Inside Story - April 2003 - RC Groups

# Inside Story - April 2003

## More on motors: Joachim Bergmeyer

who wrote last months feature on motors has provided me with two helpful files to help people make their own measurements and determine motor constants of their own motors.  Here are the two files:

Motor constants spreadsheet: Just insert your measurements, included are results Joachim got for a pager motor.

Motor constant example: An example calculation for the M20LV(KP00) motor.

## Introduction

In the January 2003 Inside Story column Graham Stabler wrote about how actuators work.  In the February 2003 issue he wrote a bit more about how to drive the actuators.  With those as background, this month I will delve into measuring actuator output.  This will not be a column on how to design an actuator.  Rather I show how to measure actuator torque and then present torque measurements for commercially available actuators and along the way I'll present some other interesting tidbits.  I want to thank Matt Keennon, Dave Lewis,  Jean-Daniel Nicoud , Rick Ruijsink, Bob Selman, and Graham Stabler for educating me about actuators and giving me valuable advice and suggestions as I progressed.

## Measuring Actuator Output

The actuators we use have rotary output so we should measure their torque.  I measure torque at zero deflection.  Stall torque is the force multiplied by the distance from the pivot or axis.  This distance is the arm length.  Since the torque from actuators is small, I express torque as g-cm (gram centimeters).  The torque from a Hitec HS50 servo at 4.8 volts is 640 g-cm.  For comparison purposes and to give a sense of scale let's look at the Selman Std actuator.  Its arm length is 6mm.  At 4.8V its output force is 10.1 grams.  So, the torque in g-cm is

Torque(g-cm)  = Force(g) x ArmLength(cm)
= 10g x (6mm/10)
= 6.1 g-cm

The Selman is the torque champ, yet develops less than 1% of a micro servo's torque.

## How do we measure torque?

The diagram below shows the basic relationships.  In the case of a remote actuator, where the magnet has a control arm attached, mount the actuator to a balsa stick or similar held above a digital scale.  3M foam tape works well if there is no mounting bracket. Connect a thread from the actuator arm to a weight that is at least double the maximum output force.  With power to the actuator full on, move the stick up or down so the actuator arm is horizontal i.e. zero deflection.  Turn the power off and use the "zero" or "tare" function on the scale to reset the measurement to zero.  Turn the power back on and observe the negative reading on the scale.  This is the force to be multiplied by the arm length to obtain torque.  The force reading should be taken immediately upon supplying power to the actuator.  The longer the power is left on the more the torque will decline as will be seen later.

The basic components of the setup are shown in the left picture below.  The Selman actuator is mounted on a CF rod that extends vertically from a balsa stick. I glued a z-bend of brass wire to Kevlar thread, with a loop in the other end for the weight.  Brass wire does not interfere with the actuator magnet.  My weight is two one-ounce lead weights with a hook glued to it.  For a test stand I clamped the stick in a hobby vise.  The picture below on the right shows it all set up.  I use an adjustable laboratory power supply so I can set the voltage precisely and this can be seen in the background.  For BIRD style actuators I mount the stabilizer surface to the stick, drill a hole in the control surface and insert the z-bend.

## First, Some New Actuators

Since Graham covered various actuators in the January column a couple of new ones have hit the market.  Let's review them before looking at actuator performance.

## Bob Selman Designs MiniAct

The Bob Selman Designs MiniAct

A very new development is the Bob Selman Designs MiniAct.  This can be purchased either as a complete actuator or as a kit to retrofit existing BIRD style Dynamics Unlimited actuators.  It is available both from Bob Selman Designs and from Dynamic Web Enterprises.  For the rest of this article I'll refer to this actuator as the "DU/BSD Mini" since it uses parts from DU and BSD and is available from either.  The kit consists of plastic moldings and a pivot pin that allow a DU actuator to be easily converted into a remote actuator.  The plastic moldings weigh 90mg, only slightly more than the wire used in the regular BIRD style DU actuator to connect the magnet to the control surface.  The plastic housing has a hole at one end so the actuators can be mounted about one inch apart on a carbon fiber rod and oriented 90-degrees to each other to provide centering force for the control surfaces.

