Bobby Watts Tail

[lazor 22]

Same as 255 stock tail???

[clytle374]

Originally Posted by lazor 22

Same as 255 stock tail???

It has a reducing collar to fit the smaller boom. The hub, grips and blades are different. Not sure about the shaft/pulley.

Cory

[Ampdraw]

Anyone know if this tail has built in pitch (opposing torque) with the bell crank at 90 degrees??

Been looking into tail geometry alot and the v2 tail is balls'd up imo. Has the same "issue" as the 450 pro.

When the bell crank is at 90 degrees to the boom the tail blades actually have pitch with torque (nose left)! Meaning that when the tail blades are in the correct position for rate mode setup the crank lever is forward alot...introducing non linear travel in a big way.

Here is a vid that explains different heli tail geometry
http://video.helifreak.com/?subpath=...otherhelis.wmv

Amp

[Picofly]

Originally Posted by clytle374

It has a reducing collar to fit the smaller boom. The hub, grips and blades are different. Not sure about the shaft/pulley.

Cory

Shaft and pulley are identical. Tai fin is different (G10 instead of CF). IMHO, Gaui 255 tail fin is much better - more clearance and better material.

[newhelliguy]

The arm has an offset.

When you put the blades even the arm in forward of 90(first picture).

When you place the short part of the arm where the ball attaches at 90 the blade are set to turn left (second picture).

This is a great tail upgrade for this heli.

[Ampdraw]

Originally Posted by newhelliguy

The arm has an offset.

When you put the blades even the arm in forward of 90(first picture).

When you place the short part of the arm where the ball attaches at 90 the blade are set to turn left (second picture).

This is a great tail upgrade for this heli.

Thanks for those pics. Seems the geometry is pretty much the same. The first pic shows that at 0 pitch the bell crank is not 90 degrees... rather it is forward. The second shows that as the bell crank is pushed backward so it is closer to 90 degrees but this introduces nose right pitch (with torque). The bell crank should be under 90 degrees with the rest of the geometry as it is.

Why don't they do the bell crank with a 90 degree bend, and geometry so that the slider ends up centre of full travel with built in pitch opposing torque??
So rate mode setup would yield a 90 degree bellcrank, which would be parallel to the 90 degree servo horn, with the slider in the centre of travel and the pitch at 8 degrees leading edge toward the case (ie nose left)??? As it should be.

Can't think of a positive for introducing differentiation in the tail like they do.

All it would need would be a redesigned crank and a longer tail shaft. Maybe shorter slider to grip links would get there and they could keep the same length shaft.

Comments welcomed...

Amp

[clytle374]

Originally Posted by Ampdraw

Thanks for those pics. Seems the geometry is pretty much the same. The first pic shows that at 0 pitch the bell crank is not 90 degrees... rather it is forward. The second shows that as the bell crank is pushed backward so it is closer to 90 degrees but this introduces nose right pitch (with torque). The bell crank should be under 90 degrees with the rest of the geometry as it is.

Why don't they do the bell crank with a 90 degree bend, and geometry so that the slider ends up centre of full travel with built in pitch opposing torque??
So rate mode setup would yield a 90 degree bellcrank, which would be parallel to the 90 degree servo horn, with the slider in the centre of travel and the pitch at 8 degrees leading edge toward the case (ie nose left)??? As it should be.

Can't think of a positive for introducing differentiation in the tail like they do.

All it would need would be a redesigned crank and a longer tail shaft. Maybe shorter slider to grip links would get there and they could keep the same length shaft.

Comments welcomed...

Amp

I think you might be over thinking it. The tail seems to have plenty of travel in either direction. The piro rate terrifies me, either direction. That angle is there is put the ball link back onto the center of rotation. Note where the pivot point is. ie ignore the angle of the arm, and take the angle from the ball to the pivot bearing. 90* is close to 0* pitch.

I'm not explaining that the best, if it doesn't make sense let me know and I'll take some pics.

Cory

[Ampdraw]

Originally Posted by clytle374

I think you might be over thinking it. The tail seems to have plenty of travel in either direction. The piro rate terrifies me, either direction. That angle is there is put the ball link back onto the center of rotation. Note where the pivot point is. ie ignore the angle of the arm, and take the angle from the ball to the pivot bearing. 90* is close to 0* pitch.

I'm not explaining that the best, if it doesn't make sense let me know and I'll take some pics.

Cory

Centre of rotation?

Yep when I am referring to the 90 degree bellcrank I'm looking at the control rod link ball against the pivot point.

Not an issue of travel really regarding piro speed etc. More a case of being able to set up rate mode and then having exactly the same gyro limit for both sides.
This allows the gyro to work with the same resolution both sides.

