[al]

Hi all,

maybe this is complete nonsense, but I have this idea since while but I am not deep enough in LP to acutally try it:

To reduce workload of the tail servo on a heli, could one not mix a bit of the yaw gyro feedback into the main motor, just a little "D" or such.
Result would be that the torque of the main rotor is modified a bit (within a small range, of course) and could handle small tail corrections.

Something similar is done in CD-players' track control: The laser is fine-adjusted by a magnetically moved lens, the coarse adjustment is done via the track motor. In the helicopter case, the main motor torque would do the fine, the tail servo would do the coarse work.

The ESC would have to be a rapid one (it is in my heli, anyway), and it would not be ran in gov mode.

The goal would be to reduce the average overall activity of the tail servo, of course.

Any thoughts?

[daveapplemotors]

Interesting question there. I use a governors on my 250s to make sure that the headspeed stays constant.

Correct me if I am wrong here:
Variables for the tail servo are inner loop (TX inputs) and outer loop (wind and heli torque).
And you are right. There is nothing in CC3D that helps with tail mix or pre-compensation.

Lowering the headspeed when the load is heavy (to the tail servo):
How could it be done quickly? Would it help the tail make a large change? And when that change is done the headspeed would try to come back to optimal (how?) and necessitate more corrections--by the pilot and/or CC3D for both pitch* and yaw 

Raising the headspeed when the load is heavy:
Again how will it be fast? And the tail servo might tend to overcompensate because the tail will spin faster, too. Will it go into a loop that needs it's own PID? And when the headspeed returns (how again?) back to the desired speed it will need some correction.

Lowering headspeed with light loads might leave the motor unprepared for a sudden pitch* increase. Raising it with light loads might work but be hard to fly. 
 
I use 'tail mixing' in my TX, a DX6, and have 42% selected at full pitch (pos or neg) to compensate. Linear tail curve for me with 21% correction at 50% pitch. Since those mix numbers are surprisingly high I am wondering how much throttle change would be needed to help the tail servo.
 
Tail mixing by the TX gives a 'heads up' on the coming load to the tail for pitch changes. I have  counterweighted (Chinese weights) tail grips to lessen the load considerably on the tail servo.     

*pitch here means total pitch. Not the pitch in 'pitch and roll' as CC3D defines it.

[al]

Well, My thinking was that a very minor speed change on the main rotor (say, +-2%) will not change thrust a lot.

But it may be enough to introduce a bit of momentary torque to the vehicle to correct little tail movements.
As soon as the +-2% correction range is used up, the tail servo would have to care for the rest.
My thinking was that all the very little tail servo corrections (that take place on one place of its potentiometer) could be reduced a bit (maybe 50%)....
Acutally, my tail servo gets hand warm in very moderate hovering/cruising with no visual oscillations.
As said above, the idea could be complete nonsense...

[daveapplemotors]

I see what you are saying.

And since I must correct 42% tail mix do you think that you could get a meaningful change of headspeed without upsetting pitch and responsiveness? And how fast could headspeed/thrust react?

I believe the larger force wanting to rotate my helis counter clockwise is from blade pitch not headspeed.
Try increasing throttle suddenly without any pitch change. Just program a flat zero pitch curve to make that test. I will try that test when I return later today.

As for pitch's influence on torque: with 0% tail compensation I get about 90 degrees of yaw left in a pitch pump. Old time heli pilots learned to add right yaw with pitch manually. Move the stick at an angle instead of straight up. That works too.

Pre-comp similar to Beastx would start tail servo corrections with the pitch stick. It would probably be much easier to implement. CC3D could use that in my opinion but since quads don't need this the programmers don't care.

Tail servo heat: CC3D says to match the output frequency to your servos per manufacturer's specs.
Channel 4 on output page if I remember correctly. I burned up a CC25a ESC and a tail servo fooling with the output frequencies so I just set everything to 50hz. My human reaction times are much slower so I don't notice the difference between 330 and 50hz.

I have a couple of Eflite DS76t tail servos that heat up a lot no matter what frequency I use. And that is before I spool up! I can't use them. HOT. 'Hand warm' would be my best friend because I really like those servos because they are 7 grams, fast and powerful enough for my 250 helis.