Choosing the right motor for project

[pityocamptes]

One of the members here has been kind enough to help my son and I with a school project. Basically it is a pan/tilt gimbal that will have a mannequin head on the top that is controlled wirelessly by a head tracker - so as your head moves the mannequins head should also follow suit. I don't foresee the head weighing more than a couple pounds, though not sure of the gimbal weight. The member needs the type of motor I should use, and since I am not very familiar with this I am not sure. He indicated that it should not be a servo, though for the project it would be nice for the head to center upon start up or by pushing a button, etc. - not sure if that is relevant. Also, the motor should keep pace with no matter how fast the user moves their head, slow, to very fast - basically replicating the users head movement. Any help appreciated. I looked at some stepper motors, but not sure. DC is fine as the other components for the project are also DC. Inexpensive motor would be great... :grin: Thanks.

 

[BradtheRad]

A servo is really the answer. It is able to move quickly or slowly. Action is firm, with no loose play. It can be sturdy enough to move the head (perhaps needing to be counterbalanced), and hold position. It always knows its center position automatically, and goes there when you send 1.5 mSec pulses.

Other than a servo, you could try motors and cams, one tilting the head up and down, the other moving it left and right. You can expect to have to gear the motor. And you'll need to sense head position somehow.

 

[chuckey]

Just twisting my 66 year old head from side to side as fast as I can. I reckon I can go from 90 degrees to -90 degrees and back in .7 of a second with any over shoot. Which in rotary terms is only 1.4 revs/sec or 84 RPM.
A servo motor from a flying model aircraft will not be fast enough, a stepper motor would be, but would be expensive. I would use a 12 V 3W or more direct drive motor with a pot on its shaft to sense its position.
Frank

 

[KlausST]

Hi,

I would use a 12 V 3W or more direct drive motor with a pot on its shaft to sense its position.

With the weight of some pounds I doubt a 3W motor has enough power to accelerate and decelerate the head fast enough.
Without any calculations I tend in the direction of 100W.
A calculation could help.

On the other hand moving of the head will not need 100% of time full power. I assume there is time to cool down.
So I think it is allowed to overpower a low power motor for a short time. How much? I have no clue.

Klaus

 

[c_mitra]

Just twisting my 66 year old head from side to side as fast as I can. I reckon I can go from 90 degrees to -90 degrees and back in .7 of a second with any over shoot.

I can move 120 deg horizontally in about the same time and about 45 deg in the vertical direction in about a sec. Hoping that the head being moved does not have a brain and will weigh only about a kg, it is possible to calculate the power needed, albeit approximately. For a trial project, an approximate value is good enough.

Taking roughly the rotational speed to be 1 rev/sec or omega about 2*pi (6.5). The moment of inertia of the brainless head is about 1000*5*5=25000. The rotational energy will be 25000*7*7 or about 0.12J

Keeping a lot of margin, a small 1-5W motor (actually two of them on gimbals) will be good enough.

 

[chuckey]

I have been thinking and to keep the inertia of the system down (no flywwheel to be accelerated or braked), I said " direct drive" - no gear box. Now I realise that if the motor is a cheap one it is only capable of stopping on a commutation point. So if its a three pole motor, it will have three segments over one revolution, so it will have a very coarse action -120, 0, + 120 degrees. I think more then 3W is required but the motor must be a multipole one. I wonder if a battery drill motor might be OK. Perhaps you could take apart a battery drill and "borrow" the motor?
In addition you will need a power amplifier with a centre tapped power supply so the ampliifier can go +- and a 10A power supply.
Frank

 

[c_mitra]

How about a stepper motor? The step count can be maintained by the driver software and each step can be small- 1.8 or 0.9 degree, good for the present purpose. Small steppers are not expensive and be easily interfaced with digital logic.

 

[KlausST]

Hi,

Small steppermotors without gear box won't have enough torque to move the head that fast.
The problem is acceleration and deceleration.
They are very critical... once the torque is too high they will stall.

I'd go for a motor with gear box.
The max used RPM will be about 60.


Klaus

 

[pityocamptes]

Re: Need some help with schematic and (code?) for RX / TX servo controller project

Unfortunately the member that was going to help me was not familiar with servos which is what other members suggested I use. Can anyone else help me with this project? Thank you.

 

[BradtheRad]

Re: Need some help with schematic and (code?) for RX / TX servo controller project

Unfortunately the member that was going to help me was not familiar with servos which is what other members suggested I use. Can anyone else help me with this project? Thank you.

(Your post was moved here because it has to do with the servo vs motor question.)

In our best hopes we expect it should not be too hard to build an interface that allows our electronics project to control how things move. In fact it appears unavoidable, that we must do a lot of work to reach success. Seems there's always more experimentation needed, or a more powerful motor, or a more robust power supply, etc.

 

[D.A.(Tony)Stewart]

HOw about a windshield wiper mechanism?

A servo needs design specs.
First design the head and define the torque to accelerate and stop the rotation and maximum shaft speed, Then a suitable pully size can be determined with motor torque and speed requirements from your specs.

This also needs optical end stops and centre sensing by counts or optical sense. A band drive goes from one roating shaft to the next with same different diameters for mechanical advantage. Stepper controllers are either Fwd/Rev pulses or pulse count and direction logic.

Tilt? that's another mechanical design spec issue.

One the mechanical specs are defined for torque and speed, the motor choice is possible.

 

[pityocamptes]

Thanks! So the consensus is that a servo will NOT support the gimbal weight and head weight? However, I from what I understand, a stepper might work, but might not spin fast enough? Plus I assume stepper motors do not have an auto center... this project is turning into a nightmare. Thanks... I am sure I will have plenty of more questions.

 

[pityocamptes]

Maybe a link to a servo, motor, stepper to help out? Not sure what I should be looking at, as mentioned previously, self centering is what we are aiming at...

 

[BradtheRad]

A windshield wiper motor (post #11) is a suitable type to consider. Easily available. Sufficient power. You might get it to run slowly enough, so that it is usable without additional gearing.

A hint about the self-centering feature. One led, 3 photosensors. Move the head toward whichever sensor is dimmer, until all sensors have equal illumination.

Are you thinking about the mechanical transfer of motion? Say, rigid sticks pushing or pulling the head along each axis. Or, a pin moving in a slot (two of them, one for X axis, another for Y axis).

- - - Updated - - -

You also need to turn off the self-centering function during the time a human is in front of the mannequin. Have you considered how to detect this? There are a few different methods. One photosensor? An array of photosensors? Motion detector? Proximity detector? Etc.

 

[c_mitra]

As both the motors are also on the gimbals, they need to be physically small. It will be a two axis gimbals and suitably balanced so that the head does not fall off. The controller needs to know both phi and psi values (current and destination) so that the motion appear smooth. It should be possible to use both relative and absolute positions. The head should be in several pieces so that it can be wrapped around the moving system.

Mechanical part of the moving system is also somewhat complicated. Such gimbals (on which you can fix a motor) are not standard and need to be fabricated.

 

[CataM]

Here is a project about head movement and the owner do not mind to share information.
https://www.edaboard.com/threads/342623/