Another issue that is introduced is the amount the tail blades actually move for a given push rod movement (differentiation). With the crank arm forward of 90 the blades will move more for a push movement than they will for a pull. So the servo has to travel farther for a pull movement to create the same pitch change.

This is not easy for the gyro to work with. If it tries to do both an equal pitch change one direction then the other (as an example and from rate setup 8 degrees) it will have to move the servo say 30 degrees back and then say 40 degrees forward in order to make this happen.

Amp

[clytle374]

Pivot point or center or rotation, same thing. I guess center is centre depending on if you live on the top or the bottom of the world.

I thought maybe you were talking about the angle of the cf arm, it looks a little deceiving at first.

I walked a mile for this picture, so everyone else has to look at it too. Yep, it has 0* pitch with the arm at 90*.

Another issue that is introduced is the amount the tail blades actually move for a given push rod movement (differentiation). With the crank arm forward of 90 the blades will move more for a push movement than they will for a pull. So the servo has to travel farther for a pull movement to create the same pitch change.

That might be exactly why they don't have hover counter torque set into the tail with a 90* arm. This way the gyro has a truly centered control over the tail.

I know for a fact rate mode setup is necessary for a rcmart gyro. But I'm wondering if something like a Quark isn't better off without it. Tell the qyro where zero tail thrust is at since it obviously knows that there will be a non-constant thrust required all the time. I set mine up at zero since the book didn't say other wise. And I can't fly hard enough to get it to mess up, but I can't fly very hard.

Cory

[sco88]

Originally Posted by Ampdraw

Centre of rotation?

Yep when I am referring to the 90 degree bellcrank I'm looking at the control rod link ball against the pivot point.

Not an issue of travel really regarding piro speed etc. More a case of being able to set up rate mode and then having exactly the same gyro limit for both sides.
This allows the gyro to work with the same resolution both sides.

Another issue that is introduced is the amount the tail blades actually move for a given push rod movement (differentiation). With the crank arm forward of 90 the blades will move more for a push movement than they will for a pull. So the servo has to travel farther for a pull movement to create the same pitch change.

This is not easy for the gyro to work with. If it tries to do both an equal pitch change one direction then the other (as an example and from rate setup 8 degrees) it will have to move the servo say 30 degrees back and then say 40 degrees forward in order to make this happen.

Amp

But equal pitch change doesn't give you equal yaw rate because one works with the rotor torque and one against - perhaps they are trying to equalise the yaw rate with the geometry.

[clytle374]

Originally Posted by sco88

But equal pitch change doesn't give you equal yaw rate because one works with the rotor torque and one against - perhaps they are trying to equalise the yaw rate with the geometry.

I think that is one of the reasons amp wants the 8* build into the tail at center. But mostly to make the gyro work less.

I was thinking about that last night and realized something I read in the Quark manual might influence this. It states that if the prio rate exceeds 3 rev per second the sensor will be overrun, and will then send the servo to the travel limit. So up to that point the gyro is controlling the rate, not the max tail pitch. Of course if you exceed the 180rpm piro rate then that will be an issue.

I'm guessing on lift off it finds about where the tail holds and works from there. And then any movement commanded by the receiver is then used as a 'degrees per second' command and controlled through a closed loop with the sensor. Of course this is all speculation, unless someone at one of the major gyro companies wants to share the formula the gyros are using.

Cory

Then I thought, lets ask them. Spartan's FAQ on the subject. http://www.spartan-rc.com/resources/...=2&category=20

[sco88]

But if you could equalise the yaw rate with the geometry that would make life easier for the gyro since positive or negative rotation at the servo would give the same yaw rate.

[clytle374]

Originally Posted by sco88

But if you could equalise the yaw rate with the geometry that would make life easier for the gyro since positive or negative rotation at the servo would give the same yaw rate.

In human terms yes, in computer terms I doubt it. Without yaw commanded the gyro is sensing moment and making changes to the servo to stop it. Then lets say we give it 1/4 stick to the right, the gyro reads the input and turns it into "move right at X degrees per second" it starts moving the servo and comparing the outcome to the desired speed(X). It increase the servo until it moves at X. As inertia is overcome the speed will increase to over X and the servo will move back towards center. I would also assume that as it approached X, it would start backing down servo position. I can almost promise you that the gyro doesn't say "I need to turn right at X" and respond with anything like "that means the servo need to be at Y" like a human would learn to do. If it worked that way there would be a couple of dozen settings required to set up the gyro, or some fuzzy logic to learn it.

