Welcome! to the South Leicester Aeronutz web site.
Muscle Wire Actuators.
An actuator is a servo but has no feed back to
show the output arm’s position.
Flexinol (Muscle Wire) is
made in California by Dynalloy. It is
a nickel-titanium alloy (nitinol)
When stretched about 4 per cent and then warm
it up, it shrinks back to its original size, with
considerable force.
One way of warming it up is to pass a current
through it, which is what we will be doing here.
Flexinol looks like a shiny, very thin metal wire.
It comes in a variety of diameters. It is
supplied on a card with instruction on use. There is also a
book on suggested uses. It is general said that the
instructions on the packet are very conservative and you
should roughly double the current. The book is OK but it is
90 per cent irrelevant for our use.
At the time of writing the following actuators are in use for controlling rudders and elevators, althought Flexinol could be used for many ohter uses as well
A) BBA. Bob Bailey combined rudder and elevator actuator - see below
B) Matt's single actuator - see below
C) Graham Stanley's, rotating tube, ortho band return, double actuator
| D) Mark's
APPA single actuator magazine article. Feb 2004 |
 |
thanks to Ken and R/C Model Flyer for letting us include it here |
Latest News. mid Dec. 03 Matt has done some tests which showed Bob's actuator moves non proportionally. Matt has written software that corrects this by telling the Rx to give the wire a little more power in the mid range of the movement. Don't panic ! The actuator is Ok without the software, you could probably programe the Tx to do this?
Bob Bailey (Nov 2003) in Florida has been playing with the 025LT . He has made an actuator that uses brass springs to centralise the rudder better. He sent one to Matt who has tested it, looks to be good. It is a remote actuator with a chassis and combines both rudder and elevator. Matt says its really good Matt has sent us a nice little MPEG of the BBA in action. Conditions used for the vid : JMP Combo receiver, with new custom software, not using diodes, 3.5V power to the JMP, trim set to neutral 'off' for the BBA. Tx is a super cheap Futaba, no special modes.
Vid clip of Matt's test plane for his BAA actuator
Bob described his actuator on the RCGroups chat site. Bob is happy for us to reporduce it here, the full discussion starts here
BBA Actuator comentary by Bob live in Florida ....... take it away Bob
Pictured below is a .1g muscle wire dual actuator, which 30 degrees of throw each side of center. It is self-centering, with good centering accuracy and force. The reaction/return speed is good. Center/full throw/center less than .5 seconds. Current draw is approx. 50 ma necessitating an amplifier to drive it. Materials used in it's construction include a .4 mm double-sided, copper-clad board, a .25 mm brass spring rod, .1 mm brass sheet and .001 muscle wire. The difficulty factor in building the first one is definitely a 10, but after you learn the tricks it is on par with building a .25 g magnetic actuator .
The springs provides a force for centering and return. The off-center pivot, in relation to the muscle wire attachment points, provides a center equilibrium. When power is supplied to one side of the actuator, the muscle wire contracts, the control horn rotates, and the spring stretches. When power is removed, the muscle wire relaxes, the springs contracts, and the control horn returns to neutral.
The main body of the actuator is made of .04mm copper-clad board. Thickness of material is not critical for function, only weight. No dimensions on this part are critical, just so they're similar. Thicker material can be cut narrower on the long part to save weight. This part of my actuator weighed .07g. I machined mine with a Dremel tool and a cut-off wheel, and deburred with a file. Insulation grooves can be cut, filed, milled or etched. Another option is to make a 2 piece part, more weight-less work. Long part can be made of copper-clad or small brass tube/rod. What I foresee for the future is this part being an extension of the circuit board on the end of one of Koichi's IR receivers with a 90ma battery magnetically attached to the board. Weight of unit-.5 to .7g. Basically a self-contained drop-in unit, no wires except for the 2 running to the motor.
