Stevens Aeromodel 39" SHAFT (Super High-performance Aerobatic Flight Trainer) 400

Introduced to the world at the 2007 NEAT Fair was Stevens Aero's answer to the question, "what comes after my aileron trainer?"

Going straight up


Seemingly out of the blue in early September 2007, Bill Stevens (head honcho at Stevens Aeromodel in Colorado Springs) popped into the Parkflyer forum here on RC Groups with a thread entitled “Get SHAFTed by Stevens AeroModel NEAT 2007”, and then proceeded to introduce to us an elegant and light variation on the “Ugly Stick” theme designed to fill the gap in the SA product line between the stik and his all out aerobatic airplanes such as the Stella, the Re-Groove and the various scale aerobats such as the CAP 232s and Edge 540. He dubbed it the SHAFT: Super High-performance Aerobatic Flight Trainer.

Preproduction kits were available to the lucky folks who attended the 2007 NEAT Fair in New York, and production kits were released a week later. My airplane is from that first production batch.

Starting with the simple wing on top of a box fuselage concept, Bill added style and flair and his signature strong but very lightweight structural design. He also gave it rugged and resilient landing gear based on the design used on the SQuiRT series of basic trainers. This appealed to me as I LOVE to shoot touch-and-go landings. Many of my planes have disappointed me in the landing gear durability department but my SQuiRT certainly hasn’t.

He designed it from the beginning to have a removable horizontal stabilizer, making it easy to pack flat for travel. I modify many of my airplanes this way, which is sometimes easy and sometimes not so easy. With the SHAFT I didn’t have to modify it at all.

Because of the light and strong structure (it’s about 1/3 less than a very popular ARF that’s about the same size), the SHAFT can perform very well with moderate amounts of power. I have found that it will take off and climb readily, loop (both inside and outside), roll and do maneuvers that are combinations of these on only 75W (verified with in-flight logging). But if you put 150W or more of quality brushless power on the nose, hang on!

Here, I’ll take a close look at the kit, and then tell you how it flies on several power systems, from moderate to wild.

Wing Area:340 sq. in.
Empty Weight:9.0 oz. with no motor, ESC or battery, but with radio system installed
Flying Weight:17.2 to 20.5 oz. depending on motor/battery/ESC combination
Wing Loading:7.3 to 8.7 oz/sq. ft. as flown
Servos:four required: rudder, elevator, two for ailerons. I used Hitec HS-55s
Transmitter:Multiplex Royal Evo 9 Synth
Receiver:A minimum of four channels required. I used a Castle/Berg 4L
Battery:various 2s and 3s lipoly from 910 to 1650 mAh – see sidebars
Motor:several – see text and sidebars
ESC:several – see text and sidebars
Manufacturer: Stevens Aeromodel in Colorado Springs
Available From: the manufacturer

Kit Contents

The kit includes everything you will need to build the airplane except covering, adhesives and the electronics.

Kit Includes:

  • 12 sheets of precisely laser cut balsa and plywood parts
  • Wire for pushrods, DuBro EZ connectors, control horns, EZ links
  • Pre-bent landing gear and 2 1/4 inch EFlite wheels, retainers
  • Carbon fiber strips that are used to reinforce the control surfaces
  • Even your first couple of sets of rubber bands for wing and gear mounting
  • Full sized plans
  • A 33-page step-by-step illustrated manual and a separate layout showing each of the laser-cut part sheets.

That manual includes six pages of excerpts on precision aerobatics from Scott Stoops’ excellent book “Mastering Radio Controlled Flight”. This book, by the way, is an excellent reference for all phases of RC flight.

All of this is packed in a flat, bagged package, with a thick piece of stiffening cardboard between the sheet parts and the rest of the contents. This keeps the hardware and wire from denting the wood in transit. This package is then boxed for mailing. My kit came via Priority Mail and arrived completely unscathed.

