Diversity Model Aircraft Butterfly

Daniel Griscom shows how DMA's very light Speed 400 aerobat builds easily and flies very capably.

Article Thumbnail

ButterflyOnGrass.jpg (12666 bytes)

  • Wingspan: 31 in
  • Wing area: 300 in2
  • Wing loading: 6.7 oz/ft2
  • Airfoil: SD8020
  • Weight: 14 oz
  • Controls: elevator, ailerons, throttle
  • Motor: Speed 400 7.2v
  • Battery: 7-cell 600AE (not included)
  • Speed control: 10 Amp minimum with BEC (not included)
  • Kit price: $74.95
  • Manufacturer: Diversity Model Aircraft


The Diversity Model Aircraft Butterfly is an aerobatic 7 cell Speed 400 plane. The laser-cut kit goes together quickly, looks great, and flies well, but it isn't for beginning builders or fliers. Building time is about 30-40 hours.

Kit contents

LaserSheets.jpg (30983 bytes)  OtherParts.jpg (18747 bytes)  Instructions.jpg (19383 bytes)

  • Laser-cut plywood and sheet balsa parts
  • Carbon fiber and balsa spars
  • Motor, prop and adapter
  • All hardware
  • Full size wing plan
  • Step by step instructions

Power package

DMA sells a power package for the Butterfly, consisting of a 7-cell Sanyo 600AE battery pack and an Aveox A-15 speed controller with BEC, complete with your choice of connectors. Note that the A-15's BEC is only specified to drive two servos, so if you want a rudder you'll need a different control. Also note that the Butterfly's battery compartment is a snug fit for this size battery, so be careful if you shop for a replacement. The price for the power package is $64.95; I bought one (with Deans Ultra connectors), plus a spare battery.

Tools you'll need

  • Flat building surface
  • #11 knife
  • Sandpapers, sanding blocks
  • Ruler, triangle, pencil
  • Razor saw
  • Small clamps
  • Mini Phillips screwdriver (for servos)
  • Covering iron, heat gun
  • Teenie hex wrench (0.050"?) for wheel collars

Materials you'll need

The instructions don't include a clear and complete list of parts and supplies needed to complete the kit (e.g. the quantity and suggested type of covering film, electrical tape, etc.), so here's my list:

  • Thin and thick CA
  • (2) micro or sub-micro servos (I used Hitec HS-55s; DMA also recommends the Cirrus C-10)
  • Micro receiver (I used a Hitec 555 with the case removed; DMA also recommends the Berg 6 or the Super Slim)
  • Electrical tape
  • Masking tape
  • Double-faced tape
  • Extra 16-gage stranded wire
  • 1 roll light covering, transparent so you can see the cool ribs. DMA recommends Oracover Lite; I used 2/3 roll.
  • Hinge tape (Scotch #845 Book Tape cut to 1" wide works well)
  • weight for balancing

Materials needed for options

  • Extra heat-shrink tubing (for joining aileron pushrods)
  • Balsa scraps for enhancing wing LE jig blocks, setting aileron hinge gap
  • If desired, parts for rudder (servo, linkage, control horn, hinge)


This is the first laser-cut kit I've bought, and I was very happy with the quality. There were only a couple of instances where I had to adjust the fit of a part; all others went together like a well-built jigsaw puzzle.

You can see the precision in the pictures below, showing each side of one of the ubiquitous butterfly cutouts. Each butterfly is 1.5" from wingtip to wingtip, and there are details far smaller than any die could cut.

ButterflyBottom.jpg (13577 bytes)  ButterflyTop.jpg (15292 bytes)

If you look carefully at the butterfly images, you'll see that the kerf looks wider on one side than the other. This is because the kerf edges are angled due to the laser's cutting action. The only place where this was a problem was when building the vertical fin; when the two pieces are glued together they tend to make an angle at the joint, rather than assembling flat.

Wing and tail

The wing builds quickly. Be careful of CA vapors, since you'll use a lot of it in a short time.

EnhancedLEJig.jpg (9816 bytes)  WingInJig.jpg (14017 bytes)

The wing is built with the CF leading edge rod lifted by two wing jig blocks. These are sized so that the bottom rear of the ribs are flush with the plan, making alignment easier. The instructions say to tape the leading edge to the blocks, but I found this difficult because you have to remove the wing several times during construction, and it takes time to align each time you retape. To work around this, I CA'd two 1/4" square sticks to the top of the jig blocks, holding the leading edge property aligned. While gluing them down I pressed them against the leading edge, so that the leading edge ended up being held by a friction fit. This made realigning the wing quick and easy.

In each of two Butterfly kits, I found that the carbon-fiber spars and leading edge were 1/8" too short. To work around this, I had to move each of the outer ribs 1/16" towards the center.

Each rib includes a beautiful butterfly cutout, but if you remove the butterfly centers before building (as the instructions instruct) you'll almost certainly break a number of the delicate strips of balsa between the wing bottoms. To work around this, I cut the retention tabs holding the centers in place, and then held the centers on with masking tape (with the ends folded over for easy removal). Near the end of construction, right before covering the wing, I pulled off the tape and the butterfly centers. Final score: only one strip missing.

The ailerons are built as part of the wing, and only cut off at the end of the process. It's difficult to keep CA out of the aileron cut lines in the R-3 and R-5 ribs, especially while attaching the small wing tip gussets. I ended up removing the ailerons cutting the ribs with a razor saw, and then sanding the wing and ailerons down to the cut lines.

The holes in the aileron torque tube supports were slightly too small for the torque tubes; I had to enlarge them slightly so the torque tube could move freely.

WingOnPlan.jpg (14825 bytes)

The small wing tip gussets need to be beveled before attaching, and it may be easier to attach them after the ailerons are removed from the wing (which in turn may make it easier to keep CA out of the R-5 cut lines).

Tail.jpg (14943 bytes)

If you're not using a rudder, you'll need to glue the two pieces of the vertical fin together. The taper on the edge caused by laser cutting tends to push the joint into an angle, so you might want to sand the edges square before gluing.

Bevel the front tab of the elevator horn to match the bevel at the front edge of the elevator.

Servos and linkages

Servos.jpg (15126 bytes)

I needed to space the servos away from R-1 ribs using sheet balsa. To do this, I used thin double-faced tape to glue the servos to squares of 1/16" sheet balsa, clamped the assemblies to the R-1 ribs, carefully aligned them, and then glued the balsa squares to the ribs. This let me align carefully before committing. For extra stiffness, I dripped thick CA across the foam tape.

Be careful when installing the aileron torque tubes; they're different lengths, and it's easy to confuse them. When gluing the aileron horns and ply rings to the torque tubes, use slow CA and kicker so as not to get glue where it shouldn't be.

I didn't like the plan's method for making the aileron links adjustable (putting "V" bends in the middle): I found them hard to adjust, and they added more springiness to the links. Instead, I used the method described for the elevator link: fabricate link in two pieces with overlap in middle, connect with 1/16" heat shrink tubing, and then lock with CA. Make sure that both wires extend all the way through the tubing, or the CA won't be able to wick in.

The instructions mention the possibility of adding a rudder, but don't include any details. I think it would be somewhat difficult to do; there's little room for another servo, and routing the linkage would be tough.

The instructions specify the necessary travel for the aileron linkages, but don't specify the resulting aileron throws. I had good success with plus or minus 1/2" of throw.

Receiver, speed control, servo wiring

I found this to be the most difficult part of the kit to figure out, since it involved so many variables. I'll describe the arrangement I ended up using; it may not be the best, but it did work.

I used a Hitec 555 receiver with the case removed. I foam-taped it to the outside of the right R-1, between the two CF spars.

I mounted the Aveox A-15 speed control to the outside of the left R-1, with the battery wires facing the rear. The battery wires were routed through the circular hole in R-1, ending in a Deans Ultra jack. The motor wires were routed through the butterfly cutout in the front of R-1 (be careful of the strip between the wing bottoms!) through the hole under the fuselage rod, where they were soldered to stranded 16-gage extensions, long enough to reach the motor. I mounted the speed control enable switch in the balsa bottom front, and routed the receiver cable through the circular holes and across the bulkhead.

TopCenterNoBatt.jpg (18070 bytes)  BottomCenterBatt.jpg (11310 bytes)

To install the battery, connect its jack to the speed control plug, and then slide it along the right R-1, keeping the connector between it and the left R-1. To make removal easier, add a loop of electrical tape around the front of the battery, with a tab on the top that you can pull on.

Another option is to mount the speed control in front of the leading edge, fastened to the fuselage tube. If you do this, be careful not to move the CG too far forward.

I made a couple of changes to the battery hatch. First, the plans say to drill a 1/16" hole to fit around the CF spars; you'll have to make it a bit larger. Second, I found that the bottom of the hatch doesn't hold onto the leading edge tube; heating and bending the bottom edge to fit the contour of the tube helped keep the hatch closed.

Motor, landing gear, fuselage tube

It's difficult to get the wheel hubs centered correctly. My procedure was to thread a hub, a wheel, and a hub onto the landing gear wire, and then tape the hubs to the wheel. I then adjusted the hubs until the assembly rotated without wobble, tacked the hubs in place with CA, removed the wheels from the gear wire, and then thoroughly CA'd the hubs in place.

When gluing the landing gear to the motor mounts, only glue to the front mount. This allows the gear to rotate in the back mount, increasing the springiness.

CenterView.jpg (9686 bytes)  FrontAndHatchOpen.jpg (15974 bytes)

The plan specifies that the front of the fuselage tube protrudes by 4.25", but that you can adjust it between 4" and 4.75". I found that a 4.625" overhang gave me a CG that was 3/16" in front of the CF spars, which is right in the middle of the specified range. If in doubt, go for too much overhang, since it takes much less weight on the tail to adjust the CG than on the nose.

Most experts say that you should break in motors such as the Speed 400 used in the Butterfly, but I didn't. Next time.


The instructions say to be careful not to over-shrink the covering. Guess what: they aren't kidding. I shrank the covering when the ribs weren't securely stuck to the covering, and cracked several of the ribs (see picture below). I fixed each rib by poking a thin piece of music wire through the covering and then, while using the wire to hold the rib in place, thoroughly ironing the covering to the rib

CrackedRibs.jpg (15305 bytes)

Another bad place is the vertical section of the wing that faces the ailerons. Shrink this too enthusiastically and you'll get the most wonderful bends in your balsa spars.

In order to make it easier to tell the wing top from bottom at a distance, I added a 6" wide by 3" tall white butterfly silhouette to the top left wing surface (scanned and enlarged from a rib butterfly cutout, and cut from white Oracover). Although you can see the butterfly through the transparent film from the bottom, the four thicknesses of film darkens it somewhat. When flying, though, it was difficult to see at all at a distance; if I did it again I'd have larger butterflies on top of both sides of the wing (although too large and it would obscure the wing's pretty internal structure).

ButterflySilhouette.gif (2985 bytes)

The completed plane weighs 14.0 oz with battery. The DMA web site and instructions list a target weight of 12.5 oz, but the two other finished Butterfly weights I've seen were also 14 oz, so I'm not too upset.

Carrying case

I was quite worried about hanger rash, since the wing and tail aren't made to survive being bounced around a trunk. So, I made a carrying case out of a large piece of double-thickness corrugated cardboard. The plane fits in upside down, and is held into two notches by crisscrossed string. About an hour of work, and now I can transport and store the plane much more safely.

CarryCase.jpg (9249 bytes)  CarryCaseTiedown.jpg (10220 bytes)


I discovered that with the power on and the throttle set to anything above zero, the servos (throttle included) twitch every 3/4 second. My guess is that the speed control is interfering with the uncased receiver. The effect doesn't increase with distance from the transmitter, and doesn't seem to make a difference in flight.

Since the Butterfly has no rudder, and the wheels are ahead of the center of gravity, taking off from the ground can result in the plane swerving uncontrollably (the beginnings of a ground loop). Taking off directly into the wind helps. Short, thick grass seems also to help (adding drag to the tail?).

It needed quite a bit of right trim in order to fly level. After landing, I found that this trim moved the ailerons 1/8" away from their centered positions. I must have built a twist into the wing, although I was pretty careful and it looks straight to me. Once trimmed, it tracked very nicely. It's got enough power to do loops until you get bored. The climb at full throttle was about 30 (it might have been better with a broken-in motor). Stalls are fairly well-behaved. It can slow down nicely for landing, but it's definitely too fast for most indoor flying sites. Although I didn't try it myself, Radio Carbon Art's "Electric Airshow" has a segment with the Butterfly doing extensive acrobatics. Once I'm more confident...

The plane is extremely quiet even at full throttle, because there are no surfaces near the prop. It's nice to be quiet, but then again it's nice to hear when the motor quits. Landings were easy and smooth, perhaps because of ground effect on the large wing.

At a distance, it's difficult to tell the top from the bottom, especially with the transparent covering. The butterfly helped, but probably the best indication is the vertical fin, which shows up fairly well.

To get an idea of flight time, I flew two flights fairly conservatively and timed the results. Both flights started from the ground, and lasted until the battery was completely dead and the plane was back on the ground. The first flight lasted 9 minutes 20 seconds, and in the second I caught a couple of thermals and stretched it to 12 minutes 20 seconds. Not bad for an aerobatic electric plane.


There are a few possible enhancements for the Butterfly. The easiest would be to switch from NiCad to NiMH batteries. 1000mAh NiMH cells are the same size as 600mAh cells, so theoretically you could get an extra 50% flying time. Unfortunately, NiMH cells don't source as much current as equivalent NiCad cells, so you'll sacrifice some performance.

A oft-cited way of compensating for the current reduction is to switch from a 7-cell NiCad battery to an 8-cell NiMH battery, but there's not enough room in the battery compartment for any more cells. I'm told that some people move the rear bulkhead back in order to make room for an additional cell, but that reduces the already tight space for the servos and shifts the CG even further back. I've bought a 7-cell 1000mAh NiMH battery pack, and I'll see how it does.

A second enhancement would be to include a rudder. This is mentioned in the instructions, but there are almost no details. Finding a place for the servo, and a good way to route the linkage, would take some thinking.

A final enhancement would be to switch from a 7.2V motor to a 6.0V motor. This would draw more current, and thus produce more power and better performance, at the expense of flight time.


I really enjoyed building and flying my second Butterfly. "Second Butterfly?" I hear you cry. Well, I lawn-darted my first one during its maiden flight. It was a somewhat gusty day, and I was having trouble controlling it, when suddenly SPLAT. My best post mortem guess is that the wind tipped the plane on its side, and when I thought I leveled it I actually ended up inverted. I then lost a bit of altitude, and pulled back on the stick to compensate. There wasn't much left.

In the spirit of getting right back on the horse, I bought and built another kit. It took about six months to make time to build it (adding the butterfly decal for orientation help), and another four months to get up the gumption to try flying it. The first few flights were nerve-wracking, but after three or four flights I started to relax and enjoy. In truth, this plane is meant for intermediate and above fliers, so perhaps it wasn't the right plane for me at the time.

While building the two Butterflies, and later while writing this review, I called or emailed DMA's owner, Steve Belknap, a number of times. He was always very responsive and helpful.

All in all, a very nice kit and plane.

ButterflyOnRug.jpg (8519 bytes)


Thread Tools

Quick Reply

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Steve at Diversity model aircraft flyinjim Vendor Talk 13 Feb 06, 2018 04:37 PM
For Sale RARE Diversity Model Aircraft Butterfly Aerobatic Parkflyer kit $60.00 dawnron1 Aircraft - Electric - Airplanes (FS/W) 5 Dec 01, 2005 07:44 AM
Diversity Model Aircraft/ Steve Belknap jimmy6819 Vendor Talk 1 Sep 24, 2002 08:56 AM
WTB Diversity Model Aircraft Skat pylon racer smsgill Aircraft - Electric - Airplanes (FS/W) 0 Jul 27, 2002 12:59 PM
Diversity Model Aircraft MaxFx Vendor Talk 2 Oct 03, 2001 08:40 PM