Of course this is a complicated formula instead of the feel and learning that your mind would use to accomplish the task. I also greatly doubt that it has any learning capability, it is too small and the tail would behave funny at first on each flight. For a better idea how this stuff generally works look here http://en.wikipedia.org/wiki/PID_controller.

I'd safely bet money that PID is what makes a HH gyro a HH gyro instead of just a rate gyro.

The wonderful part of these little gyros is how little tuning is required. I'm guessing these tiny helis would benefit much more from the pc link software than the larger helis since the tuning of the PID loop depends on the weight of the heli. Probably why the tails are more prone to bouncing at piro stop than the big helies. (or so I've gathered)

Cory

Such conversations I find require lots of hand gestures. Sorry, I was busy typing

[sco88]

Very definitely they are advanced PID controllers. I believe they also have a 'learning function' - the micro rondo manual describes a function where it learns ' the hovering offset' - this is what the manual says;

If your tail rotor does not stop evenly on both sides, μRONDO has the possibility to learn a hovering offset. Thereby the tail rotor
control is learned and stopping will be considerably smoother on both sides.
In order to learn this offset value bring your helicopter to a quiet hovering flight and switch the sensitivity channel Gain1 quickly four
times in a row. Whilst doing that and even about five seconds afterwards you should not move the tail control at all if possible and
the helicopter's nose should, if it is not completely calm, turn towards the wind. If the learning was successful μRONDO will confirm
with a slight tail twitch.
The stopping should now be the same on both sides.

[heathy]

Well, if your running the v2 tail then with the arm @ 90% you will have a automatic 8% for rate, thus making your gyro work less with the servo horn @90%

Not sure about over versions or the bobbywatts.

[clytle374]

Originally Posted by sco88

Very definitely they are advanced PID controllers. I believe they also have a 'learning function' - the micro rondo manual describes a function where it learns ' the hovering offset' - this is what the manual says;

If your tail rotor does not stop evenly on both sides, μRONDO has the possibility to learn a hovering offset. Thereby the tail rotor
control is learned and stopping will be considerably smoother on both sides.
In order to learn this offset value bring your helicopter to a quiet hovering flight and switch the sensitivity channel Gain1 quickly four
times in a row. Whilst doing that and even about five seconds afterwards you should not move the tail control at all if possible and
the helicopter's nose should, if it is not completely calm, turn towards the wind. If the learning was successful μRONDO will confirm
with a slight tail twitch.
The stopping should now be the same on both sides.

I really hate to call that learning, seems more like programming to me. Looks like μRONDO went about solving it differently. But really the programing to find that value at lift off would be quite easy, but not what I consider learning.

Cory

[Ampdraw]

Originally Posted by heathy

Well, if your running the v2 tail then with the arm @ 90% you will have a automatic 8% for rate, thus making your gyro work less with the servo horn @90%

Not sure about over versions or the bobbywatts.

Hmmm. Couple pics below of my tail.
Perhaps it is because I'm running a V1 bell crank mount?? I have V2 side plates, V2 slider but am still running the single arm bellcrank on the original V1 mount that screws to the side plates.
Perhaps this is longer than the proper V2 one??
Mine measures 7mm from the mounting face of the side plate to the centre of the hex machine screw.

1st pic shows 0 pitch and the short part of the bell crank is introducing no differential as it is exactly 90 degrees to the tail shaft. The long part of the bell crank is already forward and is introducing diff as its forward arc would create less pitch than a rearward.

2nd pic shows long side of the crank at 90 and the pitch at that point is incorrect. In fact it is almost spot on opposite of what would be required for rate hover idle up. This geometry also has differential due to the short part of the bellcrank although being it is so short it would have far less effect than than if it were long side of the crank as in the other pic.

Thinking if I shorten the crank mount at the case side (if the threads are deep enough I can get better geometry.

I am not that concerned with stick input. Rather how the gyro "reads" the mechanical setup of the tail and how effectively (not easily) it is able to achieve what it wants.

From my thinking (and I've been known to be wrong hehe) if the gyro "feels" the same range both sides (of rate pitch) and that range is similarly linear then the gyro will be able to be more accurate and require less corrections to effect the same goal.

The vid link I posted earlier shows that most helis (larger in most cases) have a 90 degree bellcrank (Gaui 200 is maybe 110 degrees). The 90 bellcrank at 90 degrees puts the slider at the correct pitch for rate mode. Then you have servo horn parallel with bell crank, slider in centre of travel with blades at rate pitch and everything ends up nice so you can get the same gyro limit (not rudder endpoint but these should be close) and a nice linear control for the gyro.