Kit Requires:

  • A minimum of a four channel transmitter, or at least five if you want separate control of each aileron for differential or flaperons.
  • A matching receiver. An end-pin configuration will fit more easily. I used a Castle Creations Berg-4L.
  • Four micro servos. Suggested (and used for this review) are four Hitec HS-55 servos for rudder, elevator and one for each aileron.
  • A couple of 6-inch servo extensions for the fuselage servos and a short Y-harness for the ailerons. Note that Hitec servos have longer leads on them than most of their competitors, so if you’re thinking of using something else you might need 12-inch extensions instead, and it might take some creativity to connect the Y harness to the servos in the wing – such as preinstalling it before covering the wing.
  • A 75-200W power system (motor, ESC, battery). I used three combinations (so far) - see the sidebars.
  • At least one roll of lightweight covering (such as Solarfilm or Solarfilm Lite)
  • Adhesives. Thin CA is used for most joints (after they are assembled), but there are some situations where medium or thick would be in order. No epoxy is needed.

There is a big warning at the top of the manual about using no more than 200W of power on this airplane. It is so light and lightly loaded that it will perform well on 100W and be nearly “ballistic” on 175W so there’s no good reason not to heed this warning. The simplest way to do this is to use a system that draws between 10 and 15A on a good 3s lithium battery while swinging an 8 inch or larger prop (there’s enough ground clearance for at least a 10 incher). Of course one could also use 2s and run higher currents or even 4s on a really low Kv motor I expect. Using a nickel battery would be a little more difficult since the battery compartment is shaped and sized for a flat li-poly pack. The extra weight would also detract from the airplane’s sparkling performance.

The SHAFT has a flat “firewall” that is best suited to rear mounting of an outrunner brushless motor. The clearance hole for the rearward-projecting shaft and collar of such motors is already in the firewall, though you may have to enlarge it a bit for some motors. Also laser cut into the firewall is the slot for passing the motor leads back to the ESC. This slot is big enough that most suitable ESCs can be fed through it as well when hooking things up, making the lack of a nose area hatch on the airplane a non-issue.

Pilot holes are laser cut for a cross-shaped mount for motors such as the recommended Hacker A20Ls, the Eflite Park 450, the AXi 22xx series, Atlas 23xx series, Hyperion Z22xx series, HiMax 28xx outrunners, CommonSense RC E5 series, etc. Bill recommends attaching the motor mount with #2 sheet metal screws. I glued some scrap ply on the aft side of the firewall to add a bit of thickness and then tapped the holes for 4-40 threads so I could use short 4-40 nylon bolts to provide a bit of a breakaway feature in case I do something stupid with the airplane too close to the ground.

One could also use a geared inrunner (perhaps something like a Castle CM-2042 or 2054 in a Cobri gearbox) or even a lowly GWS EPS-350C brushed motor with either the Eflite or Cobri firewall mounts if desired. You’d have to relieve the slot in the firewall to clear the back of the motor. Another alternative if you use a GWS EPS gearbox would be the Lighter RC carbon and fiberglass bulkhead mount for the EPS box .

Building It

Stevens Aeromodel kits are renowned for the precision of cut and careful selection of the right grade of wood for the right use. The SHAFT continues this tradition. As is typical with an SA build, if you simply follow the manual everything goes together nicely. By doing so myself I found nothing that needed to be changed or done in a different order and only a couple of small errors that, I am sure, have since been corrected. If you want to get a very good idea of how the airplane goes together, the manual, less the layouts of the laser-cut sheets and the Scott Stoops book excerpt, is posted on the SA site HERE.

For another view of how it goes together, there is also a comprehensive build thread by Bob McGee here.

Parts fit is almost without exception very nice, just a bit firm. Because of the interlocking features in all the assemblies you can dry-assemble most any of the major components and they will stay together without a drop of adhesive being applied. In the build of the SHAFT there are several places where you are explicitly told NOT to bond things until several assembly steps are completed. Please pay attention to this.

