problem seem to go away and the end points then becomes the same when you flick the switch..

As I said, I don't really understand what is happening and also think there is some odd behavior. Mike is the right person to comment on this though.. ;)

João
My er9x/Ersky9x/eepskye Video Tutorials
https://www.youtube.com/playlist?list=PL5uJhoD7sAKidZmkhMpYpp_qcuIqJXhb9

Donate to Er9x/Ersky9x:
https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=YHX43JR3J7XGW

User avatar
MikeB
9x Developer
Posts: 15719
Joined: Tue Dec 27, 2011 1:24 pm
Country: -
Location: Poole, Dorset, UK

Re: Curve bug ?

Post by MikeB » Sat Aug 08, 2015 12:24 pm

I've not had time to look at this, but if you have dual rates set to less than 100%, then you don't reach the end of the curve I think.
The dual rates are applied to the stick values BEFORE the curve in the mix is applied. So if you are at on 80% of the stick input then you are not at the 100% point of the curve, so changing the second point will indeed change the output position.

Mike.
ersky9x/er9x developer
The difficult we do immediately,
The impossible takes a little longer!

User avatar
Kilrah
Posts: 9611
Joined: Sat Feb 18, 2012 6:56 pm
Country: Switzerland

Re: Curve bug ?

Post by Kilrah » Sat Aug 08, 2015 12:36 pm

Yep as the stick range is "trimmed" before getting to the mixer you're only using a portion of that curve if D/R is used. You'd have to ignore the D/R settings and recreate its behavior in the mixer if you wanted the entire curve range to be used.

willhac
Posts: 89
Joined: Tue Feb 19, 2013 11:25 am
Country: France

Re: Curve bug ?

Post by willhac » Sat Aug 08, 2015 12:57 pm

Thanks for your answers, I agree with that.
In both cases !AIL and AIL the D/R is set to 80%.
I don't think changing the second point of the curve should move the output for the first point of the curve if I don't modify it...
Try to move from -50 to -75 the second point of the curve and you will see the output for -100 is changing.


User avatar
jhsa
Posts: 18022
Joined: Tue Dec 27, 2011 5:13 pm
Country: Germany

Re: Curve bug ?

Post by jhsa » Sat Aug 08, 2015 1:00 pm

MikeB wrote: The dual rates are applied to the stick values BEFORE the curve in the mix is applied. So if you are at on 80% of the stick input then you are not at the 100% point of the curve, so changing the second point will indeed change the output position.

Mike.
Correct, but that is not the problem I guess..

If you change the curve 2 to -100, -50, 0, 50, 100, and move the Ail stick to the ends, the end points don't change when you flick the AIL switch.. The dual rates are still applied to both mixes, so why does it behave differently with a different curve that has exactly the same end points values? Why are the servo end points different? I think they shouldn't. It seems to be wrong somehow..
I'm testing it in eepe..

Both curves have exactly the same values at the beginning and end. The values should be exactly the same with switch ON or OFF?
João
My er9x/Ersky9x/eepskye Video Tutorials
https://www.youtube.com/playlist?list=PL5uJhoD7sAKidZmkhMpYpp_qcuIqJXhb9

Donate to Er9x/Ersky9x:
https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=YHX43JR3J7XGW

User avatar
Kilrah
Posts: 9611
Joined: Sat Feb 18, 2012 6:56 pm
Country: Switzerland

Re: Curve bug ?

Post by Kilrah » Sat Aug 08, 2015 1:16 pm

willhac wrote: I don't think changing the second point of the curve should move the output for the first point of the curve if I don't modify it...
Try to move from -50 to -75 the second point of the curve and you will see the output for -100 is changing.
Again it's normal. When your stick is at -100 with a DR at 80% the mixer gets a -80, so you don't start on point 1 of the curve, you start in the middle between point 1 and point 2. As a result, if you move point 2 (which changes the slope between point 1 and 2) the output will vary.
capture_002_08082015_151200.png
capture_001_08082015_151120.png
If you don't want that leave the DR at 100%, and set mixer weight at 80% instead. Then you'll be using the whole curve.

User avatar
jhsa
Posts: 18022
Joined: Tue Dec 27, 2011 5:13 pm
Country: Germany

Re: Curve bug ?

Post by jhsa » Sat Aug 08, 2015 1:30 pm

so why is not the curve -100, -50, 0, 50, 100 affected then? The dual rate setting is the same for both. The end points are the same on both curves. Why isn't the servo end point different when flicking the switch and using this curve then?

João

EDIT: hmmm, changing all the limits and sub trims to default I think I understand why.
Because with this curve 80% of the stick value corresponds to 80% on the curve itself as it is a straight line, so no change on the behavior..

Try it in eepe. program another model, but without expo, and keep all sub trims and limits default. you can then understand it better.. ;)
My er9x/Ersky9x/eepskye Video Tutorials
https://www.youtube.com/playlist?list=PL5uJhoD7sAKidZmkhMpYpp_qcuIqJXhb9

Donate to Er9x/Ersky9x:
https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=YHX43JR3J7XGW

willhac
Posts: 89
Joined: Tue Feb 19, 2013 11:25 am
Country: France

Re: Curve bug ?

Post by willhac » Sat Aug 08, 2015 1:46 pm

OK I understand now... I'll set the D/R back to 100 and use the mixer weight.
It was the fastest way to reduce the ailerons after a first flight !
Thank you for this explanation !

Post Reply

Return to “er9x model setup”