Bell and Hiller Rates

[mickeyjohnston]

Tomorrow I'm going to try to make some adjustments to my 3G system. Can someone explain to me in different terms what the Bell and Hiller Rates do. Everything else makes sense for the most part.

Thanks in advance

[HiroboFreak]

The software has info at the bottom of the app when you mouseover the adjustable parameters.

If you don't know what they do, why would you want to change them?

[mickeyjohnston]

I want to make sure I'm adjusting the proper thing for how the heli is acting. So am I correct in saying the hiller rate is like changing the weight of paddles?

[wasabro]

Originally Posted by mickeyjohnston

I want to make sure I'm adjusting the proper thing for how the heli is acting. So am I correct in saying the hiller rate is like changing the weight of paddles?

To my understanding Hiller Rate use to stabilize the drift and Bell Rate control how fast the heli react for a given cyclic input.Please correct me if wrong.

[mickeyjohnston]

Hirobofreak, Thanks but I'm aware of the mouse roll over. That why I said in different terms. I only have an hour for lunch tomorrow, I wanted to get this dialed in so I can move on to my other heli's. I hate taking up air space at my local field by tuning a heli. I want to get the electrics set for the weekend.


Thanks Wasbro

[rbeall]

I have been thinking about these setting the most. I have been trying to come up with what they are actually doing in software to make the "bell-hiller" type setup. The Bell-Hiller setup is not exactly described by what they say their controller is doing. Here is my guess as to what they say they are doing.

I'll write it in pseudo code and equations so maybe it will be more obvious specially beings they say not to go below 100% I'll explain how I would have done it differently and in more detail below

gyro_feedback = ((pilot_input*control_rate)-gyros)*total_gain_compensate;
raw_servos = (bell_gain * pilot_input) + (hiller_gain*gyro_feedback);

I would have set it up exactly like it is on helicopters where the user has the ability to vary bell-hiller mix. It is not two gains it is really one. By making it two gains they have to add their stipulation that it has to be above 100% because their implementation is flawed in my opinion. So I would have done it like this:

***edited!
bell_hiller_ratio = 0.30; (or 30%)
hiller_gain = bell_hiller_ratio;
bell_gain = 1-hiller_gain; (or automatically the complement of 100% --> 70%);

***edited!
gyro_feedback = ((pilot_input*control_rate)-gyros)*total_gain_compensate;
raw_servos = (bell_gain * pilot_input) + (hiller_gain*gyro_feedback);

With that being said, beings I am pretty certain that this is how they have their software set up you should probably model your gain adjustment of Bell and Hiller on the 3G keeping in mind they must always add to 100% For unobvious reasons to common users, going over 100% really throws a lot of nonlinear quirks in your tuning process. Bottom line is keep the total of Bell and Hiller at 100% always!

My general gouge numbers would start with 50% 50%. The problem is, you have two moving targets when you set it up this way. So that's why I say 50-50. Say for example your gains for pitch and roll (total gain, lock gain, and stop gain) are way wrong. If you have Hiller greater than 50% then that percentage of bad or wrong control inputs are going to the heli.

With that being said, you should be able to fly H=100% B=0% if your gains (total gain, lock gain, and stop gain) are pretty close. Which isn't that hard to do (wikipedia PID controller).

With a little bit of back ground in full scale helicopter flying, controls engineering, and aero engineering I would have to say that it is no surprise that there are as many problems as I am seeing here with their setup. Even with the 3G link software, I think they set it up so that people could still get themselves into trouble. Especially beings half of how they describe things is either partially incorrect or completely misleading for most people. IE less hiller is good for 3d....ehhh more hiller is good if your pitch and roll gains are rock solid. It really is a big picture type of thing that they just are not capturing.

[mickeyjohnston]

Thanks for the reply. I'm slowly understanding it and how the two work together or in some cases fight eachother. If I understand it right, the 50 50 ratio will only work if Align made these settings linear and equaling the same so they don't fight each other. I'll be really curious with your findings after you test the Align system.

Let's say I have both at 50% and I turn hiller up to 60%. This should make the heli roll faster. Now if I put the hiller at 50% and turn up the bell to 60% this should also roll faster. So my question is, would I feel a difference on how it would fly based on how you understand the software? How should it feel if the software is correct?

[rbeall]

Just to maybe explain it slightly different.

If you increas Hiller to 60% then Bell must go down to 40%. Never change one without the other. If you do it will be harder to logically tune. As far as feel goes, because there are two things going on here, it is somewhat hard to describe how it will feel airborne.

100% Bell would feel like you had just hooked the servos directly to the reciever outputs.
100% Hiller would feel like flying a heading lock gyro on the tail except for all three axis

Both of these don't really equate to being "faster" on the controls one vs the other see why? The second order effect however has the potential to do exactly what you are talking about. So there is kind of a paradigm shift here and it is based on what is doing what.

In a nut shell here is the simplest way to look at it:
Bell-Hiller mix is the "being on rails" feel. The higher the Hiller, the more you are controlling the body rotation rate on all three axis vs direct control. Just becuase you are controlling body rates vs direct inputs doesn't make it faster or slower of a response. Here is why:
The description makes it seem like lowering Hiller would make it faster on the sticks. Well that is true for one case out of many. It could actually make it slower, it depends upon a bunch of different variables.

Bell-Hiller is not the most direct way to make the controlls faster. The best way to make the heli more sensitive to the sticks is a couple of things:

slow sticks: high exponential, low control rate gain, low tuned total/lock/stop gains with high Hiller low Bell, low swash return speed, high cyclic pitch motion delay.

fast sticks: the opposite of the above.

All of these variables can fight eachother, overpower, tradeoff, etc. So to say it is just one thing it quite the understatement. Some of things things will actually do the oposite of what Align quotes if others are overpowering the ones above them. For example, Decreasing Hiller gain is suposed to increase agility. However, if control rate is really high and total/lock/stop are really high but stable the stick speed may actually go up.

I hope this made it as clear as mud. I probably made it harder for you to understand, but you kinda have to get a handle on how each one of knobs you can tune affects the other. They do different things in different ways but the toatal outcome can be predicted if you have a general understanding of what they are doing.

[mickeyjohnston]

Once again thanks. I understand this more now then when I first posted.

[Heechee]

What's the 3G system's default for Bell and Hiller?

[mickeyjohnston]

rebeal,

What I noticed today is that Align is putting the default for the 250/450 at hiller 60 and the bell at 50. The 500/700 is hiller 50 and the bell at 70. What is your thought on these settings and why they would do this?

[rbeall]

Note I changed the equations. I made a pretty significant mistake somehow. Your question made me catch it and here is why.

Hiller is lower on the larger helicopters because they are naturally more stable.
Bell is lower on the smaller helicopters because they are more manuverable.

[mickeyjohnston]

Rbeal thanks for educating me on these adjustments. Someday I'll be able to pass on this to others.

Would raising the Hiller create more bobble while pro flipping? From my understanding raising the hiller could make the heli almost fight your inputs, am I correct? More like a gyroscopic affect?

Here's what's confusting, if changing the bell makes it more responsive, could you active that by other adjust like cyclic pitch or 9 expo?

[rbeall]

increasing Bell doesn't always make it more sensitive. IE if you had high control rate gains and high total/lock/stop gains then increasing Bell might actually make the conrolls softer. It all depends..

The gyros feedback is only used as a method of control not necesarily stability. Stability is a by product but they are using them for controlling the heli's body rotation rates. When you move your stick say 10deg (say for conversations sake its one to one control value going to the helicopte cpu) the computer says ok the pilot wants 10deg/s rotation rate of pitch. It uses total/lock/stop gains to get the servos to do what ever they have to to get that rate. Kinda like holding a slow piro rate on the tail with a stiff wind. The heading hold gyro doesn't care about the wind it will hold as best it can a constant piro rate even though the wind can be pushing it faster at times or pulling it slower at times it will always manipulate the controls to get the commanded rate. If it is zero or centered stick then it will stay there.

Don't look at hiller/bell as a stability or controllabilty issue. It is more of a "how much" fly-by-wire control do I want to give the computer. High Hiller give more control to the computer and zero gives none.

So if Hiller control from the computer is low because total/lock/stop gains and control rate gains are low, it doesn't really matter how high you raise Hiller, it still isn't doing much. That's why you have to look at the whole picture and tune each setcion. They all have to play nice with eachother.

Bell is the easiest to understand. It is just a direct input from the transmitter. It doesn't care about gyro feedback or body rates or anything. It's pure direct link multiplied by your Bell gain percentage.

[Vinger]

From Wikipedia

Stanley Hiller also arrived at a method to improve stability by adding a bar perpendicular to the rotor, but he added short, stubby airfoils, or flaps, at each end. Hiller's "Rotormatic" system was used to deliver cyclic control inputs to the main rotor as a sort of control rotor, the flaps providing added stability by also dampening the effects of external forces on the rotor.

from RC Helimag http://www.rchelimag.com/pages/howto...owto=22&page=1

Background:
Most RC helicopters in use today have a flybar. The flybar provides stability without the expense of maneuverability to our helis. In order to better understand why flybar ratio adjustments do what they do, we need to understand how they work through the history of its design.

