Pan/Tilt Head need help designing circuit

[Donny_97]

Hello everybody,

I am planning on making a motorized Pan/Tilt head to go on a camera jibcrane. My electronics knowledge isn't too well, so I'd wanted to ask all the amazing people here.

My plan was to use two 12volt DC motors, powered by a battery pack, to drive the axis, controlled by a joystick controller with speed and ramp (damping) controls, and power switch. (see picture below)
**broken link removed**
(so I do not need the tilt/pan direction and dead spot knobs.)

So basically I need the joystick left/right to control one axis moving both directions, and joystick up/down to control the other in both directions. The head should be able to move 360º, for which I think DC motors are better than servo's (right..?).

It will be wired from the controller to the head, so no need for wireless fun.

If you guys could give me a schematic and parts list for this specific build, it would be great! I'd love to have some sort of cable coming out the controller to power te motors, but that's a thing I could find out later.


Thank you guys in advance for helping me out, if you´d like more information, let me know!

Greetings,
Donny


Below are 2 pictures to give a general idea of the build I would like to make.

A jib crane with a pan-tilt head mounted at the end.
**broken link removed**
most common way of driving the axis.

 

[Inquisitive]

Hi Donny,

Welcome to ETO. It may take 24 hours to get an answer, as this is an international forum.

 

[alec_t]

The head should be able to move 360º

Indefinite angle, or would it 'return to zero' after 360°. If the former, then cables are going to get tangled unless some sort of slip-ring system is used. Presumably the elevational rotation (tilt) would be less than 360° ?
Do you want the movement to be to/through some predetermined angle? If so, that will require positional feedback. Or is it just a matter of 'left abit', 'right a bit' ?

 

[Donny_97]

Indefinite angle, or would it 'return to zero' after 360°. If the former, then cables are going to get tangled unless some sort of slip-ring system is used. Presumably the azimuth rotation would be less than 360° ?

You are right, my idea was to pull the wires through the ball bearings I will use, those are wide enough to pull some wires through. But, it won't be making 20 revolutions or something, it is more that you can get one specific camera move done, and then of course turn it back if you did those great moves.

And I will have a look into the slip-ring option, sounds great to me.

 

[Donny_97]

Indefinite angle, or would it 'return to zero' after 360°. If the former, then cables are going to get tangled unless some sort of slip-ring system is used. Presumably the elevational rotation (tilt) would be less than 360° ?
Do you want the movement to be to/through some predetermined angle? If so, that will require positional feedback. Or is it just a matter of 'left abit', 'right a bit' ?

It basically is a matter of eyeballing on the monitor where the camera should be pointed to, so yes, the 'Left a bit, right a bit" statement is fairly accurate. No need for positional feedback, since I do not need a pan tilt head that could be controlled by a computer. I would just want to eye-ball the position via the monitor I have, and control the head with the joystick controller. Hope that is possible!

 

[Joe G]

have you looked at power sideview mirrors, I don't know how heavy a camera you have and/or the capabilities of the mirror motors. But it seems just about what you are looking for. The ones I got have 2 motors that pan and tilt the mirror, by ways of a rack-and-gear, controlled by a joystick on the drivers dash, and run on 12Volts. I dont know what they came out of though.

 

[alec_t]

Good idea. Perhaps some additional reduction gearing could be used if a slower pan/tilt or greater torque is needed.

 

[Donny_97]

have you looked at power sideview mirrors, I don't know how heavy a camera you have and/or the capabilities of the mirror motors. But it seems just about what you are looking for. The ones I got have 2 motors that pan and tilt the mirror, by ways of a rack-and-gear, controlled by a joystick on the drivers dash, and run on 12Volts. I dont know what they came out of though.

Thank you for the advice, but I am afraid that won't do the job, my camera weighs about 4kg, and the mirror motors aren't powerful, fast and smooth enough I think..

 

[tomizett]

In my opinion you'd be far better of using stepper motors than DC ones - they are designed for this kind of low-speed positioning. If you're not too experienced in electronics, there are plenty of driver modules you can buy off the shelf. This is probably a job for an Arduino or something - there are experts here who'll be able to advise you about these better then I can!

I wouldn't think about continuous rotation through slip rings just yet. Remember that you'd have to route the camera signal that way too, and that will be a very high speed signal.

Do you have a plan for the mechanical side? That won't be trivial either.

 

[Donny_97]

In my opinion you'd be far better of using stepper motors than DC ones - they are designed for this kind of low-speed positioning. If you're not too experienced in electronics, there are plenty of driver modules you can buy off the shelf. This is probably a job for an Arduino or something - there are experts here who'll be able to advise you about these better then I can!

