Distinguished travel rates and limits

[GoogleCodeExporter]

The facts :
Travel rates and limits of servos have the same value, so uncentering stick 
trim affects the linearity of the servo displacement when order reach 100%.

The problem :
With pitcheron, dual servo tail or V tail, using 100% elevator with aileron or 
rudder uncentered trim produces an unwanted roll.

My proposal :
Allow, by option, to have a travel limit equal to 125% (for example) travel 
rate.

Original issue reported on code.google.com by f.ague...@wanadoo.fr on 30 Jul 2012 at 10:07

[GoogleCodeExporter]

I'd do the opposite.
Of course it's "psychologically" more logical to use 100% in mixes, but as 100% 
already result in much bigger throws than on another radio... I'd just use a 
lower weight on the controls. 
For example, on a delta model, if you set all mixes at 100% then putting the 
stick in the corner will result in clipping (each servo should go to 200%, 100% 
due to elevator and 100% due to aileron). The solution is to set mixers at 50%, 
that way you never exceed 100% throw on any servo and thus don't have any 
clipping (BTW, might be a good idea to fix that in the templates). 

So in your example, use let's say 80% on your elevator/aileron mixes. Limits is 
100%, so you have some margin for trimming. Pretty much the same as using 100% 
for the mixes and 125% for the limits actually :)

Original comment by bernet.a...@gmail.com on 30 Jul 2012 at 10:14

[GoogleCodeExporter]

You are right for the rates (in the general case), but I don't agree the 
logical : in my opinion, the throws are adjusted to have the expected behaviour 
of the plane and not to have some margin for trimming.

Bu my examples where not the best... I should have told about butterfly 
function, witch use a lot of throw and so will be affect by aileron trim.

Original comment by f.ague...@wanadoo.fr on 31 Jul 2012 at 6:35

[GoogleCodeExporter]

No, on a complex model the LIMITS should be adjusted for the maximum mechanical 
throw you'd want each control surface to reach, all mixes combined, to avoid 
forcing on the servos/linkages, and to guarantee symmetry (for example your 
rudder can mechanically move +45°, -50°, you want the max so you set limits 
so that it reaches +/- 45°).

Then, the actual control throws dedicated to each function get set in mixer 
weights.

On a simple 4CH model without any advanced mixes it's perfectly fine leaving 
all mixers at 100% and setting throws in the limits screen, as there's never 
more than one mix affecting each control anyway. But as soon as you have 2 or 
more, you need to decide on their individual ratios.

Original comment by bernet.a...@gmail.com on 31 Jul 2012 at 7:18

[GoogleCodeExporter]

Yes, setting 100% rate is the good away (it could be consider as an activation 
more then a rate), but even for a simple model you shoud use stick dual-rate to 
adjust function travel (on opposition to servo travel) instead of travel gains 
/ limits.

For a complexe model :
- travel gains symetrise the throws (I'm very happy to read you about this 
essential point !)
- travel limits protect the linkage

So, why limits and gains, each with its own and different utility, should have 
the same value ? That's not logical ! 

Original comment by f.ague...@wanadoo.fr on 31 Jul 2012 at 7:36

[GoogleCodeExporter]

Because we have a radio with a very limited processor in terms of available 
program and storage memory, and it's simply not possible to do everything, and 
some simplifications are needed!

Storing separate gains and limits would require a lot of resources, and only 
cover one rare case (for example you have an aileron that has 30° up and 40° 
down mechanical throw, now you "need" to fix limits to +/-30° to maintain 
symmetry, while with separate gains you could use the remaining 10° for 
example to mix flaps in).
And there's already a way to do it, now you could set limits for +30°/-40°, 
then use a 3-pt curve on the aileron mixes to achieve symmetry on that function.

Original comment by bernet.a...@gmail.com on 31 Jul 2012 at 8:27

[GoogleCodeExporter]

Yes, I know that the management of the memory is very hard. That's why I 
suggest a very simple rule (travail limits = travel gain + xx%, for every 
servo), just to have a little margin (10% could be sufficient in most of the 
case) against the trim effect.

Original comment by f.ague...@wanadoo.fr on 31 Jul 2012 at 8:37

[GoogleCodeExporter]

I think we can close this one? 

Original comment by bernet.a...@gmail.com on 16 Aug 2012 at 3:46

[GoogleCodeExporter]

OK : this issue was solved in the same time as #66.

Original comment by f.ague...@wanadoo.fr on 16 Aug 2012 at 3:52

[GoogleCodeExporter]

Original comment by bernet.a...@gmail.com on 16 Aug 2012 at 4:05

• Changed state: WontFix