Muscle
wire itself has no centering, so I added a spring into the
system to provide a centering force. This force can be varied
by tension pulled on the spring. However, this led to a
second problem. The two wires pulling on one of the control
horn pivot produced a lot of side load friction. The harder
you pulled, the more friction you produced. On a pivot the
sized of the average needle, I had 10 degree dead spot for
center. My current solution is to cut down the size of the
axle to the smallest size practical, and a semi loose fit on
the control horn to cut down on the number of atoms rubbing
each other. My current dead spot is one - two degrees. In
future experiments, I plan to try a pivot similar to that on
a balance beam scale. Basically, a narrow point going into a
wide groove, so control horn teeter-totters instead of
rotating. Very close to zero friction, but not sure how it
will transfer current. In designing this, I went through
three feet of muscle wire, two and a half inches at a time.
Made prototype with last five inches.
Throw-30 degrees each side of center. Both actuator and
control surface horns have four holes each, so maximum
control surface throw can be as much as forty-five degrees,
or as little as fifteen degrees. Adjustable at five degree
increments to suit model.
The brass
rod I used came from KS Metals (that rack of brass shapes in
almost every hobby store). The o.d. of the rod is .01 inches/.25mm
This rod serves two purposes: conduct current to muscle wire/act
as spring. The brass needs to have some resilience. You can
use larger rod, but it will need to have a longer overall
length to keep the same pull, or you could acid etch it to
the proper size.
Fig. 1: Flatten 1/8" (3mm) of end of rod with smooth jaw
pliers. Heat flat part of rod with flame to anneal the metal
so it can bend without breaking. Scrape flat part of rod
clean with razor blade to insure good electrical contact.
Fig. 2: Bend rod in middle of flat end a little past 90
degrees
Fig. 3 Clamp end of muscle wire in a pair of hemostats.
Bluetack hemostats down to work bench. Wrap mw around 5 times.
Fig. 4: Wind rod toward hemostats until rod is about 3mm from
tip. Tack down loose end of mw by bluetacking magnet to
workbench. Lay mw on top and place another magnet on top of
that.
Fig. 5: Push mw wraps down rod and into corner of bend. Use
tweezers to push if necessary. Tease out any looseness in
wraps by tugging on wire. Bend rod all the way down and
squish with smooth jaw pliers.solder joint to seal conection
Hover a hot soldering iron 1/8" above the first inch of
mw to keep it from curling when removed from bench.
Fig. 6: Cut off brass rod to 1/2"(13mm) overall length.
Repeat Fig. 1 and Fig. 2 on remander of brass rod.
Fig. 7 Bluetack first to workbench. Wrap mw 5 times around
new rod. Wind rod to achieve a 2 1/4" (6mm) distance
between rods. Push wire into corner. Crimp, tin, clip off
wire and cut rod to 1/2"
Repeat for second actuator. Make mw length as similar as
possible.
Fig. 1:
Bend brass rods 180 degrees around a 1/16"(1.5mm) radius.
Bend should be opposite of the side mw exits. I took a cheap
pair of pliers and modified them with Dremel to get the bend
I wanted. I then practiced with 1/2" pieces of rod
before committing to muscle wire rod. Nothing is really
critical here. It just needs to be bent and look good.
Fig. 2: Bluetack board to workbench. Bluetack brass rods and
lead wires for a 50ma draw onto board as shown in drawing. Be
sure rods are plumb, flat, level, square, parallel, etc.
Solder rods and lead wires to board.
Fig. 3: Turn board over and bluetack down to workbench. Place
scrap pieces of 4mm PCB on top of brass rods. (These are
temporary shims. Do not solder.) Position second set of brass
rods over first set. Bluetack down. Position leads and 100ma
positive lead, solder. Remove from board and look at it. If
anything is wrong, now is the time to fix it.
Fig. 4: Cut 2 1/16"(1.5mm) square piece of .4mm PCB.
Insert between brass rod mw connections and solder. Turn over.
Solder other side.
As it
turns out, the brass I had came from Special Shapes,
a division of KS Metals. My hobby store stocked them both in
the same rack. If you go to the KS Metals site, and look at
the cut-to-size section, the rod will show up there in the
size menu.
The simplest solution for your problem is to buy the 020
brass rod, go to Radio Shack, and buy some ferric chloride
etching solution. You will need this to make the control horn
anyway. Take the lid off of the solution, and put the rod in
the bottle. Check approx. every 15 minutes, until the lower
half of the rod looks half the diameter of the top half.