An important result is that the same coils and magnets when utilized with the BSD kit generate 40% more torque than a set of plain DU actuators.  Why does the BSD kit develop more torque from the same magnets and coils?  First, the magnets are spaced 2mm apart, making the magnet effectively bigger and moves the magnet mass into the stronger magnetic flux lines close to the inside surface of the coil.  It also may provide more leverage since the magnets are further out from the pivot axis.  However the exact mechanism has not been investigated and the technicalities are beyond the scope of this article.

Another performance benefit of the BSD kit is the fact that the magnets are precisely positioned inside the coil.  Even small deviations away from perfect centering can cause differences in torque between the two deflection directions of 10% or more.  It is virtually impossible to get the magnets perfectly centered in the coil when they are held within the coil by a flexible wire, as in the DU "BIRD on a wire" method.  So, it will also be almost impossible to get the same torque in both directions with a plain DU actuator.

Over 1,000 RFFS-100 systems have been sold since they were introduced in January 2002.  As a result, there are probably more DU actuators being used than any other.  The recent widespread adoption of very light lithium polymer cells for micro planes makes it difficult to balance a semi-scale plane with nearly two grams of actuators mounted way back in the tail surfaces.  These kits, at \$5 each, allow the DU actuators to be easily converted to remote actuators and moved forward near the CG to control the surfaces via either lightweight control rods or pull-pull threads. The increase in torque is an added benefit. These kits fill a widespread need and should be on the must-have list of anyone flying with an RFFS-100 system - along with upgrading the PIC chip to get more torque and other enhancements.

## Didel

 Didel, Pico, Nano, Micro and Mini "Birds"

Didel has recently introduced a line of actuators in 0.28, 0.50, 0.70, and 1.25 gram sizes that they call Pico, Nano, Micro, and Mini Birds, respectively.  They are unique in that a radially oriented magnet is used, rather than the traditional magnet where the poles are oriented toward the flat sides. This allows a larger magnet to be used for a very close fit within the coil. The magnet is thus closer to the stronger magnetic flux lines from the coil, and more force is generated.   They incorporate a molded plastic housing to hold the coil and the magnet and a molded plastic cover to protect the coil (shown removed on the top actuator in the picture).  Coil wires are soldered to one end of the molded-in pins and the user solders the wires on the other end of the pins to go to the radio. This is a neat feature, as the user does not have to directly handle the delicate and very thin coil wires.  The metal control arm on the actuator has a range of holes to allow fine tuning the relationship between the actuator control arm and the control surface arm. The control arm has integrated stops to precisely limit actuator throw.  This is a very well thought out family of actuators and fills a void in the range of commercially available actuator sizes.  Best of all, they have good torque to weight performance.

## Actuator Characteristics

I performed all the tests at 3.5 volts.  One of the most common combinations used for micro models is the M20 motor equipped KP-00 gearbox and a single 145mAh Kokam lithium-polymer cell.   After two minutes of running this combination with a typical prop the battery voltage under load will be about 3.5 volts.  This represents a reasonable approximation of the torque that would be obtained in practice.

Characteristics for commercially available actuators are presented in Table 1.  The actuators were powered directly from a laboratory power supply.  This means there was no voltage drop due to the internal resistance of a receiver's coil driver.  Consequently, these measurements represent the torque that an actuator would put out in ideal conditions with 3.5V applied directly to the coil. The arm length in millimeters and force in grams are presented so the reader can compare other actuators to the ones measured here.  However, raw force should not be compared across actuators because the arm lengths differ.  Instead actuators should be compared based on torque, which is independent of the arm length.  Since BIRD style actuators like the Mueller and DU do not have a control arm, I measured the force 6mm from the center of the actuator in the control surface.