One small but significant detail that applies to this airplane (or really any laser-cut kit build): There are always the little bumps on laser-cut parts that are left behind from the cutting/breaking of the “retention gaps” in the sheets. Whenever these appear on part edges where that part bonds to another part, these bumps should be gently removed so that the mating parts may seat fully against one another with no gaps. This is especially critical for the wing spar web-to-lower spar cap joint. There are a few other places in the SHAFT where this situation occurs as well. Just bring the bump down to the surrounding edge of the part with a couple of swipes of a fine sanding block before assembly.

When I build a review model I make notes on anything that I need to comment on in the manual as I go. For the SHAFT the notes are remarkably few. Here, more or less in construction order, are some comments:

There is only one shaped part that isn't keyed into place somehow - the plywood wing spar web doubler. Take some care to locate it properly on the spar web.

Next, I suggest using some scraps of 1/8 inch thick balsa from the sheet that contains the tail surface parts to help keep the aft end of the fuselage aligned at fuselage assembly step 14. Put one scrap in the slot in the top fuselage sheet where the fin will later be installed and put another down flat on the board and rest the stabilizer platform on it. This will help hold things square as you do the final bonding of the assembled fuselage.

The total assembly job, including stopping to do the tapping of the motor mount holes I mentioned, took about eight hours. I went slowly and carefully and took photos as I went, so it could be done more quickly.

The structure sanded and ready to cover weighed only 4.6 ounces. The landing gear, including the tailskid assembly, adds another 1.0 ounces.

I covered my airplane in transparent orange and cream AeroFilm, or at least that was the plan. As it turned out I used AeroLite for the cream-colored portions of the wing, the elevators and the rudder because I pulled the wrong roll out of my covering stock. The cream on the fuselage and the ailerons is regular AeroFilm. Covered, the structure weighed 5.6 ounces – a weight gain of only one ounce even.

That said, depending on color choices (the fluorescents are the heaviest) you could add quite a bit more weight. Some care is therefore required, especially when choosing what the cover that nice long aft fuselage and tail feathers with.

Before you cover, you’ll save yourself some time and a bit of aggravation if you put some strings or threads in the wing and fuselage to help pull the servo leads through. You can get by without them, at least if you use transparent covering so you can see where the leads are as you try to feed them through, but in hindsight it would’ve been easier to have the pull strings than doing what I did.

As is my custom, I topped off my covering job with a custom cut vinyl “SHAFT” and my AMA number from Greg Judy at Vinyl Graphics by Greg.

Power System number 1 – Atlas 2312/26 and Atlas Black 20A speed control

Hobby Lobby supplied me with an Atlas 2312/26 and an Atlas Black 20A controller for this project. The motor looks and feels well made and runs smoothly and quietly. It has generous and flexible leads attached to it (in stark contrast to many of the “inexpensive” choices out there). It also comes with a complete set of hardware for rear mounting but no prop driver for using it front-mounted. The motor has a 3.17mm shaft (the same size as the AXi 2208, 2212 and 2217 series) so you’ll need a prop driver for that size if front-mounting it. This is not an issue with the SHAFT since the airplane is set up for rear mounting.

The motor weighs 2.1 ounces as supplied, and 2.5 ounces ready to mount in the airplane with the rear mount kit, prop driver and 3.5mm bullet connectors installed. The prop shaft on the rear mount adapter is 5mm in diameter, so you’ll need the metric centering rings for APC slowflyer props (or the neat aluminum ones from Hyperion).

A little static testing with a Kokam 3200 mAh 3s battery and some APC slowflyer props yielded the following performance figures for this motor (numbers midway through 10s at full power under the control of my Medusa Power Analyzer Pro). Note that with smaller batteries that will fit in the SHAFT these numbers will be somewhat lower:

Prop RPM at full throttle Power In
8x6 SF7950 150W
9x4.7 SF8450 130W
9x6 SF6625 180W
10x4.7 SF6775 175W

In flight temperatures ranged to less than 10 degrees above ambient, showing that the motor is not overworked when run on 3s and the 9x6SF – at least not on a plane as slippery as the SHAFT.