Remove from acid and clean with water, then take 600grit sand
paper, and polish. I did this just a few minutes ago and the
resilience is indistinguishable from 010.
The output arm itself will be available on Gordon Johnson's
website for download directly onto glossy photo paper.
Transfer onto brass sheet and acid etch.
The
control horns are etched from .005 brass sheet from KS Metals.
The etching is a fairly easy process, very useful and well
worth learning. A file will be posted on Gordon Johnson's
website. Download this ono your computer. Print this out
using a laser jet printer, on glossy photo paper. When
printed, you will have a sheet with over 500 sets of control
horns on it. The next step is to transfer this print onto
brass. The laser jet printer uses toner, which is basically a
vinyl plastic dust melted onto the paper. You are going to
remelt this onto the brass. Leave brass in bag to avoid
finger prints. With a pair of scissors, cut a 1 1/2"(75mm)
strip, bag and all, off the bottom of the brass sheet. Slide
this strip out of the bag onto a flat non-heat conducting
surface such as glass, tile, marble etc. Cut a slightly wider
strip about 1" longer off of the control horn page.
Center face-down on brass. Pull tight and tape down ends.
Next, apply heat and pressure to back of paper. You can use a
clothes iron, covering iron, or the side of a soldering iron.
I use a soldering iron with a flat brass fitting, which I
bought in the craft section of Wal-Mart. I use enough heat to
turn the back of the paper a light to medium brown. Next,
place brass and paper in a bowl of water for 5 minutes. Paper
will lift off brass, leaving the image. Remove brass from
water, and inspect image. If spots are missing, not enough
heat and pressure were used. If the small adjustment holes
are missing, too much heat and pressure were used. Cut off a
new piece of brass or clean this one with Scotch Brite or
steel wool. When you get the technique down, and have a piece
with enough good images to give it a try, take a piece of
wide clear tape and place on the back of brass with as few
bubbles as possible. Leave one end of tape long enough to use
as removal handle. Pour ferric chloride acid in plastic bowl.
Immerse brass, leaving tape hanging out of bowl. Check every
half hour. When all unprotected brass is gone, remove and
wash with water. Lay tape on flat surface, and scrape off
black with razor blade. Remove good ones from tape.
Pictured below is an enlargement of the horn you will be
etching. The one on the left is the one for the actuator. The
part sticking off to the side is folded over 180 degrees, and
the muscle wire fits into this fold.
The spring on this actuator is the limiting factor on the forces that can be put into it, and the forces that can be gotten out of it. This is good and this is bad. If you bump the tail, the spring simply yields to the force. The wire does not break. 90% of all tail strike damage will occur at the control horn-tail surface glue joint, just like all models. Now, the bad news. If you try to get a force out of the actuator, stronger than the force of the spring, the spring will give, and you won't get it. But this is still a fairly strong force. I believe that if you take a push-rod and connect one end to an average magnetic actuator, and the other to this mw actuator, the mw will be the boss. I think the upper weight/size limit of models flown with this type of actuator will be what will fly with a 145 lipo battery. Fatigue of the mw at the control horn joint is probably the Achilles heel of this design. A counted cycle test and a redesign of the horn joint, if necessary, would be good so we can know where we're at. Professional acid-etching of horns is a good idea, but I think the design needs a little tweaking before investing in that. Looking at Matt's photos, I think the horn should be enlarged by 10%, but I won't have time to try this. Acid etching the spring is an idea I considered, but attaching mw to spring is not easy. It is much, much easier to attach mw to long brass rod and then cut and form spring, than the other way round. mw is difficult to work with. If you've worked with thin copper wire on magnetic actuators and think it's like that, I can tell you that it's not. However, after you learn a few tricks, you can beat it. One nice thing about this design is, if you can put it together, and get a little tension on the spring and connect it up, it works right off the bat. It's obvious whether it needs more or less tension for return time and centering. The others I've built, you hook it up, it does nothing. You play with it some it does something, but not much. You play with it some more, and it does nothing again. And you don't know for sure what's wrong. Very frustrating!
finishing horns
Fig. 1:
Take a standard sewing needle and cut it in half. Chuck the
sharp end in an Exacto handle so that 1/4" of needle
sticks out. Place on a spinning Dremel cutoff wheel and
rotate, as shown. This is not intended to make the needle
sharper, but to scratch the side of the needle, and turn it
into a micro-file. (A little sharper is good, though.) Use
this to finish cutting the acid-marked holes through brass,
and enlarge to proper size. Practice on reject horns first.
