HobbyKing.com New Products Flash Sale
Closed Thread
Thread Tools
Old Apr 03, 2010, 08:43 PM
Registered User
LukeZ's Avatar
Joined May 2001
2,952 Posts
Question
Unconventional Use for Magnetic Actuator - Will this work?

Hi Guys,

Please forgive my ignorance if this is self-evident. I think what I'm proposing should work but just wanted to run it by people who actually know, before I buy a twenty dollar actuator.

I have an application where I need to open/close a valve. I don't really need proportional control, so open/closed or on/off is fine. Ideally I would also be able to pulse the valve open and closed very quickly, much more quickly than a servo could, otherwise a servo would be much easier.

A solenoid control immediately comes to mind. The problems with this approach are:
1. It is hard to find solenoids small enough (this is a very small valve).
2. Solenoids are designed to be operated intermittently. This is fine for when I need a pulsating action, but when for example I need to hold the valve open for a long period of time, the solenoid coil is going to draw a lot of current, since they are typically low resistance.

So my next thought is to use a magnetic actuator of the type you guys use on your micro planes. They are readily available and are quite small enough. I presume the actuator speed can be quite rapid, much more than a geared servo, so my fluttering requirement should be met (I will be controlling these with a PIC so I should be able to oscillate at any practical speed possible).

But my question is, are these meant only to be operated intermittently or will there be any harm in having the actuator powered completely to one direction or the other indefinitely?

Looking at the Plantraco Mini-Act the resistance is 50 ohms. I don't know if this is generally representative of all similar actuators, or if there is any sort of standardization of resistance. But it seems to me at 50 ohms if I keep the voltage low enough it should be feasible to keep it continuously energized without too much current draw.

Also, let me ask this: in the "Off" or un-powered state, is an actuator's position "centered" or completely to one side?

A solenoid for example, when powered, has the plunger completely to one end - and when disconnected from power, the plunger moves via spring to the complete other end. But I'm guessing this is not the way your actuators work? They must have a power-off position of center, and then voltage in one direction causes them to move one way, and voltage the other direction the other way? This would be quite different from a solenoid which can only accept current in one direction.

I've posted a photo of a mock-up valve below. The tube is 1/2" diameter. There is really no friction in opening or closing the valve since it is balanced, therefore I'm sure the torque of a magnetic actuator should be sufficient.

The purpose of this valve is to create realistic smoke effects in very small models (tanks, small boats). My actual final valve will probably be even smaller.


Luke
LukeZ is offline Find More Posts by LukeZ
Sign up now
to remove ads between posts
Old Apr 03, 2010, 08:53 PM
I don't want to "Switch Now"
pmackenzie's Avatar
Toronto (Don Mills), Canada
Joined Dec 2002
17,088 Posts
Actuators have no centring force of their own. You either use some sort of "spring" at the hinge or add magnets near the coils. Both take away from the force available.

I wonder if in your case a voice coil type of drive would not work better, with an over-centre spring to hold it in both positions. So a short blip to the coil would make it open or close.
I assume weight would not be as big an issue for a boat.

Pat MacKenzie
pmackenzie is offline Find More Posts by pmackenzie
Old Apr 04, 2010, 12:03 AM
Registered User
LukeZ's Avatar
Joined May 2001
2,952 Posts
Pat the voice coil is a very interesting idea which I would not have thought of myself. If they do not easily come much smaller than what can be scavenged from a hard-drive, they will be much too large for a tank, which is my primary interest. The most common scale is 1/16th and the space for the smoke unit is only a couple square inches total. But I will keep doing some more reading on them.

So if the actuators have no centering, that is ok for my purposes. Is there anything other than magnetism that physically holds the magnet within the center of the coil, even when not energized?

What could potentially be even better than the linkage and moving arm I am envisioning now, is to have the smoke stack be made of something non-conductive, like styrene. The circular smokestack would actually become the bobbin for the coil, and the valve inside actually be the magnet. But I don't know how well these magnets "float" within their coil and I suspect the devil would be in the details - avoiding binding or whatever. Also, although I have seen wide, thin circular magnets, I don't know if those are the sorts that work well in a coil actuator.

