I'm working on designing a variable pitch propeller for a hovercraft project I will be building and I'm trying to come up with the best way to control it (mechanical, electric, hydraulic). Without getting into too much detail on the mechanical aspect, I would like your opinion on how feasible the control side is and help with designing the circuit (if I go with this control option).

The basic idea I had is to have a potentiometer mounted to the blade of the propeller to sense the blade pitch (actual pitch is irrelevant, it simply needs to match my inputs). The blade pitch would be changed by a reversible DC motor. Movement would be something in the range of 1.5 degrees per motor rotation and around 0.005 seconds per degree (though if this proves tricky for control it can be perhaps half that speed). Total blade pitch change would be approximately 90 degrees. Pitch should be controllable to within a couple of degrees (there will be 2 separate props for steering purposes but the prop output needs to be closely matched to go straight). Input would also be through a potentiometer.

Basically I need a circuit that will turn a motor in the proper direction until the "sensor" potentiometer matches the input potentiometer. How complicated/feasible is this?

Thanks,

Tim

 

Is this hovercraft a small-scale model operated by remote control or full size?

hgmjr

 

Umm, its super easy. Use a comparator. You will get a HIGH when they match.

You may be able to get away with using a decent opamp as a comparator, but I would use the real thing.

http://en.wikipedia.org/wiki/Comparator

What size model is this? Are you working on a 1:32 or bigger?

 

This will be a full-scale home-built hovercraft, 6' x 12'.

How expensive would a comparator circuit be and how reliable? Reliability is important in this case as it is the control for the vehicle.

 

Won't the centrifugal force of the rotating prop tear the pot apart?

 

Ugg... Friend, If your worrying about expense, dont do this. please.

And if you dont listen, use MIL-SPEC components. These are mission critical parts that will deal with safety of spinning blades.

There are also easier ways of doing this.

I build a hovercraft in highschool. I used the under-prop design for my model, but there is a problem with it. You will need an open top. If you dont have an open top, you cant 'suck' enough air to blow. You need to move air. My first design just sucked itself to the ground.

The final build used 8 shop vacs. I exhausted the air into the skirt, and it worked.

For the skirting, I used 6ft inner-tubes, which I was told were for tractors.

That worked well, until the extension cord ran out.

The mythbusters did it with leaf blowers.

If you are using a prop design, dont.

You have to understand and SPEND obscene amounts to ensure the prop will not explode in a mess of shrapnel from the rpms and centrifugal forces.

 

For the hovercraft part itself I have done all my research (over and over again) plus I am an industrial mechanic with a tendency for overkill when I build things. I am not worried about the mechanical, plus it will be thoroughly tested from a safe distance.

When I mentioned expense I wasn't implying I was going to totally cheap out on this. I am simply looking for the best way to do this, and total expense is part of that criteria. To be honest it will probably be mechanical control but I wanted to explore this option.

As for the pot tearing itself apart, the prop components would actually be fully supporting it. In case it wasn't clear, the pot isn't measuring the shaft rotation, but the change in pitch in the blade, which will be a much slower movement.

 

Is the pot being mounted on the propeller blades? If so, it seems that it would rotate at their speed, isn't it?

From what I know the propeller blades pitch control in big size vessels use not that kind of sensors.

With that I mean that you could try to reproduce that kind of control which, honestly, I do not actually know how is done.

Blades there could be turning at around 100 RPM.

 

If I remember correctly, it was actually around 400rpm.

As for pitch control, the way to do it, or at least the way I did it on the model, I used a part from a Mongoose bicycle. There is a device that you can mount your handlebars to that allows you to spin them around without kinking your break cables. It is basically two unions together with bearings that transfer the front and back break line cable through the device. I used one for each side of pitch adjust ring. One for increase pitch, the other for decrease.

Gear box and horse power are important. More drag, the more torque you need to keep the blades at the same RPM.

 

Hi Timplett,

Even if G forces are not an issue, I don't think a pot, where a metal wipper slides on a resistive surface, is going to be long term reliable enough for your application.

A non-contact, angular position sensor is likely more suitable for the job.

Google: angular position sensor

Regards,
Ifixit

 

You may want to look into pitch control of helicopter blades.

 

nomurphy

I am in fact essentially copying the pitch control of helicopter blades, though in a simplified manner as I do not need the pitch to change at different point of rotation. What I am looking into here is how I am going to go about linking this to the controls.

retched

If I understand what you are saying correctly this is essentially what a helicopter does, and essentially the design that I am using. The device you referred to is typically called a gyro.

ifixit

Thanks for the tip on the angular position sensor. This does seem to be a much better solution, unfortunately as it would have to be mounted within the hub itself the multiple leads work against it. Any sensor mounted within the hub will need rotary contacts to transfer the signal back to the controls. A pot would require 2 contacts, this sensor would require 8. Either that or the control circuit would have to be fully contained within the hub with only the 2 leads from the control input having to be transfered through a rotary contact.

From what I am getting from your guys' input I think the mechanical control will probably be the route I will take. The electronic way had so benefits I liked (transferring electrical energy to a rotating shaft rather than kinetic energy, which requires more bearings, wear surfaces etc., and the possibility of making it a huge R/C model ), but I think the (relative) simplicity and reliability of the mechanical method will win out in this case.

I am still open to any input as I plan to continue to fiddle with this design even though I am not likely to use it. You might give me an idea that would change my mind.

Thanks,

Tim