As lovely as the scenery is in Kloner's video, there
is evidence of "unwanted movement" - for example at 3.40 onwards, look carefully at the top left corner - there's a little vertical "dip" just as the horizon touches the top of the frame. Also some nasty shaking in the last few seconds but I don't know if that's the stabilization or the gimbal mount itself moving.
Nobody is saying that RC servos are obsolete or that belt drive on BLDC is invalid - just that they can't/don't provide the end results that a direct drive system can. Close, sure, but not the same.
As to IrisAerial's question, wind is certainly a factor. Any imbalance will be exploited by the wind since the "centre of pressure" is unlikely to be coincident with the centre of gravity (which, through balancing,
should be the same as the centre of rotation). Wind pressure will therefore act just like imbalance of the load, requiring the motors to "do work" to hold position. How well they can do that depends on the reserve of torque available and
that depends not only on the motor specs but also the particular tuning setup (especially the "gain").
As with any actuator system, the ideal is to have a substantial reserve of "power" available. Playing "near the limit" yields problems. This is directly analogous to the motors/props on the craft - we want to hover at around 50% throttle so that we've got plenty of "control bandwidth" available (above and below that figure). An underpowered craft that requires, say, 80% throttle to hover runs out of puff if a sudden input of 25% throttle increase is demanded. In the same way a direct drive gimbal, which
is operating "near the limit", can have insufficient reserve available to respond to a disturbance.
This is why IMO the AV200IDD approach
should prove more successful since, through gearing, it will have more torque in reserve (together with some other factors) and will be more forgiving on the tuning. However, their gearing system has a "compliance" (a rubber interface between the driving wheel and the driven wheel) which
will rob some of that advantage.
So, no "revolution" yet - just a step in the right direction - but we've still got some way to go. Right now balance and tuning are critical - IMO the best choices in the current crop are those gimbals which provide "micro-adjustment" of the balance (like the Halo gimbals). See also the purely mechanical hand-held "gimbal" from
FlowCine - check out the micro-adjustment system available there. Unfortunately engineering on this level costs money!