Luke
LukeZ is offline Find More Posts by LukeZ
Old Apr 04, 2010, 02:04 AM
Registered User
FLY...Please's Avatar
Dayton, Ohio
Joined Jun 2005
177 Posts
Luke,
May I ask why you need it to work so fast? Servos are pretty fast (1/5 second). 1/5 of a second should be at least as fast as the fluid time-scales. In other words, it is going to take more than 1/5 of a second for the smoke to respond to an open valve. For example, if you were to cycle the valve open/shut at a frequency of 10Hz, you would not see disperate smoke clouds, just a continuous line of smoke.
- John
FLY...Please is offline Find More Posts by FLY...Please
Old Apr 04, 2010, 02:19 AM
Registered User
LukeZ's Avatar
Joined May 2001
2,952 Posts
Actually the smoke is driven through the stack by a fan, so its movement speed is higher than might be expected. The purpose of having it oscillate is to make the smoke appear to pulse in time with the increasing speed of the "engine" (actually electric motors in our models). The most common version of this effect is created through the use of a diaphragm pump, which serves the dual purpose of both moving the smoke and creating the pulsating effect (the pump speed is proportional to the speed of the drive motors).

However these diaphragm pumps are noisy and undependable over the long term. Instead people have been turning to the small 1" diameter computer fans for motive power - which works very well, but at the expense of any pulsating effect.

This effect may be subtle but with land-based models where weight is not the consideration it is with aircraft, subtle details are often gone after to great lengths to achieve. Some RC tanks are indeed not much more than very complicated static models that are probably rarely driven, but used more for show purposes.

My thought was that an actuator can be very, very small, should be quiet other than for whatever noise the linkage mechanism makes (the arm between the actuator and the valve control horn), and able to be controlled as rapidly or slowly as we like. And if the "magnet as valve" idea would work then it would indeed be absolutely noiseless since there would be no parts moving against each other.

I guess I just need to order an actuator and do some testing. If there were something that made this clearly impossible (as opposed to perhaps just a waste of time), someone surely would have told me.

Luke
LukeZ is offline Find More Posts by LukeZ
Old Apr 04, 2010, 02:36 AM
Registered User
LukeZ's Avatar
Joined May 2001
2,952 Posts
Here are two videos which show the various effects. This first one is using the stock diaphragm pump to move the smoke. You can clearly see the pump doing its thing to the left. Although in this clip the engine speed was never taken above idle, you can see the speed of the pump is already faster than any servo would be able to oscillate. Notice how the smoke coming out the exhaust has a very nice pulsating effect.

Heng Long smoke generator modded (0 min 17 sec)


Now here is a video of a modified unit using a computer fan. The smoke output increases with engine speed but it comes out like a cigarette stream, which is not quite so realistic:

Modified mini H.L. fully proportional smoker test.AVI (2 min 24 sec)


Hopefully with the actuator one could achieve the reliability and quietness of the computer fan, but also have a pulsating effect.

Luke
LukeZ is offline Find More Posts by LukeZ
Old Apr 04, 2010, 07:46 AM
If it flies - I want one!
Petem's Avatar
Werribee, Victoria, Australia
Joined Jul 2008
1,077 Posts
Rotate it?

Luke,
What effect would you get if you simply rotated the valve continously at the desired speed?
The result should be the valve opening and shutting rapidly at twice the rpm, and then you could drive it with a geared or direct drive motor.
Sounds like an interesting project.
Cheers,
Peter
Petem is online now Find More Posts by Petem
Old Apr 04, 2010, 02:15 PM
Registered User
LukeZ's Avatar
Joined May 2001
2,952 Posts
Peter, you're right, that would indeed be the best way to achieve the pulsating effect. I considered it. However, I also desire the ability to do full open and full close sustained, and the mechanism starts to become rather complex if you want to achieve that with a motor.

