Dedicated to the Promotion of Electric Propulsion in all types of Aeromodeling

Multiplex Gemini(s) After Two Tries Let’s Get the (Final) Pair Really Right


For most RC pilots a hundred flights on one of these simple foam airframes,  when flown off of something recognizable as a runway,  seems like a reasonable  expectation.  The build-up will be directed to that goal,  even though we expect more out of ours.  Even landing them on long grass as a crop instead of compressed sand.  A Gemini could be an aerobatics trainer,  although your first experience with ailerons and “full house” (zu Deutsch nicht volles Haus,  sondern ein USA RC Fliegerausdruck for Alle drei Flugaxen plus Motor gesteuert) ought to be already in grip.


But when we just wanted to fly relaxed and enjoy an outside experience,  despite of,  or maybe because of,  a cumulative thirty other flying things between the two of us,  the Multiplex Gemini was a favorite.  Compared to our F3A (all out precision aerobatics  see VisionAir) and 4D (flat plate things with huge control surfaces,  see Park Master or DW Airplanes) types:  Knife edge flight (only level,  no loops),  aileron rolls may be axial or not depending on speed and asymmetrical upright/inverted flight are the weaknesses,  and the strength.  The Gemini is a stable platform able to buck some wind and it doesn’t require constant/absolute attention required when remote piloting “pattern” and really agile stuff.  


Real Flight (5.5) doesn’t have anything really comparable,  the closest might be the virtual Park Master.


If you want to “stunt” some,  have ordinary flying skills and want to fly carefree,  the Gemini was a great choice.  Was as they are being phased out!  But ours would never have made it that many flights without some select reinforcement,  and only hitting the ground when intended.


Much of the journey isn’t airframe specific,  I keep my notes in this form as the procedure is entertaining.  The Editor and I go clear back to my brain damaged four years of life,  back when most people didn’t believe in electric powered RC flight,  now a decade and a half mostly recovered from,  he enjoys his role as publisher.


So,  here in condensed form is our experience,  flight and build logs,  compressed down to what functioned best,  with some things to avoided too.  It was,  and is,  an Radio Controlled flying adventure.  And not just made three flights and everything is great of most reports.  We had twenty different versions of propulsion and four kinds of servos along the way.  As in doesn’t anybody else ever sit there frustrated with things that don’t go right on the first try,  or on flight (380) accidently discover a combination,  just slightly different then the twenty previous ones,  which is great?  We fiddled with and fly our stuff until they are really right.


About a Stock Gemini
If we exclude the transmitter,  receiver and the other stuff required for general RC flight (such as the battery charger),  from the economics,  figure the cost of a new Gemini with all new parts and three batteries was in the four hundred to six hundred dollar range.  As always,  if you don’t have a clue,  just follow Multiplexes recommendations.  Propulsion wise in the USA that means the “Tuning” i.e. heavy (135) gram motor.  The original was an (80) gram motor,  still available for the Park Master.  Which papers out to around five bucks a flight for a hundred flights.  To keep that in perspective,  at three flights a session that’s nine months of once a week flying.  If you have an inbuilt Gemini kit please let the Editor know as I’ll buy it off of ya,  can’t get them new anymore.


If you had assembled a foam airplane lately,  and all your electrical and radio stuff worked correctly on the first try,  figure about six to eight hours to assemble a Gemini.  Add another hour to tape the leading edges and another hour or two to iron on the excellent decals they come with.  But why hurry?  Unlike an ARF,  you get to chose the components and there is still some satisfaction to having had a hand in assembling what you fly.  That choosing the components is something edited out by the censorship of omitting experience beyond just making three flights.  I’ve seen published reports where all they made was one flight!  We made (450) flights and assembled four airframes for this report…


Of all the Multiplex lineup the Gemini is the easiest to glue together straight.  Who’d have thought of that as biplanes are usually a real expletative deleted to get the wings aligned with each other and the rest of the airframe.