I wouldn't think about continuous rotation through slip rings just yet. Remember that you'd have to route the camera signal that way too, and that will be a very high speed signal.

Do you have a plan for the mechanical side? That won't be trivial either.

The arduino idea sounds great, but is there no analog way to do this? An Arduino is also great to program some movements maybe.. And the stepper motors are a great idea, I don't know if they are fast enough, because I'd like to do some very fast movements with them.

And the mechanical part is as good as ready, I can build the assembly out of aluminum tubing and plates. Got a lathe and mill, so that's not the problem

But could you maybe tell me someone that is an expert on this?

 

[alec_t]

I'd like to do some very fast movements with them.

How many degrees per sec, roughly? That would affect the choice of motor/gearing/circuitry.

 

[Donny_97]

How many degrees per sec, roughly? That would affect the choice of motor/gearing/circuitry.

At minimum speed about the slowest you can get, at highest speed around 150-170 degrees per second, if that is even possible. 130 degrees per second is really the minimum for the fast movements..

 

[alec_t]

So, a speed less than 0.5 revs per sec and a load of, say, 5kg. Don't know the radius of gyration, but I think that setup should be well within the capability of a powerful stepper motor, although I've no experience with those. Other members here should be able to advise.
As an alternative to a stepper motor, perhaps something like this could be used with a well-balanced mount?

 

[Donny_97]

So, a speed less than 0.5 revs per sec and a load of, say, 5kg. Don't know the radius of gyration, but I think that setup should be well within the capability of a powerful stepper motor, although I've no experience with those. Other members here should be able to advise.
As an alternative to a stepper motor, perhaps something like this could be used?

That may be a very good idea, yes! I will keep those in mind for sure.

Now I need a schematic and parts list of how to make this thing.. (for which I am not smart enough to design..)

 

[alec_t]

Well the circuit could be quite different, depending whether you go for the stepper-motor or geared DC motor option.

 

[tomizett]

A similar application is in moving head lights - the loads and speeds are in the same ballpark. These typically use large stepper motors (maybe 2 inches square) geared down via a toothed (or "timing") belt. The gearing gets you finer movement and the ability to swing the load faster (as more torque is available) and the slight spring in the belt helps remove vibration from the stepping of the motor, making it quieter than a geared system. As a side benefit, you get to route the cabling through a hole in the large driven pulley.
In this aplication some rather cleaver trajectory planning is done, to reduce stress, achieve the shortest traverse times and give a natural looking movement. I don't know the details, but there will be some maths to do if you want that kind of behaviour.
In fact, if you could lay your hands on a scrapped moving head fixture, it would be almost what you need ready made...

The arduino idea sounds great, but is there no analog way to do this?

I'm a big fan of analogue ways of doing things (I'm normally the first to say "surely you don't *really* need a micro to do that?"), but this could be a moderately challenging design for someone who is relatively new to electronics. I don't want to put you off by any means, but it just depends how much you want to learn about electronics design, and how much you just want to get the job done.

Digital and module-based approaches seem quite popular in hobby electronics at the moment. My father, who's been in electronics most of is life, recently bought some stepper driver modules from eBay and has been pretty pleased with them; you simply set a pin for direction and pulse each time you want it to step (of course, they're actually micro-stepping). This kind of module would be very easy to hook up to a micro like the Arduino - in fact, it becomes more of a wiring job than a real electronics project. But it's a possible aproach if you're more interested in the destination than the journey.

If you wanted an analogue aproach using DC motors it would be, conceptualy at least, pretty streightforward - and would certainly make a really nice project. My aproach to this would be to place a servo potentiometer on each axis to give you absolute position feedback and control the motor velocity in proportion to the error between the current position and the desired position set by a joystick or control knob. DC motors (unless very highly geared) don't hold their position like steppers, so some kind of absolute position feedback would serve to keep the camera steady while not panning. The motor would be driven from the error signal by a hefty H bridge... not efficient, but simple.
A system I encounterd recently nested another feedback loop, using a tachomiter to measure the speed of the motor. Needless to say, it worked beautifully*.

If I were you I'd have a look around at what parts you have available, and how they fit in to your mechanical and financial constraints. Also think about what power source you want to use, as this could have a bearing on what would be suitable.

 

[Donny_97]

A similar application is in moving head lights - the loads and speeds are in the same ballpark. These typically use large stepper motors (maybe 2 inches square) geared down via a toothed (or "timing") belt. The gearing gets you finer movement and the ability to swing the load faster (as more torque is available) and the slight spring in the belt helps remove vibration from the stepping of the motor, making it quieter than a geared system. As a side benefit, you get to route the cabling through a hole in the large driven pulley.
In this aplication some rather cleaver trajectory planning is done, to reduce stress, achieve the shortest traverse times and give a natural looking movement. I don't know the details, but there will be some maths to do if you want that kind of behaviour.
In fact, if you could lay your hands on a scrapped moving head fixture, it would be almost what you need ready made...