The Atlas Black 20A electronic speed control is rated for 5 to 10 nickel cells or 2-3 lithiums. It has overtemperature protection and options for timing, hard or soft low-voltage cutoff, brake on or off, two startup choices and a helicopter governor mode with three different RPM ranges. You can also reverse the motor rotation by programming as an alternative to swapping two motor leads. A jumper-based programming card is supplied with each Atlas ESC, making choosing among the available options fast and easy. This idea was introduced by Jeti with their Advance Plus range of controllers and has now spread quite widely, and with good reason: it’s the fastest and simplest way to set several options on the ESC and no computer is required. Just set the jumpers for the desired settings of each option, plug the ESC into the indicated spot on the card, apply power either via the motor leads or a separate receiver battery, and it’s done.

Various options can also be set with transmitter stick programming. Only one option can be set per power cycle. This makes it relatively easy to change a single option if that’s all you want to do. The supplied instruction sheet is devoted to how to do this. I first encountered this programming method, with its distinctive tone patterns for each of the options, with the Hyperion Titan 20A controller I’ve had for some time. Since the tone patterns and the options are very similar, I suspect that the Atlas controllers all come from the same manufacturer. Note that only the Hyperion version has the interface plug for direct programming via the Emeter.

The battery-eliminator circuit of the Atlas Black 20A is of moderate capability (based on the information on the Hobby-Lobby web site – no specs are provided with the unit), recommending only two servos on 10 nickel (or 3 lithium) cells. Since the SHAFT is a four-servo airplane, this bears watching and making sure the ESC has some cooling airflow (which is built into the SHAFT’s design). Also, fortunately, I am using Hitec HS-55s in the airplane, which aren’t as current hungry as some small servos. In-flight servo draw peaks at only 700 mA as reported by the handy little Dimension Engineering ServoSense device. I also got an in-flight log of servo draw with my EagleTree MicroPower data logger with the servo current expander. It confirmed peaks around 700 mA with averages well below half that. So….no issues with 4 HS-55s in the SHAFT on this ESC anyway. In flight temperatures (with a temperature sensor taped to the regulator side of the ESC, seldom got above 20 degrees above ambient.

The Atlas Black 20A ESC is about average in size and weight for similarly rated ESCs these days at 0.9 ounces out of the bag, and 1.0 ounces ready to use (with Anderson Powerpoles intalled on the battery leads). It comes with 3.5mm bullet sockets already installed on the motor-side leads and the mating bullets are also supplied for use on your motor. These I installed on the Atlas motor. There is an on/off slide switch attached The motor and battery leads are about 2 1⁄2 inches long, the receiver lead is 9 inches long and the switch lead is about 4 1⁄2 inches long.

On the test stand the ESC started the Atlas motor quickly and smoothly and ran it without a hiccup. There is simple set of arming tones when you power-up to let you know things are live. The low voltage cutoff (I have it set to soft – or “reduce power” on the programming card) engages properly to maintain the battery at or above 3V per cell (this also confirmed with in-flight logging). If the battery falls below that level and stays there, the motor is stopped altogether.

One negative, though: I don’t care for the throttle response of this ESC very much. It is very heavily damped and sluggish to respond which adds a little drama to maneuvers that require wide throttle excursions like proper looping figures, or even touch and go landings.

With that one comment – which is really a personal preference – I can recommend the Atlas Black motor and controller as a worthy way to power your SHAFT.

Final Assembly and Equipment Installation

Good quality DuBro hardware (horns, EZ connectors and EZ Links) is supplied in the kit along with 0.045 inch wire for making the pushrods. There’s no reason not to use the supplied hardware and materials, as they’re more than adequate for the job. For installation, again, you can’t go wrong just following the manual. I did run into a few very minor things at this stage, however.

The first, and most significant, is that the cutout in the bottom leading edge of the rudder to clear the elevator joiner on my airplane wasn’t big enough. It needed to be enlarged at least 1/16th of an inch both aft and up in order for the elevator to have free movement at the same time as the rudder is deflected much off of center. As I understand it Bill was going to modify the parts to eliminate this so by the time you get yours this will be likely already fixed.