The single most important assembling a Gemini step,  omitted with staggering frequency,  is to take five minutes sanding the leading edge of the wings smooth.  Use grit in the (320) to (220) range (i.e. fine,  about what you would sand car paint with),  watch it as the foam of the wings is soft,  which helped keep the weight and cost down.  While you are at it the trailing edge of the wings also often have a slight lip,  which screws up airflow all out of proportion to your expectations.  It takes a pleasant hour with a sanding block,  the best quality packing tape and a covering iron to make the wings leading edges last a long time.  Without some hardening the leading edges quickly abrade away,  it still flies,  but not near as well.


Me,  I enjoy working everything over for an hour or so sanding.  That makes applying my usual tape (leading edges and hinges) and fiberglass easier.  It helps with the decals and any paint too.  I’ve done more packing crate (foam and tape) airplanes then the Harbor Soaring Society allows (our manufacturer of the world’s best electric RC motors has been open about his disapproval too),  it takes me an hour to tape the leading edges and hinges of a Gemini.  Even at that about one strip in five of tape I have to remove and re-do.  Use a covering iron takes out the wrinkles,  more important,  the heated tape binds much better then cold pressed.


I don’t know what the difference really is between the Elapor specific CA (glue)  and the many other fine CAs out there,  but I preferentially use the Multiplex glue.  Even if it costs a couple of bucks more,  somehow Elapor goes together with it better.  CA is the only glue I use on Multiplex airframes.


The basic servos are either Multiplex Nano(s) or HiTec HS-55s (plus cable extensions not required for the Multiplex servos in the Gemini,  which might swing the economics in their favor,  the Multiplex quality control may be better too) in the eight gram size,  which may be expected to outlast the airframe and give satisfactory performance.  The HiTec HS-65s (costs about twice as much per servo) heavier at twelve grams (identical external dimensions) with their better precision and stronger,  more durable gear trains are worth every bit of their additional price.


From Mini-Mag experience somewhere out beyond (300) flights the gear trains of HS-55s,  and possibly the electronics inside (that’s a mechanical potentiometer in them),  wears out.  We haven’t even approached the limits on HS-65s.  Wreak an Elapor airframe,  cut the servos out,  scrape off the glue and foam,  and use them in the next airplane.  I’ve had the twice as big HS-82s last through three Twin Star IIs for (600) flights.  If you have them,  the twice as big ((18) grams verses (8)) servos fit in just fine with just trimming the foam a little.


However, if you misunderstood that the little aluminum connectors for the wires from the servos to the control horns must be both snug and still turn,  things don’t do well.  Even locked on hard,  so the wires have to bend,  things go sort of ok for a while,  until the servos quit.  Don’t forget to CA glue the little brass nuts on.  If,   with the parts in your hands,  that didn’t make immediate sense,    get yourself some help!  Those little bitty parts are essential,  and prone to rolling off somewhere,  try assembling over a towel.


It is exquisitely easy to change the propulsion of (almost) all Multiplex airplanes.  Two screws hold the motor mount plate on,  the motor is hung off the back,  changing stuff can’t be any easier.  That handing the motor off the front to a plate along with impact resisting foam makes for durability the likes of which we didn’t even dream of not long ago.  The modern aluminum motor mounts are fine,  the plastic ones warp.  It depends,  the “frame rails” in form from the hard plastic inserts glued into the fuselage sides may not hold up all that well either,  those four screws for adjusting motor angle (down and to the right) dig into them.  


Motor Mount


Out around flight (160) I sanded the nose of Gemini 2 flat and fiberglassed on a wedge of plywood.  By which I adjusted the down thrust until it was about right.  About right as in on application of full power it first pulls forward and down (figure with a (135) gram motor in thirty feet forward about two foot down),  then it angles up a little up.  With the wings two different sizes and the motor in between there is no perfect angle with the Gemini airframe,  and it will vary depending on the motor and propeller.  If you are inclined to flying a Gemini slow,  aileron rolls at low speed are corkscrews,  fly a little faster and they straighten right out.  That’s most pronounced with a heavy motor,  ten or eleven inch prop with the pitch half the diameter (10X6,  11X6).  See above,  the Gemini isn’t a precision pattern airplane.