I'm a big fan of analogue ways of doing things (I'm normally the first to say "surely you don't *really* need a micro to do that?"), but this could be a moderately challenging design for someone who is relatively new to electronics. I don't want to put you off by any means, but it just depends how much you want to learn about electronics design, and how much you just want to get the job done.

Digital and module-based approaches seem quite popular in hobby electronics at the moment. My father, who's been in electronics most of is life, recently bought some stepper driver modules from eBay and has been pretty pleased with them; you simply set a pin for direction and pulse each time you want it to step (of course, they're actually micro-stepping). This kind of module would be very easy to hook up to a micro like the Arduino - in fact, it becomes more of a wiring job than a real electronics project. But it's a possible aproach if you're more interested in the destination than the journey.

If you wanted an analogue aproach using DC motors it would be, conceptualy at least, pretty streightforward - and would certainly make a really nice project. My aproach to this would be to place a servo potentiometer on each axis to give you absolute position feedback and control the motor velocity in proportion to the error between the current position and the desired position set by a joystick or control knob. DC motors (unless very highly geared) don't hold their position like steppers, so some kind of absolute position feedback would serve to keep the camera steady while not panning. The motor would be driven from the error signal by a hefty H bridge... not efficient, but simple.
A system I encounterd recently nested another feedback loop, using a tachomiter to measure the speed of the motor. Needless to say, it worked beautifully*.

If I were you I'd have a look around at what parts you have available, and how they fit in to your mechanical and financial constraints. Also think about what power source you want to use, as this could have a bearing on what would be suitable.

Thank you so much for your reply!

This really is some great advice, and your statement about the destination and the journey is beautiful, but for me, right now, since I'd like to have this build soon, it would be the destination.. (I will be learning electronics any time soon, because it is a beautiful world)

Also I think the DC motors would do a great job since I'd love to have this build analogue which isn'd possible with steppers if I am not mistaken.. Sorry totall newbie to motors.)
Anyhow, the belt driven DC motor option sounds very attractive.

If you could maybe give me the best options for the parts, that'd be great.

Maybe you can also explain me how to use a joystick in this build, the joystick doesn't need to be touch sensitive since I'd like a potentiometer as speed controller (both axis at one potentiometer).

 

[tomizett]

I'd love to have this build analogue which isn'd possible with steppers if I am not mistaken

Well... you can drive steppers, without a using a microcontroller, from an analogue control circuit, but it's going to get very complicated very quickly. Probably not something you want to get into.

Now, I'm no expert in motors either, so I'd like to ask a question to others on the thread: If I want to design an absolute position servo using a DC motor and by controlling the velocity of the motor in proportion to the error in position, what is the best way of driving the motor - voltage or current? That is, does the drive voltage or drive current correlate most linearly with motor speed? I'm thinking that the OP would probably rather avoid having to fit a tachometer on the motor, with all the extra mechanical and electrical complication that brings. In any case, voltage control will surely be easier so may be the best bet...

Now, I'd suggested a linear bridge, but I suppose PWM could be a very attractive solution given that it could be pretty low-frequency (high-kHz??). Any thoughts?

From a mechanical point-of-view, I think you'll need quite a lot of gearing with a DC motor - more than you'll comfotably achieve with pulleys. A worm-geared assembly such as suggested in #13 seems ideal - better ones are available but get expensive quickly. Likewise, propper servo potentiometers are also expensive. Perhaps someone here has experience of using ordinalry potentiometers for position sensing?
...even better, you could considder the newer technique of using a Hall effect absolute rotation sensor.

 

[tomizett]

Hmmm... A brief look round the web turned up this as the most likely looking angular Hall sensor:
**broken link removed**
It seems to be able to give an analogue output proportional to the rotation. Allegro seem to be the big makers of these things, but there was no mention of analogue outputs - just serial interfaces.

There was a thread somewhere here about using these things for position sensors on excavator arms... think it may have been TCMtech?

 

[tomizett]

Nope... Wrong!
jpanhalt was the instigator, here:
https://www.electro-tech-online.com/threads/angle-encoder-selection.143503/#post-1206917

<Edit>
Another likely part linked from the above thread:
https://ams.com/eng/Products/Magnetic-Position-Sensors/Angle-Position-On-Axis/AS5162
</edit>