Second, the hole in the basswood elevator joiner that receives one of the pins from the supplied DuBro control horns needs to be drilled out – there’s no way you’re going to just force the horn’s barbed pin through it like you can through balsa on the rudder, ailerons and the other hole in the elevator assembly. A 1/16th inch bit isn’t quite big enough, but you can sort of ream the hole with a bit that size until it’s large enough.

Speaking of the control horns, they also need to have their holes drilled out to fit the supplied pushrod wire. A scrap of pushrod material with the cut end left sharp will do the job for you. I almost wonder if the control horns in my kit were misßpackaged by DuBro, because they were labeled as suitable for 0.047 inch wire but the holes were more like 0.032 inch, though I don’t see another version with smaller holes listed on the DuBro web site.

Also, the aft fuselage holes for the servos are a very snug fit on HS-55s. They will go in, just don’t be in a hurry to get them there.

Finally, the patterns on the plan for the pushrods for the elevator and rudder are at least two inches too long. So just make them that much shorter when you bend them up initially and you won’t have to shorten them later.

Rigging for flight

Thanks to the inclusion of EZ Connectors in the supplied hardware and the short, straight, and all external pushrod runs, setting up the controls is very straightforward. I used the suggested initial control surface throws and dialed in a bit of expo as Bill suggests in the instructions. This required me to set the low rates on the transmitter to between 50 and 60% of throw depending on the control even with the mechanical setup – which holes in the servo and control horns to use – as recommended in the manual.

I want to reiterate the caution in the instructions and on the plan – the ailerons need to be rigged so that when at neutral they are parallel to the top of the fuselage aft of the wing, NOT the bottom of the wing. This is easy to do if you just mount the wing to the fuselage when doing the final setup of the ailerons, then sight along the bottom of each aileron towards the top of the fuselage.

The recommended balance point is 3 inches aft of the wing leading edge. With the ESC stuffed into the nose compartment, the drive battery needed to be all the way forward in the battery compartment under the wing (but not projecting into the nose) to have my airplane balance at that point. If you use a lighter motor, or perhaps heavier covering on the fuselage or tail, then the battery might need to project into the nose compartment. This does leave plenty of room for moving the balance point aft as your skills increase without having to do anything more than shift the battery.

Also, there are some provisions for mounting the rudder and elevator servos inside the fuselage at the back of the compartment under the wing, and I have read of some folks doing that. It would be worth considering this if you are using an exceptionally light motor, or if you chose really heavy covering for the aft end of the airplane. This installation is not covered in the manual, though. You’ll have to at least make the holes in the bulkheads and for the pushrod exits, ideally before closing up the aft fuselage, if you want to go this route.

The wing is held on with only four #16 uncrossed rubber bands. This looks far too little to most of us “old hands” but I assure you it works. It helps that the leading edge of the wing is also well-retained by the fuselage sides. This arrangement allows some “give” in the event of a less-than-graceful landing such as might come from practicing crosswind landing technique. Been there, done that, no damage done, including a badly botched landing that had the airplane hit first on one wingtip and then the rudder. All I had to do was get the wing back on straight and take off again.

The landing gear also retained with a couple of rubber bands. That, coupled with the gear’s ability to swing fore and aft in the slot on the fuselage bottom, gives a gear that stays where it’s supposed to on the plane but can absorb all sorts ungraceful handling without any trouble at all and with no collateral damage to the fuselage. It REALLY works well!

Flying it – From Mild to Wild!

As I usually do, I tried several different power systems on my SHAFT in order to get a sense of how it responds to varying amounts of power. The first power system I selected is in the middle of the range of the recommended power systems. This was the Atlas 2312/26 (also known as 2312/09), running on 3s li-polys. Later I tried a CommonSense R/C E5-L-10 first on 2s and then on 3s.

I had a small installation glitch in that the retention collar that is used on the Atlas motor shaft is a little larger in diameter than those used by AXi and some others, so I needed to enlarge the clearance hole in the firewall slightly. A small Dremel sanding drum made quick work of that little detail.