From my extensive and obsessive research many of the top flyers set up the gyro for either rate or centre slider depending on the geometry. The paramount goal is to get the same limit either side. If the slider ends up with the correct pitch then they do a rate mode setup. If rate mode setup puts the slider off centre then they don't do rate but rather set up for 0 pitch.

Here's an interesting thread that got me thinking:
https://www.helifreak.com/showthread...y+rate+anymore (particularily finless's posts and the responses back and forth including mention of how Nick Maxwell adjusted a setup for exact equal limit).

And the 450 pro thread:
https://www.helifreak.com/showthread.php?t=132747 especially "tail geometry", "the proof" and "tail fix mod".


Another exaggerated way to think of it...
Say the slider ends up quite close to the tail case for rate mode and has alot of travel the other way (toward the hub). If you set up the gyro limits to use all of this travel you will have a small "compressed" range for the gyro to use on one side, ie the gyro only has a limit of say 80 points to work with and on the other side has 140. Resolution on the 140 side would mean that side can be more accurate than the other.

Kinda getting really into the nitty gritty but I'm enjoying the discussion.

I've been running my Quark with rate setup. When I set the limits I maxed the endpoint and with a 6.5mm ball position get approx 140 toward the hub and 100 or so toward the case. I had mushy right piro stops and bounce on left stops.

Recently I set up slider for 0 pitch and then incramentally made changes. Now when setting up limit I stop at the 100 mark when the Quark led shows blue and pauses the servo movement. This is set both sides. My heli feels much more consistant both ways...in piro and on stop. I also removed the -32 expo I had on rudder (to reverse the built in + expo) and suprisingly the stops improved. These changes made quite alot of difference and I was able to bump the gain up a few points on gear channel. I set up gain from the bottom up, not from wag down.

Amp

[clytle374]

Amp, I think I'm with you now.

I was confused at first thinking you were saying the Bobby Watts tail isn't set up for 90* when you said the long bell crank arm was forward, and just assumed you were looking at the apparent angle of the cf arm. I guess you were talking about the v2 tail, since I see it clearly in your pictures. Did I understand you right this time?

Maybe you should turn the tail over so the natural pitch is the right way. IDK.

I agree that it is important for the gyro to have equal control for both directions. If not, you'd need 2 gain settings, one for each side.

Cory

[BHChieftain]

Hi,
Which tail shaft is used in the Bobby Watts Tail? Stock 200 V2 shafts?

Chief

[Ampdraw]

Originally Posted by clytle374

Amp, I think I'm with you now.

I was confused at first thinking you were saying the Bobby Watts tail isn't set up for 90* when you said the long bell crank arm was forward, and just assumed you were looking at the apparent angle of the cf arm. I guess you were talking about the v2 tail, since I see it clearly in your pictures. Did I understand you right this time?

Maybe you should turn the tail over so the natural pitch is the right way. IDK.

I agree that it is important for the gyro to have equal control for both directions. If not, you'd need 2 gain settings, one for each side.

Cory

Comparison of both your pic of the Watts tail and my 1st pic of the v2 tail they both have very similar geometry.
At 0 degrees pitch the short part of the crank is at 90 degrees to the tail shaft BUT the long part is forward by quite alot. Being the crank is not made with a 90 degree bend we can only either get the short part at 90 or the long part at 90.
My theory is (and my thinking may be flawed but I'll give it a shot until the cows come home lol) having the long part at 90 is more important than the short, as the long would introduce more non linear slider movement being off 90 degrees than the short part would.

If the tail is set up for 0 pitch (which is the better option imo for the 200) then the long part of the crank is forward. If set up for rate mode the long part is even farther forward. Forward crank (pull movement on the control rod) is pitch against torque ie nose right. Now being the arm is forward this side has less resolution as the arm being forward means that as the arc increases (arm moves forward more) the amount of pitch vs control rod movement ratio changes. The more forward it goes the less actual pitch change you get for that movement. It would seem to me this is the side you want the resolution on...ie it would be better for the crank to actually have a less than 90 degree manufactured bend than the current situation.

Looking at my tail if I reduce the bell crank mount by 1.5mm where it mates to the side plates effectively pulling the slider closer to the case and leaving the hub the same distance I should see 90 degree bell crank (long side 90 degrees to cf tail control rod) and also rate hover pitch of approx 8 degrees nose right.

Currently 90 degree on the long side of bell crank yields 8 degrees nose left ie with torque. Same thing with the Watts tail. If the cf part of the crank arm is moved back so the control rod link ball and the cranks pivot point are at 90 degrees to the tail boom (or cf tail control rod) the blades are pitched for nose left.

Amp