Place the horn on a plastic surface (a bottle cap will do
fine). Insert the micro-file as shown in fig. 3. Apply a
little pressure and rotate. The center holes are enlarged to
the size of the .010 brass rod. The horn should rock side to
side 5 degrees to get the proper looseness. The adjustment
holes need to be the size of the center strand of 1/16 7x7
stainless steel aircraft cable, available at most hardware
and boat stores. Cut off short length, unravel and keep
center strand. Weedless fishing hooks also have a similar
piece of stainless wire.
Fig. 2: You'll need to make a tool to thread the horn. It is
shown here so that I don't have to make another drawing. Cut
a 2" by about 1/8" strip of brass. Cut a taper on
the end so the end is the same width as the horn extremity.
Smooth the edges. Next, take the micro-file, and round the
edges of the horn indicated by arrows. Round the inside of
the box, also. Finish the remaining edges of the horn with a
jeweler's file to your standards.
Fig. 3: Pin down one end of the horn and sand clean the other
end of the horn with 600 grit. Rotate and repeat. Flip and
repeat.
Fig. 4: Insert .010 brass rod through both mw and over the
side of the actuator body with positive lead. Check to see if
rod is 90 degrees to actuator body when both wires come
tight, but springs don't move. If so, scratch a mark on
actuator body just above the brass rod.
Fig. 5: If rod is not 90 degrees, then the wires are not the
same length, or the springs are not bent the same. This can
be adjusted by spreading open or squeezing together the
appropriate spring. If you don't do this, the actuators will
not operate the same. Scratch the mark in fig. 4 when done.
threading horn
I spent a
little extra time making these drawing. I hope their meaning
is clear. If you're left-handed, do everything opposite and
it will work fine.
Fig. 1: Bend horn extemity a little past 90 degrees. Tape
horn down to work bench. Bluetack actuator next to it. If mw
is curled, apply radiant heat with a soldering iron. Place or
weight down mw, as shown. Slide a loop of fine copper wire
over the mw and through the square hole in horn. A single
strand of multi-strand copper will do.
Fig. 2: Pull mw through square hole and pull snug so that you
can clearly see the routeing of the wire. Slide horn
threading tool under right side and over left side of mw loop.
Rotate tool toward horn.
Fig. 3: Slide tool into bend on horn. Remove Bluetack and
slide actuator away from horn. Check to see if routeing looks
like fig. 3. Remove copper wire. Pull snug, and bend
extremity down nearly 180 degrees.
Fig. 4: Place actuator on horn 90 degrees to one another.
Pull fairly snug.
Fig. 5: Double check for the tiny loops just to be sure
routeing is right. Pull wires tight. Check fore equal tension
and that horn is square to actuator. Squish extremity flat.
Check to see if brass rod goes through both holes and turns
freely. If not, align with micro-file.
Fig. 6: Bluetack .010 brass rod on actuator and solder. If
you get solder on horn part of rod, pull off the rod, and try
again.
Fig. 7: Clip rod 1/8"(3mm) from actuator body. File ends
so the horn goes on easy. Hold horns square to actuator. Pull
to, and slide over rod. If horns do not come out parallel to
one another, or square to the body, relieve some tension on
spring, and try teasing into position. If that doesn't work,
remove horn, open fold-over a little with Exacto blade, and
try again. (It can be a little off alignment without too much
harm.) Put a slight bend in the rod to retain horns. Clip off
excess rod. Ta da! Actuator is done.
A few well-deserved acknowledgements:
To Matt Keenon for a starting place, basic design and
invaluable info on mw. Also for sending Bob Selman a roll of
mw to sell, without which I wouldn't have gotten far.
To ulR Team for cool designs that sparked my imagination.
