moving weight inside a tube ?

Hi, I’m looking to move a weight inside a tube based on the angle of the boat. I think a gyro is the answer but i’m not familiar with gyro or how they work any idea ?


Laterally or Longitudinally ?

laterally - it’s for a trimaran. was thinking that a gyro which is designed to stabilise a plan/heli could do that - as the boat heels the gyro could control a servo to move the weight up to weather. for now i’m looking for something cheap and simple. Some gyros are expensive but other are cheaper but don’t know how they work and what to feature to look for when looking at the various gyros.

heard recently some guys use them in RC cars now to reduce skidding to improve speed over a course…

Hi Gilbert,
I’m not sure it will work because the response time of a gyro is very fast. This means that a servo will continuously get a signal to move for any little variations of tilt also when not required. In my opinion some additional electronics are need as such to tell to the servo when to move. A sort of minimum tilt angle before start moving and displace a ballast.
Probably a pendulum and a crossing detector could trigger an adjustable pulse width generator to the servo. Still some electronics are needed anyhow to control left and right pulse width.
Another idea is to combine a pendulum and a potentiometer as replacement of the servo one. No electronics needed and keep the proportionality.
The external servo wiring shall be in accordance with left/right tilt. see simple sketch


I have made a “proof-of-concept-thing” on the system claudio is explaining.


I think it should work, but have never tested it on water. Of course size of drums and servos should be adjusted to have quicker response. This was only to test the concept.

I just used a servo which could be modified to spin continuously and then moved the pot meter outside and attached a pendulum to it. On top of that a couple of end-switches. Also has a small cheap gyro in the system to react to quick changes.

I don’t know if an expensive gyro will keep the heading , but this cheap one looses the zero-point pretty fast, which would mean your cat would eventually capsize (maybe).

The system was intended for my next project when i finish the two IACC120’s I’m doing now.

Hope you can use it as inspiration.


so if I understand the 2 weight will work in opposite direction right ? with the pendulum of the potentiometer moving to leeward it would instruct the servo to move the ballast to the weather ? not a bad idea at all. In the light wind I’d be more inclined to move the ballast to the leeward hull to keep the boat heeled and helping the sail keep their shape - how could I do this ?

would a 3 position switch do it ? one to move the ballast to port hull, central position to move ballast to centre hulll and last position to move ballast to starboard hull. the switch on the Tx would be on a channel connected to a small sailwinch to move the ballast in the tube across the boat ? could that work (though manual)



Propperly it would be possible to make a simple logic control to have 3 fixed positions.

But there is another good thing about to the system in the video. As it is still connected to the RX, you can actually offset the zero-point by moving the stick on the transmitter. It is what I am doing in the beginning of the video, but the as platform is not moving much, it is not clear how it works.

But it should be possible to move the stick on the transmitter and have 2 deg leeward heel if light conditions.

It is so simple, that it is almost hard to under stand. If you have an old servo, I suggest you take it apart and play around with it. Then it becomes very clear how it works.


As complement of what Anders says,
the potentiometer pendulum when is in the vertical position for instance and installed inside the hull, the ballast servo will be is also in the neutral position and the mass sitting at the boat center. You can still play with the joystick, even not necessary, so far the servo is initially set at the neutral position.
You do not need anything else, no gyro, no electronics, just straight forward.
Up to you to decide how to operate the servo and ballast movements.
I do not know How much space you have inside the hull, but for sure the shifting system that Anders showed is a good solution.
The ballast could travel on a tube and the strings connected to the servo winch as normally dine with sail sheeting.
The potentiometer pendulum may go left and the servo may turn right, it depend on the wiring position on the potentiometer.
A will insert a small sketch to explain and as Anders says is more difficult to explain then to do it !

the sketch include mono hull balance mass with a servo arm capable to turn 180° if this may make sense !!

I will take apart an old servo this weekend to check as Anders recommended - your drawing and Anders’ video start to make sense no - so the potentiometer sends a signal to the motor of the servo to tell it to turn one way or the other ? sorry if it’s a basic question but my knowledge of electronics is quite poor - I’m alright on basic electrics and soldering to all my electrics in my boats but that’s where it stops so I appreciate your help on this.

Indeed it sounds simple, effective and in theory I shouldn’t have to spend more money as I should have all I need already… I’ll need to spend few hours in the shed to test that out.

Your drawing is what I sketched out (sorry only on paper) : a small sailwinch in the main hull linked to the ballast in the tube by poleys and a simple sheeting system that goes on either end of the tube to move the ballast inside the tube as the sailwinch turns one way or the other. very similar to your drawing. but I didn’t know how to control the sailwinch in a ‘automatic’ way so I can concentrate on the sailing of the boat. I’ll do a test hopefully very soon and let you know.

Cheers, Anders and Claudio :slight_smile:

Hi again

Well I have a little better time now to explain, was at work earlier :spin:

Lets just think how a servo works normally. Don’t know if i got it right but below is how it makes sense to me.