 Table 1.  Actuator performance at 3.5v

## Actuator Efficiency

Raw torque data does not tell the whole story.   For example, the BSD standard actuator produces 2.8 times the torque of the DU actuator.  But weighs a little more than twice what the DU actuator does.  The DU/BSD Mini (DU coil and magnets with the BSD remote actuator kit) develops 40% more torque at only 70mg additional weight compared to the DU. The Didel Microbird generates similar torque as the DU, but with a 23% weight savings.  To get a handle on the various trade offs we can produce measures of efficiency.

As well as torque, Table 1 contains three of these measures of efficiency.  The first efficiency measure is torque divided by actuator mass in grams, or T/m.  The larger actuators tend to score better because as most of the actuators shown have approximately the same resistance the weight will often come from the use of thicker wire or a larger magnet.  Both increase torque, (the latter directly and the former due to a higher number of turns) more than they increase weight.  Auxiliary components such as the pivots and arms also make up a smaller percentage of the total weight.  Some actuators stand out from the crowd such as the Didel Microbird with a measure of 2.2 while still only weighing 0.7g compared, for example, to the DU at 1.7 and 0.9g.  It has more torque at less weight!  Another is the DU/BSD Mini that has a torque/mass efficiency that is better than all the actuators that are heavier than it.  The bottom line for most models will be the weight so you should select the actuator with the highest T/m for given weight.  Conversely if you know the torque you want then selecting the actuator with the highest T/m will give you the lightest actuator.

The second measure of efficiency is torque/current, or T/I.  However, with the move to LiPoly cells, we are not constrained as much in terms of current.  In a typical M20 based system, the motor pulls about 700 mA, the receiver system about 10mA, and each actuator consumes about 50mA.  So, the current draw of an actuator is about 6% of the total and will generally not influence the choice of actuator.  In fact because all the actuators have roughly the same resistance and hence draw roughly the same currents T/I will follow the torques closely and of course the larger actuators with the larger magnets and greater number of turns will do well again.

The third (and better than the second) measure of efficiency is (torque/current)/mass, or (T/I)/m.  Overall, larger actuators seem to have an advantage in terms of efficiency.  Among the 0.9 gram and larger actuator the Didel Minibird stands out with the highest efficiency, even higher than the much heavier Selman Std.  In the middle sized group of 0.5 to 0.7 gram actuators it is the Didel Microbird that comes out on top.  In the category of 0.25 to 0.33 grams the Picobird scores well above the rest.

## Getting the Most Torque from the RFFS-100

The graph below focuses on the DU actuators generally supplied with the RFFS-100.  The torque obtained directly from a laboratory power supply is the maximum that can be obtained from any given actuator.  As can be seen, before performing the PIC retrofit or buying one of the new receivers, the actuator delivers just 34% of the torque that is possible. But, what level of torque will you get with the new PIC chip?  The answer is not a simple one.  It depends on your transmitter.

Torque potential for DU actuators

First, you need to determine whether you are getting "full duty cycle" at full stick on your transmitter.  Full duty cycle, for an RFFS-100 type of pulsed (PWM) actuator controller, is obtained when the current to the actuator is constant.  Pulsing the current causes the actuator to hum so you can test to see if you are getting the full duty cycle, and full torque, by listening for the actuator humming when the transmitter stick for one channel is pushed all the way to its stop.  If you hear no humming, the actuator is receiving full current, and generating full torque.

If your actuator still hums at full stick deflection, try sliding your trim tab for that channel all the way in the same direction as the stick deflection.  If the humming stops, this is full current, but you may not be able to achieve it with your transmitter unless it is computer programmable.  If your transmitter is programmable, you may be able to obtain full duty cycle by setting your EPA (end point adjustment, sometimes called volume) to its maximum value, often 125%.  The trim tab check, unfortunately, cannot be used when flying, so my Focus IV cannot achieve full duty cycle.

I used a Hitec Eclipse programmable transmitter for the tests from RFFS-100.  I also checked my tests with a Hitec Focus IV.  Both transmitters resulted in the same torque when used with standard settings with the RFFS-100 with retrofit PIC.  After programming the Eclipse to have 125% EPA, and sliding the trim tab to its stop on the Focus IV, then both transmitters yielded the same, but higher, torque.  The 94% and 80% of possible torque bars in the graph above are for the Eclipse with and without the 125% EPA.