To Aeronutz, my favorite, whose exemplary website gave me the
inspiration and desire to build tiny rc's, and, more
importantly, for showing me that it could be done.
mw connections
This is a
list of mw connections that I have experimented with and a 1
to 10 rating to help those who want to experiment. I only
recommend two of them. There is also a drawing of a work
board that makes it easier to manage mw.
Fig. 1: work board - First of all, forget the hemostats in
the posting, magnets are better. You will need 4 magnets, 6"
of .3mm carbon rod and a 6" x 8", or larger, white
work board. A small Formica shelf from a hardware store works
well. Glue 1 magnet to the front left corner of the board.
Glue another magnet to a small square of paper, balsa,
plastic or whatever. Glue the end of the carbon rod to the
square. Bluetack the other end of the rod to the board so
that it can be moved around to suit your needs. Place the
very end of the mw on the first magnet. Place another on top.
Run mw over 2nd magnet, pre-load rod to tension mw, and place
magnet on top.
Fig. 2: stranded connection - This one is the current 10, in
my opinion. Strip 1/4" of insulation off of a small
guage stranded wire. Scrape the metal clean with a razor
blade. This is necessary for good contact. Don't skip it.
Divide wires into two bundles to form a fork. Slide mw 1/8"
into fork. Rotate connection wire 3 times to insure good
contact. mw should pull wires into a tight bundle. If not,
tug on mw until tight.
Fig. 3: Fold over wires at mw intersection. Crimp wires with
smooth jaw pliers. Solder to seal the joint. Repeat at 2 1/4"/60mm
spacing for other end.
Fig. 4: This can be adapted to design by extending small
spring connecting board, and soldering connection to it.
Adjust actuator to proper function, then clip off unwanted
wire. The advantages of this method are - fast, easy
connection, no special tools or materials, easily adjustable
by melting solder so you may reposition it on board and allow
to cool. The wire provides a nice handle to work with. When
everything is right, you cut it off. The major drawback is
two solder conections in close proximity. When you melt one,
the other melts, too. Best solution is to bury it in
Bluetack, then make the connection. Bluetack removes melted
Bluetack. Main body of actuator will lengthen to compensate
for longer spring connector.
Solid connection: This connection is described in the
original construction text. I would reate this as an 8, in
comparison. The main advantage is a smaller package. Main
disadvantage - harder to replace broken wire.
Crimp connections: I would rate this a 5. I tried making a
swaged loop with short pieces of hypodermic needles and a
swaged crimp made from 1/16" copper tube pulled through
a jewelers' drawbar until it was teeny-tiny. I then cut off
short piece and crimped on mw. The major drawbacks are size
and crimp pressure. You are trying to put two very small
pieces together, hold everything together in the right place,
and then crimp it. When you're done, you give it a small
pull, and one swage slides off because you didn't crimp it
hard enough. However, if you can buy them pre-swaged, this is
probably the easiest way to go.
Screw connection: I would rate this a 2 in comparison.
Logically, this seems like the best solution. I pursued it
first, and I think everyone else did, too. In reality, to get
this nasy little wire to wrap around a tiny little screw, and
tighten in the right spot is not easy. If you wrap the wire
clockwise, it sucks it in. Counter clockwise spits it out. It
does make the most reliable electrical connection, but I
still don't recommend it.
Straight solder: I rate this one a 0. I stretched mw across 2
flat pieces of brass, with 60mm between them, and attempted
to solder them directly to the brass. The titanium rejects
the solder the same way wax rejects water, it just beads up.
I managed to blob solder over the wire. I checked connections
with ohmsmeter, and both showed no connection. I planned on
trying to gold plate the mw at the connection, but the
Robostore said they tried this with poor results.
This is a photo of my infra-red muscle wire combo. All up weight should be .6 grams when completed. Also shown is an experimental muscle wire control horn. It has strain relief for the muscle wire, zero friction pivot and built in stops.
