A conventional flybar acts as power steering. And ”power” steering indeed - on a .90 heli, it may take away a hundred watts or more from the main engine.
In a flybarless setup, the swash servos need to do the same job. They are more efficient than the flybar, but still they need to work very hard.
This article discusses some typical problems that can limit the performance of a FBL heli.

swash servo setup


The swash servos remain one of the main challenges in a flybarless heli. They are the link between the flybarless controller and the blades. The software running inside the little black box may be clever, but it's helpless when the heli doesn't obey its commands. If a loaded servo moves quickly in one direction but slowly to the other, the flybarless controller will do sloppy work and the heli flies badly.
What we need:
1) adequate servos with sufficient speed under load
2) a link geometry that doesn't overload them.

A consequence from 1) is that rudder servos don't work well as swash servos in FBL. On paper, they are the fastest, but not anymore under load.
For example, the Align DS 620 is lacking torque for .90 size. Use DS 610 instead. Futaba 9257 is unsuited for a 500-heli on swash for the same reason.
The solution to 2) is to choose a geometry that allows maximum servo movement for the required pitch range. Usually the setup procedure of the FBL controller includes a step to check, whether or not the pitch range is good (”blue light” test on beastX, ”Cyclic gain” calculator on V-bar). If the test fails, fix the geometry before going any further. It's worth the effort.

Power supply


At this point, we've made it as easy as possible for our swash servos to do their work. Still, that's not enough: A hard-working swash servo also needs sufficient power.
Much has been written about servos and geometry, but maybe the power supply never got the attention it deserves.
In my .90-size heli, a 6-minute flight draws up to 650 mAh. This is equivalent to an average current of 6.5 amps. The peak current is certainly several times higher. When looking at the flimsy servo wires and connectors, this is quite a lot of current.
If a regulator is used, it needs to be dimensioned according to the peak current, otherwise it will become the ”bottleneck” when it matters most.
Let's take an example:
  • the heli is performing a fast flip, such as in a tic-toc
  • The pilot releases the cyclic stick and commands the heli to stop the flip
  • Note that this is a difficult task for the heli: At one moment it's rotating along the elevator axis. An instant later, it's supposed to stand still. How soon? As soon as you can...
  • To stop the heli, the FBL controller commands a brief burst of cyclic pitch.
  • There is a lot of servo movement in a very short time: Increase cyclic pitch, and turn it back to neutral.
  • Turning the heavy blades within a fraction of a second requires quite some force.
  • The current draw of the swash servos is high.
  • So far so good. Now if the power supply isn't up to the job, the servos can't keep up with the commands.
  • The commands from the FBL controller are executed sloppily, for example causing interaction with the aileron axis.
  • The FBL controller is forced to apply corrections
  • Result: The heli feels inaccurate, wobbles and flies badly.


What can we do about it?
In addition to adequate servos and geometry - see 1) and 2) - we should use multiple wires to the FBL controller, so that the full current doesn't need to go through a single servo connector.
On the FBL controller you'll find unused channels with "+" and "-" pins that can be used for additional power supply. Note, a few connectors include only signal pins, no power supply. Check the manual before connecting anything.
Then, get a Rx battery that is powerful enough. For example, the Outrage Rx LiPos are in my experience underpowered. Instead, I'd use a high-C flight LiPo. For now, I've got 3300 mAh 2s, but 2200 - the same cell size as in a 450 - should probably do.
An on-board monitor for the supply voltage (for example V-bar or Helitron regulator) can be useful to spot problems with the power supply.

Blades


We have another option to ease the work of our swash servos: Use FBL blades.
A conventional FB blade has a leading edge that cuts into the airstream. The blade pitch tends to reinforce itself, adding ”snap” or ”punch” on a FB heli. For the servos on a FBL head, FB blades are an unruly lot that requires a strong arm to control. In other words, if servos, geometry and power supply are good enough, conventional FB blades may work very well. But if the heli doesn't perform according to expectations, it may be a good idea to put on FBL blades simply to rule out the blades as a possible cause of ”wobble” and inaccuracy.
The advantage of lighter blades is that they cause less loading to the engine for a given roll rate, so the heli can be faster on cyclic without bogging.

Rudder


While there are some FBL-specific features (torque precompensation, piro compensation), the basic rudder setup is the same as on a conventional heli. The old rule applies: Most rudder problems cannot be solved by tweaking parameters, but need changes to the mechanical setup.

There is one difference: In a FBL heli, rudder problems can become indistinguishable from swash-related issues in tictocs or the like. Therefore, it may be wise to do a careful rudder setup before starting to investigate swash-related problems.
In electric helis, a fast governor can cause torque peaks that lead to tail inaccuracies during tic-tocs. This can be improved by using ungoverned mode and running with a flat throttle curve. Reducing the governor gain should also help, allowing the headspeed to bog slightly and thus reducing torque during the abrupt move. For a small electric heli, ”flat 100 throttle” and a suitably sized pinion can be a very robust setup.

Summary


  • Use servos with sufficient speed under load
  • Set the geometry correctly
  • Use multiple power supply wires to the FBL controller
  • Use a Rx battery that can deliver enough current
  • If in doubt, use FBL blades