To reiterate, some transmitters, programmable or non-programmable, may result in full duty cycle and torque without needing any programming.  Other transmitters may not but if they are programmable this may be solvable with EPA programming.  Finally, the reason even full duty cycle from the RFFS-100 results in 6% less torque than from a power supply is because of losses due to the internal resistance of the RFFS-100 circuitry.

## Actuator Torque in Practice

The torque available at 3.5 volts for larger and medium weight actuators and various RFFS-100 setups is given in Table 2.  The RFFS-100 w/Retrofit is from a Hitec transmitter that does not give full duty cycle.  The RFFS-100 w/Retrofit 125% EPA is from the Hitec Eclipse transmitter with the end point adjustment to deliver full duty cycle.  These are intended to give some idea what you can expect if you are able or unable to get full duty cycle from your transmitter.  This allows comparisons to be made between particular actuators and actuator driver setups.  Torque from a laboratory power supply is included for reference.

One of the results most interesting to the vast number of RFFS-100 owners is that torque can be increased from 0.49 g-cm for a DU actuator and a non-retrofit RFFS-100 to 2.02 g-cm torque with the DU/BSD "Mini" kit and a RFFS-100 with a retrofit PIC chip and the 125% EPA technique from a programmable radio.  This is slightly more than four times the torque with virtually no increase in weight!  The table also shows that for planes that can afford more weight, going to the same setup but with Selman Std actuators results in more than eight times the torque.  Similarly, for planes that need lighter equipment, going to the retrofit/EPA setup and the Didel Nanobird would result in half the actuator weight but nearly 50% more torque at 0.69g-cm.  This table also has torque for the BSD DDPC (dual digital pulse converter) board when hooked up with the GWS R4P receiver.  Since this system weighs more than an RFFS-100, the Didel Nanobird is not evaluated with it.  The lower torque with the DDPC compared to the retrofit RFFS-100 is because, the speed controller used includes a 5v regulator and as this was only being driven by 3.5v, it caused a small voltage drop.

 Table 2. Actual torques using typical RC electronics

Table 3 gives torque for the smallest actuators.  Since the smallest actuators are not likely to be used with the comparatively heavier RFFS-100 system, only their torque from the Leichty integrated receiver and actuator driver system (link) is given.  Torque from a power supply is, again, included for reference only.  The Didel Nanobird, at half a gram, clearly generates the most torque.  However, it may be on the heavy size for many planes based around the Leichty system.  The Didel Picobird at 0.28 grams, however, does generate more torque than either the Leichty MX or Std actuators.  It is, however, slightly heavier than the Leichty Std actuator.  The Leichty Mini and Micro actuators are in a league of their own in terms of weight at 0.16 and 0.11 grams each, respectively.  Their torque is also considerably less.  These actuators are probably best used on the smallest and lightest of planes.  Tiny Co2 powered planes come to mind as a perfect match for these actuators.

 Table 3. Actual torques using typical ultra small electronics

## Torque and Deflection

As the magnet in an actuator deflects away from its neutral position its poles move further from the more powerful magnetic flux lines close to the coil.  Thus, torque should decrease as deflection is increased.  The graph below shows torque as a percentage of zero-deflection torque for the BSD MiniAct.  This demonstrates an interesting relationship.  As the control surface is deflected further into the air stream, which requires more force, there is actually less torque available to push against the air stream.  Still at 25 degrees deflection 75 percent of max torque is being generated.

 Change in torque with actuator deflection

## Actuator Torque at Higher Volts

So far all the results I've presented have been from powering the actuator at 3.5 volts.  It's worth a quick look at what happens to torque as voltage is varied.  For any given actuator, as volts are increased current is increased and torque also increases.  Table 4 shows torque for the DU/BSD Mini at various voltage levels.  The 2 volt level was chosen because it is roughly equivalent to two 120mAh NiMh cells under typical loads.  With LiPoly cells in widespread use almost no one will be flying an RFFS-100 or similar system on this type of pack.  Torque at 5 volts was chosen because it is the maximum voltage an RFFS-100 can take, and is what can be achieved with a DC-DC voltage booster from a LiPoly cell (if the motor doesn't pull too many amps).  Finally, torque at 7 volts  corresponds to a 2-cell LiPoly pack if there were a receiver system that could operate at 7 volts and also deliver 7 volts to the actuators.  No such receiver/actuator driver currently exists but it's interesting to explore this possibility.  The graph below plots the torque versus volts and shows that the relationship is linear.  The slope of the line may differ somewhat for different actuators because of their design, number of turns and wire size in the coil, and magnets.

 Torque change with voltage

So at first look it seems that increasing the volts 43% from 3.5 to 5.0 volts results in a 43% increase in torque and that doubling the volts to 7.0 results in a doubling of torque.  The problem is that actuators are just like motors and at higher current levels heat up.  The longer they are subjected to higher current the more the magnet and the coil will heat up, and the greater reduction in torque.  This is because as the wire heats up its resistance increases and the current goes down and hence, so does the torque.  This can be thought of as heat-related torque fade.

The graph below shows the results from measuring torque at 15-second intervals for the DU/BSD Mini at 3.5, 5.0, and 7.0 volts.  Torque at a given voltage will be at a maximum when a room temperature actuator is first powered at that voltage.  I measured torque at subsequent 15-second intervals and expressed them as a percentage the corresponding max torque.  This allows easy comparisons of torque fade at different voltage levels.  Note that these measures are at "full duty cycle" which is roughly equivalent to maintaining full stick deflection when flying.  If the actuator is not subject to sustained full stick deflection as simulated here, the torque fade will be less.  And, if the actuators are mounted in the open there will also be some cooling effect from moving through the air.

 Example of titled image

These results show that at the "normal" 3.5 volt operating level the DU/BSD Mini's torque declines to  about 92% in the first 30 seconds of operation and is down to 82% by 120 seconds.  At 5 volts the corresponding declines are down to 90% by 30 seconds and 77% by 120 seconds.  At 7 volts, however, the fade is much more dramatic.  By 30 seconds the torque has declined to 82% and by 120 seconds it has declined to just 59% of its initial maximum torque.  Since the torque fade is due to heat losses I subjectively measured heat with my finger at the end of each test.  At 3.5 volts the actuator was not noticeably warm.  At 5 volts it was warm.  And, at 7 volts it was hot and I quickly removed my finger.  What this tells us is it is probably OK to run these actuators at 5 volts with a voltage booster.  However, if there were a receiver that allowed powering the actuators from two LiPoly cells the actuators might run too hot.  At 7 volts they would either need to be used with moderation to keep them from heating up, or actuators specifically designed to handle higher current levels would need to be used.

Propellers

## History

When I first began flying micro electric powered models in the early 1990's I found the power options to be quite limited. The numbers of excellent power systems that we enjoy today were still years from coming to the market. Micro flight was mainly a build it yourself proposition. If you could not buy the parts you needed from one of the few small suppliers offering propulsion systems, the only option was build it yourself. Propeller choices were mainly limited to small glow, Co2, and rubber props.  Rubber props were the most attractive due to their lighter weight but most had either too much pitch, or blade area. The hub cross section was very often too thin to allow for mounting on a prop adapter or gear shaft.  When I progressed from direct drive to gear drives, I often used a 5.5" G-Mot Co2 prop that worked fairly well. The hub was configured to accommodate a mounting screw, and it was nearly as light as the rubber props. The blade area was better suited to small electric use, but some option for fine-tuning of the drive system was still needed.

In 1998 I was developing a model for the DEAF (Dallas Electric Aircraft Flyers) annual fly-in to compete in the "Lightest Aircraft To Fly 10 minutes" event. I had an 80-gram model that could consistently deliver 8 to 8.5 minute flights on the stock G-Mot prop. Changing the gear ratio was more effort, and represented a larger change than I wanted to make. I felt the best way to reach the 10 minute goal was to decrease the operating current slightly by adjusting the prop. I decided that the best prop for the task would be similar to the Cox 020 prop with increased pitch. I decided to see if I could make my own version of the Cox prop using the G-Mot as the base. I used an old broken Cox prop as the template for the new modified G-Mot prop. It took less than an hour to produce the modified G-Mot prop, and I was able to further customize it a bit by adding a quarter inch to the diameter of the stock Cox part. The pitch of the G-Mot also seemed to be just what I wanted. After balancing and static testing, it was time to head to the field for a test flight. The first test flight was slightly over the 10-minute goal and I could not really see any significant difference in thrust even though the current was now about 15% less at full throttle. I was lucky enough to have good weather at the fly-in that year and did indeed manage to win the event. I have been an enthusiastic advocate of this technique since that time.

 The first attempt at reshaping a prop. The gray Cox blade was used as the template for the new blade profile. The stock G-Mot prop is shown at the top.

I still employ this "tuning technique" on virtually every model I build. Even though there are many excellent power systems and prop choices available today, I find this technique to still be a valuable and powerful tool to optimize the power system for each model I build. I know a few modelers who are now molding their own props from carbon fiber. Reshaping the plastic prop is a very easy and economical method to verify performance before committing the time and materials to produce a carbon version. The plastic prop can also serve as the mold for the carbon fiber prop.

 An assortment of the popular Gunther   props.  The two at the top are enlargements from the stock prop, shown in the middle. The black prop was cut from the VL Products 7" prop. The red prop was cut from the 7" Modela  (for more pitch). The two on the bottom are reduced versions of the stock Gunther   prop.

## Making the templates

Once the paper pattern is in place, its time to start cutting. Cut the plastic away to within 1/64" to 1/32" from the outline of the pattern. The remainder of the material is removed by sanding with medium grit sandpaper. As material is removed, a burr will be raised fairly rapidly. Keep removing this burr as you work the material down to the edge of the pattern. Lightly scraping the cut edge with a knife blade will remove the burr very effectively. I use a 4-power jeweler's eye loupe to inspect the cut and verify that the plastic is removed to the edge of the pattern. The eye loupe is very critical, and helpful in completing the job accurately. Once the blade is trimmed and sanded to the new profile remove the paper pattern and make sure the edges of the new blade are free of any burrs. Don't worry about tearing the paper pattern while removing it, as it will not need to be reused.

## Profiling

Using a file, sandpaper, and scraping with a knife or razor blade begin removing material from the front of the cut edges of the blades. Its very much like carving the airfoil into a free flight hand launch glider wing but much smaller. Once again, the eye loupe is critical in doing this step accurately. Check your progress by inspecting the blade edges with the eye loupe. Its much more difficult to describe the process than to actually do it. As you work the edges down it will be obvious where to remove material as you inspect it. You will need to frequently remove the burr formed on the edges as you work the material down to its final shape. Do a final sanding with 150 grit sandpaper followed with 320 grit paper. If you are careful and do an accurate job, you might find that you don't need to balance the finished prop. Balancing and wet sanding are the final steps.

## Finishing

After the prop is balanced, wet sand with 320 grit paper followed by 400 grit paper. The 400 grit produces an acceptable finish but you can go to 600 grit if you want a little finer finish. The wet sanding only takes about a minute to complete. Be careful not to over do it, or you may alter the blade profile.

 A selection of stock and modified plastic props.  From top to bottom are the stock 7 " Modela, below is a 6" Gunther cut from the Modela, the VL P1 with two custom versions below, a stock GWS 4.5" X 4" with a 90% copy below, and a stock GWS 2.5" X 1" with three variations below. The smallest is 1.625" diameter and was used in a direct drive pager experiment.

## Conclusions:

It should only require about 30 minutes to rework most props once you gain some experience and get comfortable with the process.  Once you try it, I'm sure you will find that the results are well worth the time invested.

## Mini Article - Hot wire

cutting for micros By Peter Frostick

## Peter Frostick - Who he?

I'm of WW2 vintage, and have been an addicted model flying nut since 1950.  Surprisingly, despite a working career in model making, I still enjoy it as a hobby!   Indoor flying first grabbed me in the mid seventies with the dawn of the Peanut scale movement, when Bill Hannan and his merry men sowed the small scale flying seed for many of us.     The dormant radio bug soon returned however, and by the 1980s enormous but light models were occasionally steered around gyms by some very nervous pilots - hostility was common; so a difficult  time to 'boldly go'!.

## History (or why bother hot wire cutting)

My 'flash of light' conversion from stick and tissue (I do still build them) occurred at the local model shop, where a piece of 'Dow Floormate 200' blue-foam was waiting to be turned into gliders by the magicians at the back of the store.    I picked up a piece, which seemed a lot lighter than balsa and also a lot cheaper!    At first the process of working this foam by way of carving and hollowing out seemed very messy (still does!) so hot wire cutting in miniature was tried for the flying surfaces.   This works like magic, and saves a lot of dust and partner hostility!!    It is nothing more than normal RC wing core cutting, refined for the 'Small is beautiful' enthusiasts.   It's also very quick, and desired trimming warps can be cut into the foam, rather than the' Bend it and hope it stays' approach of former times!!

 My Tiny rubber powered Mustang with wire cut flying surfaces --- when everything else fails, this always flies!!!

Initial small scale cutting attempts were carried out using easily available 0.45mm 'Constantan' resistance wire.  Results on larger surfaces were OK, but the wire proved fragile in use, and dragged behind in the between-template regions due to insufficient tension.  Some encouraging results were obtained with surprisingly small wing sections, but it was not until indoor RC 'guru' Dr Chris Fouweather kindly sent me some really thin Nichrome wire that dependable results were produced.    The 0.22mm Nichrome is best run at temperatures a little below visible red heat, and may be used at surprisingly high tensions. The quantity of local heat required is so small that foam sections down to 1/32" thickness are cut with minimal distortion.   By positioning the tinplate section templates at different Root and tip angles prior to cutting, tailored twists can also be cut into the wing permanently.     The picture of my little 8" rubber powered 'Pistachio scale' Mustang shows an unpromising layout to trim, with a propeller half the wingspan and tiny tails surfaces; but thanks to a wire cut right wing incidence progressively decreasing 2.5degrees from root to tip, it has flown reliably straight from the model box for seven years without any hassle! --- sounds a pretty good deal to me?

 'Old Faithful' Spitfire airframe was built in three hours flat!! --  this model has flown with  I.R.,  experimental wide band FM,   proportional single channel super-regen,   and is about to be tried with a BitCar conversion.    Very good value, and still flies with the original direct drive B2 motor!

!The Bow and its "String"

The simple cutting 'bow' pictured is sturdy, and made from 2x1 inch pine screwed and glued: I usually clamp the bow to a bench, and bring the work to it for convenience.   You can tension the wire using either flight rubber or a spring: for the musically inclined, tuning to an octave above middle 'C' for a 12" wire is good starting point!    As a rough power guide, wire of around 0.22mm will need about one volt per inch length, so the humble car battery charger should be OK for the job with small projects.     To adjust the wire's power consumption, leave a loose length of cutting wire to clamp connectors to, and experiment until the heat is right.   Rules of common sense, self preservation, (and aversion to pain) apply to all these processes.   Always carry out these tasks in a well ventilated room to avoid creating your own domestic 'skunk-works'