Graham Stanley's rotating tube with ortho band rubber springs design, used on his vee tail test rig. Worked much better than the plane!
| Here is Mark's APPA Mk2 deisgned for low current consumption. The barn door rudder is supposed to push the actuator back to central, its not yet been tested in flight. |
Matt's MW
articles from RC Micro flight One and two
| Latest notes on coil actuators/MW actuators
from Matt ... Muscle wire is good were the lightest weight is critical, it can be 1/4th the weight of a coil actuator with the same force. Coil actuators are great were efficiency and low power are critical, like running off small cells, including button cells on rubber or CO2 models. Coil actuators potentially have better precision, because they do not have the 'slop' hysteris effect of the muscle wire around neutral. Muscle wire actuators are good if you need to increase the force of the acuator, you simply re-thread it with the next heavier muscle wire, the voltage will be fine, and it will automatically pull more current and give more force, this with no measureable weight increase because the wire weighs nearly nothing. Coil actuators are less tricky to setup because you don't have to the issue of muscle wire lengths and tensions. Muscle wire actuators can be very flat which makes them easy to install in tail surfaces or wings. The light weight of the muscle wire actuators makes them nice for 'add-on' features like retracts, bomb-drops etc. where the lower efficiency is not a problem for the occaisonal use, but the light weight and high force is desireable. Muslce wire acuators can break due to the stress of the wire if it goes through a hole with a tight radius. Coil actuators should last forever. They both are 'loose' at neutral. -Matt 1st June 02 |
Mw is available in the UK in larger sizes from http://www.milinst.demon.co.uk/ but smaller sizes have to come from the USA. Mark has bought some via our agents in the USA. It is sold in 1m lengths and costs approx. £14 per meter. We use small amounts of it. Maybe, three to five inches per actuator.
It is not easy to work with because you can't solder to it and it is very fine which makes it hard to see and hold. The smaller sizes tend to curl up. Always have a sheet of white paper under it. We have figured out some ways to help work and fix to it, see later one. It is by nature very, very light.
Mark and Andrew H had ago with it just
before we started using coil / magnet actuators.
Since then Andy Bircket has done some good
work with it. Take a look at his throttle for his Co2
motor on the video clip
The most experienced aeromodeller seems to
be Matt in California. Most of the notes on this page
are from him, so thanks for that Matt. Matt has used MW
actuators on many occasions and he has even done retracts
with it - Mark has one of his Mig 15 wing/retracts here in
Leicester
MW actuators are new to us and this page should be seen as a guideline to get us going rather than a cast in stone document.
The advantage of a MW actuator over a coil / magnet actuator is that they are much lighter and easy to make. They are cheap too. If we can make a MW actuator that weighs say 0.25g when a coil weighs 1g then we can make some useful savings when we use three and four function planes. Although a coil is good because it is not directly fixed to the rudder, so during a crash it just flops about.
The MW will use much more current than the
coils. How much depends on how quickly you heat the wire (which
makes the wire move quicker). The length of the wire depends
on the amount of movement you want, the longer it is the more
resistance there will be. So the higher the supply voltage
must be (Current = Voltage / Resistance).
The bigger diameter wire has less resistance
and produces more power, but heats up slower, so it moves
slower.
If we are using one or two cells then we
may need a voltage booster for the control system electronics.
This booster could also supply the MW. If the MW was say 50mm
long (2”) and we used 3.6v supply we would have a load
of say 40 mA. If the battery voltage was say 1v then the
voltage booster would use about 172 mA to produce 40mA at 3.6v.
Obviously this load would not be constant as we don't turn
all the time and we would not necessarily need to use full
rudder deflection either. But a flat battery would not like a
big load like that!
The plane could be a one cell with page motor
which would not use much current or maybe we could use the
new Lithium Polymer cells which we hope to be able to use in
the summer of 2002. These are about 5.5g 220mA and have a
voltage range of about 2 to 3.5v - so we would not need to
boost the voltage as much.
MW actuators could be good for Co2 powered planes. Some of the Gasparin motors now have throttles. Andy has done some tests on a MW throttle and it seemed to work OK. I don't think we would need to switch the Co2 motor to full off, just reduce the power from full the cruise and decent. Test will show how much of a reduction in power that might be? About 15 percent reduction to cruise? After the initial climb you would want full power anyway so it might be better to have a spring to hold the throttle on full power and the MW to switch it off, that way you don't use as much current.
I am unable to find the name of the American guy who sent this sketch of his "cross bow" MW actuator, he said it worked fine, no reason to think it didn't!