If you fly the lower ailerons stock (as in just glue the control horns on with CA) the control horns stay on (no problem with the rudder or elevator horns tearing out either,  unlike the Twin Star II),  but the aileron foam between the control horn (to which the servo is attached) bends between it and the link to the upper aileron as we observed on Gemini 1.  Gemini 2 the lower ailerons were fiberglassed the inboard three inches,  it was the top ailerons which developed a bend at the control rod attachment.  These wonderfully simple to assemble foam things have just slight differences in foam density which accounts for things like that.  


I like my flight axis balanced,  two aileron pairs were too quick relative to the elevator for me,  even with minimal throws,   so out around flight (170) I disconnected the upper ailerons and flew on with just the lower ones,  which,  by a test flight,  did not suite the other pilot.  I have a fluid,  energy efficient style (left over from when electric powered RC airplanes barely flew) the reduction in drag (those bare wires have almost as much resistance to airflow as the foam struts between the wings) makes for a better glide and low powered phases.  The other pilot alternates between airplanes,  tendency F3A (precision dynamic aerobatics) and getting faster airframes all the time.  He has now reached the state of proficiency where he needs to enter some competitions,  to get encouragement from other demanding RC pilots.


I like a simple,  if garish,  paint scheme.  Minimal paint as paint can weigh a whole lot more than most pilots recognize.  The other pilot liked both the decals which come with the Gemini,  and some additional contrasting color paint.  If fresh,  the decals (very easy to apply) press right on.  As a veteran of iron on plastic covering (and old enough to remember the hassle of shrinking paint (i.e. dope) over paper or silk) I iron on my Multiplex decals.  Ironed on the decals will outlast the airframe.  They may even have some structural value.  


That contrast (between top and bottom,  and in my case front and rear too) is more important to a Gemini then most,  we can’t see them well,  and with it ability to decide orientation,  on them out beyond a hundred yards.  When you do paint Elapor keep in mind if you use glossy paint (verses flat) that suddenly the intersections between the foam cells,  plus the inevitable uneven surface of expanded in the molds foam,  shows right up.  Then go for it.  I use spray paint in the ten dollars a can quality,  the available Elapor primer then isn’t needed.  


The Gemini is a nice looking RC airplane,  you want a spiffy looking RC airplane,  get an ARF,  and tell me about repairing it.


Over time all plastics,  rubber and any non-metallic material with some stretch and compression properties (such as EPP and Elapor foams) goes hard and brittle.  A reasonable expectation for the Gemini is around five years to eight years after production (at San Diego temperatures,  i.e. perfect),  if you fly it or not.  There’s another issue,  the wire inside the plastic sleeves rusts,  binding things.   


Additional lessons from Geminis 1 and 2,
for longer life airframes as we did to Geminis 3 and 4.  


Without taping the leading edges (after first cleaning up the casting flash from the mounding process) we’d have been ready to junk these things,  or at least replace the lower wing,  after thirty flights.  


Since ours “bonked” (zu Deutsch,  etwa ein leichter Stoss die einem Klang,  als ob man auf einem Rohr geschlagen hat,  erzeugt.  eindeutig Amerikanisch) the nose on every single landing,  gluing on the lower wing (and while you’re at it the upper one too) reduces the tendency of the fuselage to bend up right at the intersection of the nose to the lower wing.  Gemini 1 also developed a bend at the rear of the lower wing,  levered down from the fuselage bending away at the attachment bolt.  We figured,  might as well glue the wings on as the assembled Gemini easily fits on the back seat for the drive to the flying place.  For Gemini 3 and 4 the internal wing mount foam and plastic mounts (the part the screws go into) were cut out and the entire “tunnel” fiberglassed.  That not only stiffens and (less important strengthened) the fuselage up some,  it also allows positioning the battery further reward too.


Battery Placement


We like the (35) mm diameter size motors in our Geminis.  If you stick with the HiMax (80) gram motor at (30)mm or similar,  the wires from outrunner motors have enough room to get past without dragging on the rotating case.  When you go to the (35) mm size,  we ground out a trough in the nose for clearance.   There was enough fiberglass to make up for any loss in strength anyway.  As an alternate run the motor wires forward and then back through the chin duct.


Gemini 1 all the fiberglass was held on with water base hardwood flooring paint,  it was good enough,  and a whole lot easier to work with then two part fiberglass resin(s).  Gemini 2,  3 and 4 were mostly done with two component resin,  at least there were strength and rigidity was a major factor.  But waiting for it to harden is a bore,  for the finish coats I used water based hardwood flooring paint,  which may be sanded to a decent finish and seals the foam from my greasy fingers.  A final layer of (4) ounce per square yard fiberglass (as light as it gets) went over the fuselage from just rearward of the cowl forward and the whole underside of the lower wing.


It took six applications of epoxy on fiberglass,  wait,  and sand to finish each of Geminis 3 and 4,  at two hours a cycle,  with the final coat being fiberglass held on with the much easier to sand (water based) hardwood flooring paint.  Which added about twenty bucks to the price,  of which the most expensive component was sandpaper,  which also took up the most time.  Why so many cycles?  Many of what to be hardened/strengthened was compound curves,  which require really thing fiberglass cloth and you can only do a limited amount at a time.  


If I were asked to assemble another Gemini for an average pilot,  landing on more reasonable runways then we do (and not much wind,  i.e. Mission Bay) with (experienced based) expectations of more like fifty to a hundred flights (before a crash sets the statistical limit on the airframe) I’d likely do just two fiberglass cycles.  And then only on request.  But I’d refuse to touch a foam airplane unless the person commissioning the work agreed to taping the leading edges and lower wing tips.


Speaking of commissions;  For a while I had a well made (and expensive) carbon fiber motor plate,  interchangeable between most Multiplex airframes,  which,  uniquely,  included most of the needed down thrust built in.  Mine seemed to have turned into that warping white plastic overnight while I spent the night at the flying field in my camper.  I woke up that morning way late,  alone,  my head hurt,  my feet stank and at first I couldn’t focus.  When the fog inside lifted I wondered how six bottles of good quality local wine wound up on the floor,  two of them empty.  Well,  the flying field is on a hill surrounded by vineyards with a nice view of the Rhine and far back enough to be a quiet place to spend the night as the lights of the villages across the river come on when it’s too dark to see to fly.  A year later the motor plates were unavailable.


However,  while the two component fiberglass epoxy stuff in Germany has very low viscosity,  and at (45) minutes longer pot life then the twenty minutes for what Hobby People sells in California (pot life,  i.e. the time you can paint the stuff on after mixing,  keep going at your own peril as it won’t be as strong and takes days to harden up),  the German stuff (good enough for building small airplanes the pilot can sit in) is an expletative deleted to sand.  The USA stuff,  while higher viscosity (makes following compound curves harder) can be easily sanded to even feather edged.  But probably the most important difference is that at room temperatures (warmer then my basement workshop) the German stuff needs at least twenty to thirty hours to harden up,  the American stuff four to six.  Next trip Wiesbaden/San Diego/ Wiesbaden I’m bringing the USA resin back for myself and the other club members.  Both start out opaque (not quite transparent),  the USA stuff quickly yellows,  the German stuff shrinks.  


Taking off on the landing gear from a hard surface,  and the scale like “this is a model of a real airplane” (although there is no man carrying exact match) even after we went to folding propellers in the ten inch range,  requires down elevator for the first few yards,  then let it go neutral and the Gemini will rise off ground just fine.  However,  it is prone to prop strikes on takeoff,  initially half the price of the kits went into propellers.  It took a few tries to decide on the right,  for us,  prop size(s).


That from Multiplex wire landing gear is only just good enough even flying off of a hard runway.  Expect to have to hand straighten the main landing gear often (not necessarily a bad thing,  that springiness protects the rest of the airplane) and that the wire holding the rear wheel isn’t all that durable.  The Gemini is a great flier,  at a nominal purchase price,  so if you have to add a better landing gear,  it’s still a great flier and inexpensive.  We used the landing gear as an “arrestor hook” over the long grass we have to land on,  at which the stock mounting plate tore out at twenty flights.  The long term solution was removal of the hard plastic stock wire mount for a replacement carbon fiber landing gear fiberglassed on.  While you are at it,  bigger diameter wheels make for taxing easier.  A simplest solution for the too weak main landing gear is to bend up an additional wire and tape it to the original.


That simple wire for the rear tail wheel won’t last all that long.  Back when digital airborne movie cameras were new and novel (about fifteen to twelve years ago),  one SEFSD pilot mounted one facing rearward on a Telemaster,  we were surprised how much impact that little rear wheel takes.   From a recent conversion of nearly identical Fun Cub to a more durable rear wheel assembly figure twenty bucks for materials and two or three hours of modeling,  worth it,  but not until the original gives out.  Gemini or Fun Cub,  when running a (135) gram motor you need some additional weight back there anyway,  if you didn’t clearance inside the fuselage to move the battery further rearwards then Multiplex intended for the original (80) gram motor setup.  


The nose had to be beefed up as Gemini I landings were;  First hung the main landing gear up in the grass,  then bonk the nose down every single landing.  We always giggled about that as RC airplanes aren’t expected to withstand that,  and it looks,  after elegant flying,  amusingly clumsy.  Gemini II the lower wings leading edge,  the chin air inlet,  the nose,  and the servo arms hanging down took the landing skid.  Gemini I did rather well with (4) ounce per square yard fiberglass cloth (figure six layers) held on with hardwood flooring paint.  Gemini 2 had the same in fiberglass.  Geminis 3 and 4 were treated inside at the servo/air outlet areas and again at the outside there as if the lower wing snags on landing that’s where the fuselage breaks.




We like folding propellers,  easily available in Germany,  for which the nose of all four of our Gemini’s were cut back a bit to let them fold.  Only at the bottom,  it’s the “chin” which takes the most landing hits,  are the blades expected to really fold flat.  If at the sides and top the blades only fold to forty five degrees the slight reduction (at Gemini glide speeds ) doesn’t justify the large increase in modeling to get them to fold flatter.  These aren’t F5D airplanes!  The Elapor block,  cut from the wing mounts when the inside was “tunneled” was used to block off the useless air inlet.  Six multiple coats over that part of the nose and two over the black of the block.  Taper the layers as you don’t want an abrupt change in rigidity as that creates a breaking place.  The trick is to balance the “bash plate” nose to transition into the flexible Elapor,  I can’t set it to words.  Even if the Editor or I,  both educated as engineers,  could put numbers to it (it would take finite element analysis,  beyond a basic undergraduate degree,  and detailed knowledge of the foam layout),  even most engineers couldn’t read it…  Try spacing the ends of the fiberglass three quarters of an inch (two centimeter) short of each other.


Although even just glued on the control horns don’t tear out,  in particular at the ailerons the foam where it has to actuate the upper aileron develops a warp.  I fiberglassed the inboard three inches of the lower ailerons (the one with the servos) of ours.


That chin air inlet is a relic of a while back,  it no longer serves any purpose.  Unless you consider collecting grass,  mud and,  where we land,  horse droppings,  useful.  Mole hills are the usual cause.  With the electrical stuff of the last five years,  according to our fingers all that get more then warm is the motors.  Motor controllers and batteries don’t heat up the way they did even a few years ago.


The rudder doesn’t get much use in a Gemini,  if you can’t afford everything you like,  a less expensive servo (with it’s hunt for center response,  sloppier gear train and shorter service life) there might be fine.  I had an a Nano Karbonite (HS-65) in Gemini 2’s rudder,  which was cut out and installed in Gemini 4.  Since I wanted to keep flying Gemini 2,  the replacement was a physically larger Hobby People House Brand servo ((16) grams,  $11- as of 2014),  which did fine.  The bigger external size took up the foam lost cutting out the smaller servo,  the was just enough foam left.


There is an abrupt change of rigidity of the lower wing right at the servos (only the upper wing has a spar)  to length the service life (12) ounce per square yard four inch diameter circles of fiberglass were applied top and bottom.  Gemini1,  where the lower wing seldom touched the ground that wasn’t necessary.  Gemini 2,  where one wing tore off and the each side of the fuselage split right at the fuselage servos,  it was.   Geminis 3 and 4 the whole bottom of the lower wing was treated with a layer of (4) ounce per square yard fiberglass over the whole wing with three additional layers over the middle extending to the fuselage in front of and behind the lower wing with a four inch patch of fiberglass over the top too.  


Other then taping the leading edge the top wing doesn’t seem to need any improvements at all.  Oh it gets some scrapes from transport (soft and light foam ya know) but other than that time I just hit a branch the upper wings did just fine.  If you go back to the first SESFD Gemini report of mine I substituted a carbon spar for the stock fiberglass one to save (15) grams.  In between,  although fifteen grams does make some difference,  it wasn’t enough to justify riding my bicycle to one of the few remaining hobby shops (my high roof camper wouldn’t fit in any parking structure) and then taking the bus back up the steep hill (two hours round trip plus three dollars for the bus) at the moment I was ready to start building.


It’s not enough that the horizontal foam of the wing struts is glued,  you have to flex the wigs and get some glue in all the way around too.  Geminis 3 and 4 contact cement was used for the flat,  then the wings bend and CA flown in.


There just wasn’t much to be done to improve the wings,  upper and lower,  out past the wing struts.  Make one of these inherently flexible foam airframes too strong and rigid and they weight too much.  With that neat scale top surface,  that looks just about like fabric over the ribs of a “real” one,  things like inletting carbon fiber rods or strips were rejected.  I figure an additional fifteen hours over minimal assembly as reasonable for the Gemini,  when its used up,  assemble another something.  If you need a more precise airframe,  get a different airplane.  Want something spiffy,  get something else.


I’ve got to admit it though,  I really enjoyed the better controllability at the elevator of quality servos and wished I had them in the wings too.  But,  being almost,  but not quite,  completely broke at the end of every month…


The pair of unbuilt Gemini kits we found in E-Bay late 2014 were complete and in fine shape,  except that the glue of the decals had aged.  The first attempt to apply them with a covering iron resulted in a mess of little swollen pustules,  almost all of them in the gray coloring or the clear portions.  After which they were applied by hand,  just touching down and then pulling them back up until things aligned well and then gone over with a standard hair dryer,  pressing down with a soft cloth.  As part of making Gemini 1,  after three years of service,  re-fit for a few more flights,  those details applied new were holding on rather well,  


Things are looking up,  they are finally building more in Germany,  and maybe I can get life in San Diego going again.


Reasonable Quality doesn’t Overheat
One reason I take the time to put this down in print;  If you find your stuff overheating,  and you didn’t block off the front and rear air flow,  the problem is your equipment.  


Quit buying cheap stuff!  You may use the price of HiMax motors (standard for Multiplex until recently) or the Multiplex house brand Permax as a reference for the minimum to spend on motors and controllers.  Use HiTec and Multiplex as the standard for servos too.  


Repeatedly when following Internet Forum threads the problems are using servos and motors that barely function,  and often not even that.  That may partly be attributed to the censorship by omission of motor efficiencies in a vast majority of reports,  lack of commentary on imprecise servos which don’t fail during the first five flights (they didn’t even try to compare different ones!),  or even mention that the servos need to be individually centered to their control surface(s) and a general lack of reporting on the tuning as an adventure to getting your stuff to really fly right.  


The other pilots new Gemini 3 was fitted with four HS-65s,  bought new from current production,  of which only two were centered to each other as a reversed pair (for the ailerons) and none them of which were at right angles to the case.  The Multiplex Brand servos have been better in that respect,  as is generally the case as the price of the servo increases.  That’s weird,  pay more,  and get better performing stuff.  If you want the airplane to fly really right,  any airplane,  you need a computer radio to individually center the servos for equal throws at the control surfaces,  or the patience and luck to at least get the aileron servos to match each other.  At least with the SESFD we have competition pilots setting a good example,  who,  if asked,  will be glad to clue you in.


I was staunchly old school in not taking full advantage of the improvements in transmitters until a semi-pro sponsored pilot asked if he could fly my airplane.  His first comment was how did I fly so well without using exponential,  although he is the better pilot.  What we need is for small stick movements about where the control sticks are centered to give small control surface changes of angle,  with increasing large movements as we go to the extremes.  Due to mechanical geometry,  that bit about proporti