So do you need to upgrade servos if you go FBL?

[Stevec]

Hi Gang,

I am talking theoretically here, There is always a huge emphasis on speed for a tail servo, far faster than a human can actually react and this is due to the fact that we run a gyro which can react as fast as electricity flows.

with the mk1 human eyeball it is hard to detect movement that needs correction until it is well off course.

now enter the FBL controller. surely we should be sticking the fastest possible servos on the cyclic to keep up with the gyros in the controller? I have been casually looking at flybarless, the only helis I have that are real candidates are my trex 500 and clone 450's. I think I may bite the bullet on the 500 size but I was wondering why I don't see the servo speed being of that great an importance on the FBL?

Going back to the tail servo speed debate it seems to be some people train of thought that if you can put something on there that react in the time it takes my wife to lose her rag then it isn't even worth bothering with a gyro...

is it just snobbery or do we all need to upgrade the servos across the whole bird when we go FBL?

of course a decent cyclic servo still moves 60 degrees in less than 0.17 of a second which is faster than a human (IIRC human reflexes are about 0.4 of a second).

so is it simply the fact that you sacrifice torque for speed and you mist have a certain torque level to move the swash with a running rotor?

I noticed MKS are selling FBL servos now is this going to be a new trend?


sorry if this question has been answered before but I couldn't find it.

/Steve

[Si_B]

I think its more to do with how quick the electronics can react to the servo, not you !!

[Mark_T]

A lot depends on the design of the head, but generally the ability to hold posiiton accurately and move precisely, when commanded, to a new position under high loads is more important than just speed.

This is especially true when you have direct links from the servo to the swash.

Yes, I think you'll see more servos specifically designed for FBL, they are already there now from MKS, JR and probably others.

[Stevec]

I think its more to do with how quick the electronics can react to the servo, not you !!

exactly, This is why I asked the question. Lets say you are using the standard servos that come with a trex 500 DS510m, this is a pretty quick servo, and loads of people have had a great experience of using them with their FBL conversion, does it actually get better if you upgrade to super quick servos or does it not make that much difference?

is the FBL system making such small adjustments that you don't take advantage of the speed of the servos because it is only moving 5 degrees or so at a time and as such much slower servos are almost as good a super fast ones?

if a servo takes .10 seconds to do a 60 degree arc then it stands to reason that to move 5 degrees takes 0.016 second
a servo rated at .20 seconds to move 60 degrees would take 0.033 seconds...

it is half as fast but who can tell the difference in real terms?

/Steve

[PeteStewardson]

I've been wondering the same thing. I'm half tempted to convert my 700e to FBL but at the moment I run MKS DS660A+ servos which are more F3C orientated so are slower ( 0.149s ) but have high torque and accuracy.

If these are too slow it seriously bumps up the cost of going FBL as it's already an expensive swap over anyway.

[Andy from Sandy]

Other important factors in a servo are its lag between a new position command and ramp up speed and stop on position speed.

Something else is deadband. If there isn't much of one the servo will chatter away when it is supposed to be stationary and I guess if it is too large then there will be a certain inaccuracy of where the servo stops in either direction.

[Mark_T]

if a servo takes .10 seconds to do a 60 degree arc then it stands to reason that to move 5 degrees takes 0.016 second a servo rated at .20 seconds to move 60 degrees would take 0.033 seconds...

it is half as fast but who can tell the difference in real terms?

The problem is that 'reason' doesn't actually get factored into this stuff when quoting specs

Those figures you are using are usually for a servo with no load, they don't tell you how fast the servo can move to a new position and stop nor do they tell you how accurately it will stop, mostly they tell you how fast it will move once it is moving with no load.

In other words the spec is actually useless for comparing the utility of a servo in most cases

As Andy has said there is a lot more to this than just a simple speed calculation.

[barney]

You need both speed and torque no doubt about that. The human reaction having a delay makes no difference as you have already decided which direction you are moving the Tx stick so ideally it needs to respond directly since you have already told the Heli where it is going (well mostly ).

A FBL just reacts to remove the interactions and the faster it does that the truer the model will fly.

Brian

[Andy from Sandy]

A FBL controller reacts to remove unwanted or uncommanded movement. So for the model to hover hands off if the wind blows at the models side to cause it to roll the controller will react to stop the roll. If this is your description of interaction fair enough.

[barney]

There are numerious interactions it removes that you would see in a Fly-barred model. For example control slop, servo gear wear, poorly set rotor-heads, rotor down-wash, uneven lift from lead/lag blade by main blades in FFF etc these are are all taken care of electronically. Just different wording for much the same thing. Theres a very fine line between uncommanded movement and self stabilisation though, but thats another topic in itself.

For best results good servos are required.

Brian

[Paulice]

the standard servos in the 500 are up to the job without a problem,
Thats all i have in my v-bared trex 500 ESP but i found the standard DS520 on the tail lacking so i used the 600 mount and fitted a BLS251.