Getting the undies down on landing can be a problem for a
light plane using very few cells, because, by definition we
don't have any volts left at the end of the flight. We might
be able to get round this by using a super capacitor to blast
them down!
The multi engine planes have little stubby undies that fit
into the engine bay, these should be easier to make.
I have seen Matt's Mig 15 U/C work and it is a very
impressive sight, lots of power and a reliabel action. They
are help up with a magnet.
If you don't have enough volts, you could run two pieces
of MW in series? Worked for me!
Me thinks?
Make an actuator as a complete cassette which
can be fitted into the plane easily?
Linear or Rotary MW actuators? Don't forget
that if you wind the wire all the way round a drum the wire
will shrink and just crush the drum.
We don't expect to use feed back as our coil
actuators don't have feedback.
To control the MW current from a Z Tron Rx we can use the Rx output to control a transistor which supplies the MW with current from the voltage booster or flight batteries. The standard Z Tron Rx’s are limited to 25mA output from the Pic. chip. Z Tron can supply proportional or bang bang Rx’s.
I assume we will need a robust rudder hinge as the MW will pull the control horn, not rotate it like a coil actuator
Might be good for flaps, similar to
throttle? (If you see what I mean!)
Could be able to active retracts by pulling
the pin and letting a spring pull 'em down?
Maybe use a needle vale for a co2 or Gas
motor? A little bit of movement would be good here but a Co2
motor might freeze up the valve?
Contact!
Mark suggests connection the wires to the MW
with a short length of aluminium tube. The MW is squashed
against the Ali. Tube and electrical supply wire by a length
of heat shrink. Seems to work good, presumably the ali. tube
will eventually corrode - long after the plane has gone to
the great hanger in the sky!
Pull Tube Boogie
Andrew H figured out a neat way of getting the
wire through the aluminium tube. Use a hook of wire. Put the
wire through the tube and grab the MW in the middle of its
length and put it through the tube - simple!
Notes from Matt
Here is an image of a Muscle Wire installation
on the 12 gram, 6
minute Microbat ornithopter.
It had left/right and up/down controls
with the rudder and elevator.
It is a basic pull-pull system. One muscle
wire on each side of the
control surface. I usually loop the wire
through the hole on the
control horn so I don't have electrical
connections near the
hinge line. The length of the control
horn is very short,
approximately .03 inches (0.75mm) from hinge
line to the hole.
You activate one wire or the other for left
or right commands. Don't
activate them together.
Proportional PWM type control works well for MW.
You can replace a coil actuator with MW by
connecting one end of
the muscle wire to the Rx output and the other
to the battery negative.
MW will pull more current than the micro
coils, but you can get usable
movement with around 29 mA and 25 micron MW.
You can replace a RC servo by using just the
servo's amplifier.
Leave the pot at centre or replace with two 2.2K
resistors. Take one end
Of the muscle wire and connect to the battery
negative. Then take the
other end and connect it to the lead that
normally drives the servo motor.
Then it works like a normal servo. If it
doesn't seem to work right, take the
connections from battery negative over to the
positive battery connection.
Using a Servo amp. adds weight and might not
work below 2.7 volts?
Be careful of the current to the MW, don't
exceed about twice the
current recommended by the manufacturer.
Add length or fixed resistor
if necessary. If you over current or over-heat
the MW they will loose about half of
their activation force, which can leave you
very frustrated if don't
realize it.
I have never had any problems with one wire
stretching the other wire
causing damage. Only forcing it by hand or
putting too much current
through it will damage it.
The MW only works when they are stretched,
otherwise you would have no
where to contract it from, and it would just
sit contracted. You need
a spring or another MW or something to stretch
out the non activated
MW. Sometimes it slowly stretches out
over time, but this is not
acceptable for aeroplane control.
I have flown many planes with MW. On
rudder only planes, I used to
install a kick-up elevator, so up elevator was
added to right or left
rudder. Used bell crank, threads and a
spring for that.
For flying wings, use one MW on each
elevon, with a spring. The MW
pulls the elevon up from neutral for turning,
the spring pulls the
elevon down against a stop for neutral.
So you get right, left, or up
controls with 2 MW.
Matt's sketches of his MW actuator:
