Magnetic actuator

Martin
Posts: 744
Joined: 16 Feb 2018, 14:11
Location: Warwickshire

Magnetic actuator

Post by Martin »

This thread is to detail the magnetic actuators I'm making for my ATtiny85 'galloping ghost' project. I know that magnetic actuators aren't really galloping ghost, but the same driver circuit and decoder electronics/code works for both.

You see these sort of actuators used on lightweight indoor models - they're basically just a moving magnet inside a coil. When the coil is energized it causes the magnet to rotate, reversing the connection rotates the magnet the opposite way. The moving magnet is directly connected to the control surface, so there's only one moving part. For operation on a rudder, the magnet poles point left and right, and the coil axis is aligned with the direction of flight. I wanted a bigger more powerful version than the ones sold for indoor models.

3-D printing a bobbin on which to wind the coil is tricky because whichever way you orient the print there will be steep gradients or big overhangs - these don't print well without also printing support material and the support material is a pain to remove. So I opted to print the bobbins in two halves and glue the halves together. The bobbin then prints as two 'top hat' shapes with the 'lids' missing.

actuatorParts.jpg
Click for larger view.

The magnets I had lying around are really too small for this job: 2.03mm long by 3,02mm diameter, so I designed and printed a 'holder' that holds a central row of four such magnets, with a row either side consisting of three magnets - so ten tiny magnets working together as a large one - all the north poles point one way and all the south poles the other. The magnets are fixed to the holder with a drop of c.a. (superglue).

To wind the coil I used some old enamelled copper wire from a spool I've had for about forty years! There are no markings on the spool, but measuring the diameter with a micrometer, it's about 0.22 mm which I think makes it 34 SWG or 31 AWG. If and when I need to buy a new spool, I would get something slightly thinner to get more turns on the bobbin and produce a higher resistance.

I 3D-printed a mandrel to fit a small DC motor and hold/spin the bobbin for winding the coil - much easier and neater than winding by hand.

I designed the parts in OpenSCAD. I'll attach the files to the next post, but I expect if anyone else decides to make similar actuators they may decide to use different magnets and/or alter the diameter/length of the coil to suit.
Martin
Posts: 744
Joined: 16 Feb 2018, 14:11
Location: Warwickshire

Re: Magnetic actuator

Post by Martin »

Here's the bobbin-half in OpenSCAD
bobbin-half.png
click for larger view

Here's the OpenSCAD 'code'. You can see how easy it is to change the diameter and length to make different-sized bobbins.

Code: Select all

// bobbin for winding magnetic actuator coil 
wallThickness  = 0.5;
endPlateThickness = 0.8;
innerDia = 11;
outerDia = 16;
coilLength = 10;
$fn=50;
difference() {
    union() {
        cylinder(d=outerDia, h=endPlateThickness);
        cylinder(d=innerDia+wallThickness*2, h=endPlateThickness+coilLength/2);
    }
    translate([0, 0, -1])
        cylinder(d=innerDia, h=endPlateThickness+coilLength/2 + 2);
}
Here's the magnet holder - probably of less interest as you'll most likely be using different magnets. I would have used larger magnets myself, but I had the little ones hanging around.
magnet-holder.png
click for larger view

Code: Select all

//magnet holder for magnetic actuator.  Suits 4+3+3 magnets 3.05 dia by 2.05 long
$fn=50;
difference() {
    union() {
        hull() {
            cylinder(d=6, h=2);
            translate([5,0,0])
                cylinder(d=2, h=2);
        }        
        cylinder(d1=6, d2=8, h=1);
        translate([0,0,0.95])
            cylinder(d=8, h=3.05/2+0.5);
        translate([3,-1,0])
            cube([15,2,3.05+0.8]);
        translate([-6.5,-1,0])
            cube([3.5,2,3.05+0.8]);
    }
    translate([0, 5, 3.05/2+1])
        rotate([90,0,0]) {
            cylinder(d=3.5, h=10);
            translate([3.05,0,0])
                cylinder(d=3.5, h=10);
            translate([-3.05,0,0])
                cylinder(d=3.5, h=10);
        }
}
Lastly the mandrel, which is just a slightly tapered cone that the bobbin is a push-fit onto for use when winding the coils.
mandrel.png
click for larger view

Code: Select all

// mandrel for winding magnetic actuator coil bobbin 
bigDia = 10.7;
smallDia = 10.3;
shaftDia  = 2.5;
coilLength = 10;
endPlateThickness = 1;
$fn=50;
difference() {
        cylinder(d1=bigDia, d2=smallDia, h=coilLength+2*endPlateThickness + 2);
    translate([0, 0, -1])
        cylinder(d=shaftDia, h=coilLength+2*endPlateThickness + 4);
}
FBMinis
Posts: 55
Joined: 25 Feb 2018, 17:59

Re: Magnetic actuator

Post by FBMinis »

Martin, this kind of project is inspiring ;)
User avatar
tiptipflyer
Posts: 393
Joined: 16 Feb 2018, 22:49
Location: Germany

Re: Magnetic actuator

Post by tiptipflyer »

Thank you Martin,
I will have to check my magnets and what copper wire is laying around.
Thanks for sharing.
cheers
Frank
Tobe
Posts: 665
Joined: 16 Feb 2018, 06:19
Location: Varberg or Stockholm, Sweden

Re: Magnetic actuator

Post by Tobe »

I went thru magnetic actuator and tested different configurations some times ago and despite I got about the same effectiveness if not better than the commercial I was never happy with their power compare to how much power they needed. I still have some case and proper polarized magnets left somewhere in the attic. If made small for very light constructions they are however a very good option. In the early day of micro this was we had and I think HobbyKing still have receiver for them
CIMG0004.JPG
20141007_152444_resized.jpg
diy-mag-actuator-featured.jpg
Actuator.jpg
20141007_152444_resized.jpg
20141007_152835_resized_1.jpg
adamstwin.jpg
20141007_153358_resized.jpg
20141007_153526_resized.jpg
Last edited by Tobe on 26 Feb 2019, 11:49, edited 1 time in total.
Cheers,

Tobe
GarydNB
Posts: 282
Joined: 15 Feb 2018, 23:12

Re: Magnetic actuator

Post by GarydNB »

I have an adams baby that has no coil wire. Could someone give a quick tutorial on how to re-wind one? I have a spool of 34g enamelled wire with the intention of having a go but I'm not sure where to start....where do the ends attach? How many turns?
Etc. etc.
Thanks chaps
Gary
Tobe
Posts: 665
Joined: 16 Feb 2018, 06:19
Location: Varberg or Stockholm, Sweden

Re: Magnetic actuator

Post by Tobe »

2.4v - 2 coils of 700 turns each of #30 AWG (0.255mm) center tapped (2 x 8.5 ohms), so 1 coil of 1400 turns would do (17 ohms)

viewtopic.php?f=25&t=151
Cheers,

Tobe
Martin
Posts: 744
Joined: 16 Feb 2018, 14:11
Location: Warwickshire

Re: Magnetic actuator

Post by Martin »

Gluing the bobbin halves together with a drop of cyano. By coincidence, an AAA battery is a nice sliding fit to hold the halves in alignment while gluing. Don't use too much glue though, or you'll never get the bobbin off the battery!
gluingCoilHalves.jpg
Click for larger view.

A couple of views of the assembled bobbins and the magnets glued into their little holders. The 'handles' of the frying-pan-shaped holders will eventually be glued into the control surfaces: rudder, elevator or whatever.
magnetsGlued1.jpg
magnetsGlued2.jpg
Click for larger views.
Last edited by Martin on 27 Feb 2019, 16:20, edited 1 time in total.
Martin
Posts: 744
Joined: 16 Feb 2018, 14:11
Location: Warwickshire

Re: Magnetic actuator

Post by Martin »

Tobe, I agree that the actuators are 'thirsty' (draw more current) than an equivalently powerful motor drive does. I guess it's just physics working against us - the actuators are direct drive whereas the motors are geared down maybe 50:1 or more - the motors only have to produce a tiny torque, which is then amplified by the gearbox. The easiest way to increase torque with a direct drive is to increase the length of the magnet - but of course we are limited by what is practical to install in a small model.

It's easier for us now than it was back in the day of Adams and Ace actuators as we now have the super-powerful neodymium magnets they could only dream of. Also using a modern H-bridge driver chip means there's no need for a split or centre-tapped coil - so every turn of the coil can be used to drive the magnet in both directions.

My actuators seem to have plenty of drive while only consuming about half an amp - and for the sort of models I'm intending to use them on that's acceptable - it's less than the current a small servo takes (though the average may be more as the controls are moving all the time). It means I can drive a couple of actuators, one for rudder and one for elevator from the BEC of a typical small ESC.
belli
Posts: 39
Joined: 19 Feb 2018, 19:19

Re: Magnetic actuator

Post by belli »

So I just thunk sumting...

For a small F/F timer, DC brushed motor, could reversing the motor act as a dethermaliser? Saves a servo?
Post Reply