## Cutting

Close textured Styrofoam is easy to work, and even comes in colours other than blue!     The Fina petroleum company in Germany make version of it called 'Flammbar' in a light beige, which is handy for natural fabric finishes.  Its intended use, like the familiar blue, is for building insulation.    Surprisingly this material has a 'grain', and is stiffest along the extrusion axis, which the same as that of the logo printing.  Thin metal rib templates require a very smooth edge finish or they can snag the thin wire, causing surface irregularities.     An easy beginners method of cutting a small wing is to stick the templates to a trued block of foam with double sided tape to make the upper surface cut, then simply relocate the templates lower down to cut the under surface: this leaves you with a minor sanding job at leading and trailing edges, but at least you now have a pre shaped foam block to rest the work on while doing it!    Cutting speed judgment is real 'seat of the pants' engineering.    As you cut, the wire makes a gentle hissing sound, so adjust the wire speed for the loudest hiss and all should be well.  For those of you who need the very lightest airframes, some extra fine tuning is needed!   A cutting wire running too hot will not only melt too large an area of foam; but will deposit the heavy melted residue permanently on your flying surface.    When a cut is done at the ideal speed, the surface should have loose spider web of  styrene 'fluff' on it, which can be removed with sticky tape to reveal a very good surface finish --- this may sound nit picking, but the result of a poor cut on 1/32" sections can double the weight.  It always feels to me that a slightly 'forced' cutting pace gives the very best results.   This entire technique will seem tricky at first, as new reflexes have to be acquired, also you sometimes feel that five hands are needed as well; 'humanoids' can master it with practice!!  Trust me.

 This 15 gram, 13" span 1950s style power model  was originally  capacitor powered electric freeflight, but has now become a real 'hooligan' with LiPoly cell and two channel I.R. control --- fast furious fun.   Flown sensibly, half an hour's air time is possible.  The wings are unsanded, straight from the cutter!

There are a few ways of cutting flat foam sections, the simplest being to 'eyeball it' and hold two engineers squares onto the block,  then pull it over the hot wire. British modeller John Soper came up with a nice idea a few years back, in which a wire was stretched at the desired cut thickness above a flat board; the board was then tilted to 45 deg. And a suitable weight attached to a true block of foam to 'drive' it over the hot wire by gravity --- very clever, and excellent: Sir Isaac Newton would have approved!!

Another considerable benefit of the system is that different wing thicknesses can be used for root and tip wing sections by simple template positioning: a more elegant structural solution for tapered plan form wings, and lighter.

Once you get into this technique many new ideas and methods will quickly surface --- one member of the famous 'Leicester Aeronutz' club even made a low speed hot-wire lathe to produce round fuselage sections.   Of course if your name is Graham Stabler, you could even build a CNC stepper motor driven version!!

A quick bout of web-wandering should locate the resistance wire for you --- Our European needs are well served by the Scientific Wire Company www.wires.co.uk well worth a visit jut to see the varied offerings they have; even Titanium wire!

I wish you all the best of luck with this fascinating and now quite ancient process; enjoy.

Peter Frostick

## Micro Tip -Printing color and

markings directly on balsa By Paul Bradley

## Introduction

When it comes to indoor and micro models, weight is everything. We all love a nice looking model, but as the size begins to shrink the finish is often sacrificed in the name of saving weight. If you happen to use balsa for your your indoor and micro creations, I have a solution to the color and markings weight problem. Thanks to the technology of computers and ink jet printers we can create some nice looking models with no weight gain. Let me say that again .... no weight gain.

This short pictorial article describes a process for printing color and markings directly on balsa using an ink jet printer. You can create your killer design and then print it right on the balsa sheets that you will be using to build the model. While your at it, you can add some color and markings to dress up the finished model. One of the best parts is the color and markings are applied before the model is even built, so you have something nice to look at while that special bird takes shape. Want to give it a try? ... I hope so.

## What do you need?

There are two basic requirements for printing on balsa. You first need a printer that has a straight through paper path. I use a Hewlett Packard model 1120. There are many printers that have this feature. It also helps if the printer will support banner printing. The second thing you need is some graphics software to develop your artwork. Just about any program that will let you draw will work fine. I prefer software that uses vector graphics so they can be scaled without any loss of resolution. Programs that draw using bit maps can also be used very effectively.

## The process.

The photos that follow will take you through the steps. Basically, the process is  create your artwork and then print the balsa sheets. Here are the more detailed steps.

 Begin by creating your artwork. The example shown hear has been developed using CorelDraw