Bell and Hiller
There are two systems that are combined and mixed together to control our model heli rotor heads: the Bell and Hiller systems. Both the Bell and the Hiller systems get their names and designs from full-size helis. Let's explain them both briefly:
The Bell control system is based on the Bell Stabilizer bar. The Bell stabilizer bar was used in many early Bell helicopters and is most commonly seen on the Bell UH-1 Huey. The stabilizer bar is basically a weighted flybar without paddles that spins with the rotor head providing gyroscopic stability. A mixing arm on the stabilizer bar takes inputs directly from the swashplate and mixes them with the gyroscopically stabilized bar to the blade grips. Flight control systems using a stabilizer bar benefit from direct control input, but offer limited maneuverability due to the overriding stability of the system.

• On a pure Hiller setup, the cyclic input from the swashplate is sent to the flybar; the flybar is then flown to the desired disk angle. The main rotor grips are attached to the flybar, and the tilting flybar will move the main grips to follow the same circular path of the flybar. The Hiller flybar is gyroscopically stabilized as the Bell Stabilizer Bar but now offers increased maneuverability as it can be flown to a new position. The downside to a pure Hiller setup is the lag in controls. The flybar needs to move itself into its new position before the main blades do, as the flybar in a Hiller system pulls the main disk behind it. This results in a small delay before the effect is realized at the main rotor disk.

Bell-Hiller System: The best of both worlds
How do you enjoy the pros of each system while diminishing the cons? Simple. Just combine them! Most RC helis use a combination of the above systems called a Bell-Hiller setup. By offering a direct input from the swash to the main grips via a mixing arm (Bell input) and controlling a maneuverable flybar (Hiller input) you get stability and maneuverability without any delay in control input.
The Bell-Hiller mixer arm is where all the magic happens. It takes the combined input from the swash and the flybar and transmits the resulting output to the blade grips. The use of a mixer allows the blade grips to get the desired instant input from the swashplate while also getting the needed control and stabilization input from the flybar. When the flybar tilts, it raises or lowers the mix arm, and will thus raise or lower the pitch on the main blades. By making adjustments to the mixing arm you can fine-tune how much influence you want the flybar to have over the main blades.

Bell input Adjustment:
The Bell component is usually set at factory to what the manufacturer has determined to be the best for the given heli. There are very few model helis on the market with adjustable Bell input; most let you adjust only the flybar (Hiller) side of the mixer arm. If you have a machine that allows Bell input changes, and you decide to experiment with these settings, be sure to check the total collective pitch range as you do so. Changes in the Bell input affect total pitch range much more than Hiller changes. This is because the washout removes all collective pitch influence from the flybar (Hiller side), leaving the Bell input to transmit all collective pitch changes to the grips. Making changes to the Bell mix is generally done by more experienced pilots looking for the perfect mix for their flying style. Making changes to the Bell side of the input is done using the same steps as making changes to the Hiller input, so the procedure below is more or less the same.

It can seem confusing at first, but if you read the manual and try the different settings out on the bench, it will make more sense. Try tilting the flybar with the heli on the bench while watching how much the blade grips rotate on the spindle. The amount of movement generated at the grips by the flybar input increases as the flybar ratio is changed to a higher setting—just as lowering the flybar ratio will reduce the amount of movement you see with this exercise.

How to Measure Flybar Ratio:
With all the prerequisite info behind us, we can move on to working with the flybar ratio. First, let's define what exactly we are measuring, here, so it's not confused with the Bell and Hiller ratios. The flybar ratio is simply the correlation between the flybar tilt and main blade pitch, measured in degrees. The standard form is to measure how much main blade pitch is given per each degree of flybar tilt. For example, a Raptor 50 Titan is equipped stock with a 1:1 flybar ratio. What this means is that for each degree of flybar tilt, the main blades actuate one degree as well. It's a 1 to 1 relationship in this case. 1:1 is considered a fairly high flybar ratio, while .5:1 would be a low flybar ratio, as the main blades would only move .5 of a degree for each degree of flybar input.

An interesting sidenote on the Raptor is that due to its 1:1 ratio, the flybar does not need to be level when setting the pitch up on this heli if you use the flybar as a guide to set pitch to. Any heli with a flybar ratio other than 1:1 needs to have the flybar leveled while setting pitch.

Calculating the flybar ratio on your heli is quite simple as long as you have a pitch gauge with a bubble level on it. If your pitch gauge does not have a level, you can simply attach a small one to the top or bottom rail. The other tool you'll need is a protractor, or anything else that measures angles.