Inside the servo there basically is some electronics and a pot-meter. Pot meter is attached to the servo axle (to which we connect the arms or drums.)

In a magic way the electronics are keeping some resistances in a equilibrium/balance. The pot-meter has 3 wires going into it. If you measure the resistance between the center wire, and one the other two, it will be the same, if the pot-meter is in the center position. If the pot-meter is turned, the resistance changes. One side goes up, the other goes down.

When we move the sticks on the transmitter, we disturb this fine balance, and the only way the electronics can compensate to bet back in equilibrium, is by turning the pot-meter. To turn the pot-meter it has to start the motor. The motor will then run until the new equilibrium has been found.

The catch is that if you move the pot-meter outside the servo, there is noting to tell the electronics to stop, so the motor will continue to run until the equilibrium is reached. So if the gearbox can not rotate freely 360 degrees it will break the gears at some point. Most servos has a pin or something on the biggest gear in the servo to protect whatever we connect to it. So before playing around with servo you are taking apart, make sure the gearbox can rotate freely.

You will need end-switches if you move the pot-meter outside the servo. Alternatively you can make a system that relies on friction to move the weight and have a mechanical stop of the weight, in stead of switches.

Don’t know if the above helps on the understanding. But give it a try, i’m sure you will be able to see how simple it actually is, and make up a solution that suits your needs.


This is another schematic shoving the principle set up.
I do not think Anders that it is necessary an end stop switch if the pendulum angle is limited by mechanical stops !

What it is needed is to find a compromise between the pendulum swing angle and the potentiometer resistance value for the same rotation angle . I toke a 10k pot knowing that normally is used a 4.7k for 350°. I hope that 10k is not going to reduce too much the signal error amplitude due to lower current passing trough. Again a test shall be done.

Thanks Anders, Claudio,

I have actually taken apart an old servo and now start to understand what you’re talking about. Once I removed the inside of the servo I connected it to a Rx and powered up to see. The motor of the servo turned straightaway but by moving by hand the potentiometer top shaft I got to a point where the motor stopped alright. this must be the centre point and then by moving it one way or the other the motor turns one way or the other continuously. So I get why we need end point switches to stop the motor.

Don’t ask me the spec of the motor or potentiometer as there is nothing written on them - just took them out of a standard size servo.

I’ll need to see now to build a proto of the weight moving in the tube with this system next…

Cheers both for your help @-)

Hi ,
probably you miss something !
The potentiometer you have moved, shall be substituted with an external one on which you attach the arm pendulum as depicted previously.
For that purpose, you shall unsolder the 3 wires connected to the internal servo and connect them to an external potentiometer equipped with the pendulum (an arm and a weight).
As you have already observed, moving the pot, the servo turned until stop by himself once the signal comparator provide a zero voltage. This is why you do not need an “end switch”.
Assume the external pot with a vertical pendulum position and the joystick at zero or central position.
The servo will move until stop reaching the zero voltage.
Now swing the Arm pendulum by 45°, you will see the servo mowing again until it stop once the zero is reached.
Similarly will occurs by turning again the arm pendulum.
The joystick shall be mechanically blocked at the central position, otherwise the servo will start moving and searching for a new zero !
You will thus notice that the pendulum and the joystick will be acting at the same way up to a certain limit.

Now the pendulum inside the hull will be free swinging with at the extreme some internal mechanical stops as such that the pendulum cannot further rotate. The servo will “copy” that movement without any other stopper !

The question remain with the potentiometer value.
Normally the internal potentiometer is around 4.7kohm out of 350° of rotation and this valid for a full arm swing or winch.
What it is necessary in this application is to provide the 4.7Kohm variation within the pendulum swing that will much lower then 350°. Probably limited to some 180° maximum.
For that reason, the potentiometer shall have a value around 10/15Kohm, but a trial is needed !!

BTW the "pendulum function is already in your hands !!! It is the joystick itself. (see picture)
What you need is to make the same function outside the transmitter and “close” to the servo or at least inside the hull.
If you open your transmitter, you will observe as it is done inside and you will see that at each joystick function is attached a potentiometer.
Each joystick is mechanically limited for the swing !

Very clever thinking and demonstration Claudio!

I have to apologize for what I wrote before.
I dismounted from my old AC33 the winch servo with an external 10 turns potentiometer .
After removal of the ‘linked’ gear, I added an arm to the potentiometer axis with some weight in order to get a pendulum function.
I also zeroed the joystick to stop the servo turning.
I modified the pendulum arm position and the servo started mowing in one direction and continued turning without stopping independently of the arm stoppage.
This is the point where I made a mistake, of course the direct feed back link was not existing anymore.
To stop turning I was obliged to reposition the arm to the original position.
Moving the arm in the opposite direction also the winch started rotate in the opposite direction . To stop rotation was necessary to bring back the arm.

You were right, to stop the winch rotation it is necessary to break the power supply line with an End Stop switch

Sorry for that missing understanding.