For example, let's say we are doing engine startup. We want a realistic big puff as if the engine had been sitting for some time and gotten oil collected in the cylinders. Before the sounds of the engine come from the control board, we will heat the resistor for several seconds, but keep the smoke valve completely shut to allow the smoke to accumulate. When we hear the sounds of the engine begin to turn over the valve and fan may allow a few small puffs of smoke to be released, but right when the engine catches the fan will spin up to full speed and the valve open completely for a short span, allowing a nice big belch. Then immediately it will start to flutter at an increasing speed.

The valve control, temperature of the heating element, and speed of the fan will all be controlled by a microchip, so the timing of these things can become as complex as we like. I think some realistic scenarios might be obtained, such as backfiring effects, engine coasting, etc...

That is the idea anyway.

Luke
LukeZ is offline Find More Posts by LukeZ
Old Apr 04, 2010, 04:11 PM
TjW
Registered User
USA, CA, Riverside
Joined Mar 2006
220 Posts
Quote:
Originally Posted by LukeZ View Post
Peter, you're right, that would indeed be the best way to achieve the pulsating effect. I considered it. However, I also desire the ability to do full open and full close sustained, and the mechanism starts to become rather complex if you want to achieve that with a motor.

For example, let's say we are doing engine startup. We want a realistic big puff as if the engine had been sitting for some time and gotten oil collected in the cylinders. Before the sounds of the engine come from the control board, we will heat the resistor for several seconds, but keep the smoke valve completely shut to allow the smoke to accumulate. When we hear the sounds of the engine begin to turn over the valve and fan may allow a few small puffs of smoke to be released, but right when the engine catches the fan will spin up to full speed and the valve open completely for a short span, allowing a nice big belch. Then immediately it will start to flutter at an increasing speed.

The valve control, temperature of the heating element, and speed of the fan will all be controlled by a microchip, so the timing of these things can become as complex as we like. I think some realistic scenarios might be obtained, such as backfiring effects, engine coasting, etc...

That is the idea anyway.

Luke
Why not use a stepper motor? An index -- say, a hall effect switch -- for valve fully closed, and you're ready to go.
The controller would know where it is from the last time it saw the index pulse.
TjW is offline Find More Posts by TjW
Old Apr 04, 2010, 04:35 PM
Registered User
LukeZ's Avatar
Joined May 2001
2,952 Posts
TJ, that again is another good idea in principle - the problem is sourcing a stepper small enough. Do a search on the internet and you will have a very hard time finding anything near as small as these actuators. In fact I was not able to.

The only thing I did see which might work is this video on YouTube. The guy said he picked this up in the "electronics district in Bangkok." Not very promising for being able to buy them easily. But it certainly would do the job quite well.

Mini stepper motor in action (0 min 16 sec)
LukeZ is offline Find More Posts by LukeZ
Old Apr 04, 2010, 04:37 PM
Registered User
LukeZ's Avatar
Joined May 2001
2,952 Posts
I just ordered one

I just went ahead and ordered the MCA3-A actuator from Air Midi Micros.

Now I need to make up a circuit for my PIC to control it. There appears to be lots of information in Japanese on this process, but not very many details of the code.

Luke

EDIT: Or I guess I could use the trick of using a servo board and swapping the actuator for the motor. I can easily control a servo from my PIC having already programmed this code before. This would leave the PWM out on my PIC free to control the fan.
LukeZ is offline Find More Posts by LukeZ
Last edited by LukeZ; Apr 04, 2010 at 04:41 PM. Reason: Had another thought.
Old Apr 09, 2010, 07:37 PM
Registered User
LukeZ's Avatar
Joined May 2001
2,952 Posts
Ok guys, I got my actuator from Air Midi, nice little piece of work.

I'm trying to test proportional control using the PCB from an old servo. You can see in the video below that it is not anything like what I would call proportional. Full deflection works great, but it has no real range in-between. I've tried all the tricks I've been reading about like setting the travel down on the TX, etc... but this is about the best I can get.

Is there something about this actuator that won't permit proportional control? It seems to me all actuators should have the same design, which is to say, just a coil of wire. This one did have a centering magnet which I removed since I didn't need it, but it made no difference before or after.