1. The first thing you do is tilt the flybar in one direction (it doesn't matter how much tilt you put in). Measure the angle that the flybar is tilting by using your protractor (make sure that the protractor is level). Keep the flybar locked at that position and use your pitch gauge to find out what pitch the main blades are at, and again check for level. Write both of these numbers down.


2. Tilt the flybar in the other direction and re-measure in its new location. Also check the main blade pitch at this new position. Write these numbers down as well.


3. Use the two main blade pitch numbers to figure out the total travel between the two. For example, if one end was -3 and the other end was +4, you would have 7 degrees of travel between these two points. Now do the same with both flybar angle readings. Let's say, for our example, that we had 9 degrees total range on the flybar tilt.


4. Simply divide the main blade travel by the flybar tilt travel we just found to get the ratio. In this exercise it would be 7/9 or .78, meaning we have a .78:1 flybar ratio in this exercise. You want the main blade travel in the numerator because you are trying to get a ratio of how much main blade pitch each degree of flybar tilt creates. In our example, you get .78 degrees of pitch change for each degree of flybar tilt.

It's wise to run this calculation a couple of times using different flybar positions, in order to verify your results. While you do not need to know the flybar ratio of your machine in order to fly, it's interesting data to have and can be useful for fine-tuning a machine.

The flybar ratio is only one part of the equation regarding head adjustments. The length of the flybar and the dampers used will affect how a heli flies, as does the size, weight, and shape of the paddles. A flybar paddle that has a large surface area placed farther out from the main shaft (longer flybar) will have more influence than a small paddle placed closer in. For example, on a machine with a high flybar ratio you can get more powerful cyclic by using light paddles that have a large surface area, or use a longer flybar. To experiment, start with the stock heli setup and change one item at a time to get a feeling for how each option changes the heli's flight style.


Conclusion
The flybar is one of those seemingly simple, yet important parts to a model helicopter. Knowing what your flybar ratio is and what options you have on your heli makes you a more knowledgeable pilot in the long run. Next time you have a hankering to tinker, try adjusting the flybar a bit. The results achieved by small changes are amazing. In fact, a simple change on the flybar can make a heli feel like a completely new machine in flight. Whatever you do, enjoy!

So we deduce from the above info that both the Bell and Hiller control system provide extra stability to the rotorhead. It is just a problem that the Bell controls provide a certain lag to the controls, where the hiller system does not.

In the Align implementation of their electronic stabilised system, they have combined the two control systems to mimic a flybar head as close as possible. So adjusting the Hiller higher, you will have to lower the Bell mixing effect and this will result in a quicker reacting head with less stability. If you only raise the Hiller effect then you will get quicker reactions, but with the same stability, but bear in mind that Align has put a 100% cap on the two effects in their algorith so this is theoretically not possible.

So the short is, more Hiller quicker reaction and more Bell effect will slow the reaction of the head to control inputs down with more stability.

[mickeyjohnston]

So it's safe to say that the reason why the flybared 450V2 and the 450 Pro fly so different is due to the bell and Hiller setup. It's becoming more clear now. In order to perfect this bell heller ratio, it will take time and testing. I can see how it would be easier to tune this ratio on a flybared heli than a flybarless system. The FBL system has other adjustments that could be masking the bell and heller ratio along with software.

[mickeyjohnston]

Thanks again for all the replies, two days ago I was really lost. Now I have a little understanding about it. I had a person out at the field say just dont mess with them, were I look at it as why not understand it.
So correct me if I'm wrong here, with the 3G system they have an aileron control rate and roll EXP level setting. Couldn't you achieve the same feel by adjusting these settings compared to the bell and heller?

[rbeall]

This isn't complete or 100% accurate beings I havn't don't have a cable or unit to test it, but this layout is very close to how the software is layed out. A few things may be slightly different but this flow chart should help a ton!

-Beall

[mickeyjohnston]

Originally Posted by rbeall

This isn't complete or 100% accurate beings I havn't don't have a cable or unit to test it, but this layout is very close to how the software is layed out. A few things may be slightly different but this flow chart should help a ton!

-Beall

WOW, based on what you see so far, would you say it would be ideal to run the swash return speed as low as possible and turn the stabilty with the other adjustments?

[rbeall]

@Vinger

The way Align has implemented their software it is clearly evident that their Hiller/Bell gains resemble nothing of what the physical Bell/Hiller head does. Those two gains are simply something you would include on any digital control system that has a user command input. They just call them that to have the user base sort of understand them. And in my opinion it is very misleading because for the most part it they can resemble sort of what the Bell/Hiller mix does on a normal head. However buecause there is a lot more in front of the Hiller gain you can clearly see that the digial system can very easily vary from the physical description so just take that into consideration when tuning gains and be cognisant of how which ones affect the others.

**note my flow chart

-Beall