I also tried the circuit boards from several different servos just in case that was the problem. Made no difference. I should also say I'm running this at basically 5v (4AA). I know that's a bit more than usual but I don't think it should affect what I'm testing here.

I'm thinking I might try to whip up a 555 PWM circuit next and see if that makes any difference, perhaps the update frequency on the servo board is too coarse?

Here is a video of my actuator:

Proportional Actuator (0 min 28 sec)




Nothing at all like this other nice video I found on YouTube:

X-Twin Acrobat / Air Hogs / AeroAce proportional actuator (1 min 12 sec)
LukeZ is offline Find More Posts by LukeZ
Old Apr 09, 2010, 08:51 PM
Registered User
RC_AV8R_181's Avatar
state college, pa
Joined Aug 2002
326 Posts
Luke-
Assuming everything is pretty much frictionless and not binding, you might consider holding the whole setup upside down and trying it like that. It looks like you have quite a bit of mass on the axle above the pivot point, which could be causing problems.
Also, the actuator movement won't necessarily be proportional since there isn't any feedback, but the force produced by the actuator should be proportional. That essentially means you need a little resistance in the system (often rubber band hinges in control surfaces). Just sitting on the bench disconnected from the valve system, does the actuator behave proportionally?
Ben
RC_AV8R_181 is offline Find More Posts by RC_AV8R_181
Old Apr 09, 2010, 09:38 PM
I love my ID.
Fairfax VA
Joined Aug 2004
670 Posts
Quote:
Originally Posted by RC_AV8R_181 View Post
... you need a little resistance in the system ...
I agree with AV8R. I think you'll need a centering magnet back on your actuator. The magnet will work as resistance. Try installing stronger centering magnet.

By the way, I have done some RC tank modeling and PIC programming before I became addicted to Micro RC airplanes. Mostly 1:35 scale conversions. Here is a link to a thread
jingjingjing is offline Find More Posts by jingjingjing
Old Apr 09, 2010, 09:40 PM
Registered User
LukeZ's Avatar
Joined May 2001
2,952 Posts
Ben, thanks for the help. You are right, the thing really needs some kind of resistance, so I've reattached the magnet.

Just disconnected from the valve, the thing still oscillates all over the place - no proportional control, it's either oscillating, or completely over to one extreme or the other at full control throw.

However! I wired up a simple 555 PWM circuit. I started off with a PWM frequency of 144hz since that's what the schematic called for, but I got the same effect as with the servo PCB. Next I upped the frequency tenfold to 1,440hz - now this thing is humming! At the higher frequency I get the very high-pitched whine that the RobsHangar video has, and very, very smooth proportional control (as long as I use the magnet).

RobsHangar was using a PCB from a Walkera servo, you'd think it would be a cheap one, whereas I was using a HiTec board, but I guess his must have had a higher PWM frequency all the same. Interesting the differences in servo manufacturers.

I may not need proportional control but I wanted to know how it worked and how to accomplish it - and also, since my results didn't match the other guy's, if I was doing something wrong. This has been a great learning experience!

Now I need to rig up a better linkage and start programming a PIC. Future updates about the smoker specifically are going to be in this thread, if anyone is interested. If I have any more actuator-specific questions I'll post them here.

Thanks everyone for your help,


Luke
LukeZ is offline Find More Posts by LukeZ
Closed Thread


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Question Attaching sleevel to shaft with locktite: will this work? outsider787 The Builders Workshop 4 Mar 29, 2010 07:28 PM
Discussion Will this charger work? chopper320 Micro Ready-to-Fly 4 Mar 24, 2010 02:03 PM
Question Will this work on an Esky Big Lama? cfanto Micro Helis 7 Mar 14, 2010 11:06 AM
Question Will this HiTEC combo work? kelles Radios 3 Mar 08, 2010 07:37 PM
Question will this main gear work in belt cp v2 450 size heli mahmoodmusaddi Electric Heli Talk 2 Mar 03, 2010 01:48 PM