Based on the static tests, my initial propeller choice for the Atlas motor was the APC 9x6 slowflyer prop. For the first few flights I used either a 3s Kokam 1250 mAh SHD battery or a 910 mAh pack from the same battery family. This proved to be a good starting point – about 165W input at the beginning of the run.

Takeoffs, Slow Flight and Landings

With this first power setup the initial takeoff roll, on a rough short grass surface, was perhaps three or four airplane lengths. A little rudder is needed to keep the nose straight as the power (and the tail) comes up. From here it climbs as steeply as desired including going straight up. It’s easy to keep straight – one of the pluses of that long tail moment and generous rudder.

I needed some roll trim initially (later traced to a twisted left aileron). Once in the air and having that sorted I started to feel it out. One of the first things I noticed was that you can slow the airplane down much more than I’d expected. Of course with a wing loading of just over 7 ounces per square foot, decent low speed performance should be a given. Even so, I was and continue to be amazed at just how slowly it can be flown and how it doesn’t even threaten to do anything nasty even when flying that slowly. Hopefully that shows in the videos. There is no threat to snap roll and the stall doesn’t ever really “break” – the nose just drops gently.

As you can imagine, this makes landings – especially three-pointers – easy to do. There is plenty of rudder authority at low speeds to make heading corrections and the ailerons are responsive right down to the stall as well.

I recently was able to use the new 5 Hz GPS expander for my Eagle Tree MicroPower V3 logger on a calm day to determine that the minimum flight speed (while not in a high-alpha mode) is about 12 miles per hour (while flying about 2 1⁄2 ounces heavier due to the logger and accessories).

Couple the SHAFT's trustworthy behavior with the tall, wide and resilient gear and you get an airplane that’s just made to do touch and go landings. My previous favorite airplane in this size class for shooting touch and goes was my Sig Kadet EP-42 . It shares a couple of key characteristics with the SHAFT: trustworthy and predictable handling and a generous tail moment. It also shares having a fixed tailskid, which may prove problematic on a slippery surface (such as the mat runway at the Fern Prairie Modelers’ field in Washougal, WA) but elsewhere it works just fine. (I’ll find out in May.)

The SHAFT has a much lower wing loading than the EP-42 and has a wider, more resilient main landing gear, and bigger wheels – so it is even better for this kind of fun from more different kinds of surfaces.

The supplied wheels are adequate for just about any dry (or damp) surface you might want to operate off of. I suppose if you want to fly off of really swampy ground you’d want wider wheels, but I’ve been flying from some fairly soggy surfaces (typical for late fall and early winter in the Pacific Northwest) with no problems at all (other than mud being thrown up onto the aileron servos).

Power System number 2 – Commonsense RC E5-L-10 motor and CSRC 20A speed control

Commonsense RC also supplied me a system to try consisting of their E5-L-10 motor, Z-20 20A speed control, a 2s 1650 mAh battery and an APC 10x7 slowflyer prop. This is milder but a little lighter setup than the Atlas on 3s. The motor itself is a little “hotter” with a Kv of 1000 vs. 900 for the Atlas and on 3s this shows in the static test results.

The CSRC motor looks and feels well made and it runs smoothly and quietly. Like the Atlas it comes with all the bits needed to mount it either way, so of course for the SHAFT it was rear-mounted. The prop adapter was a little bit shorter, so I had to forgo the thick prop washer in order to get the prop and the spinner backplate (for the 1 1⁄2 inch Hyperion spinner I am using) on it.

Here are some static test performance figures on both 2s and 3s Kokam 3200 mAh batteries. Note that with smaller batteries that will fit in the SHAFT these numbers will be somewhat lower.

Prop RPM at full throttle Power In on 2s
9x6 SF 5500 78W
10x4.7 SF 6450 56W
10x7 SF 5275 115W
11x4.7SF 5375 105W
11x7SF 4725 128W

Prop RPM at full throttle Power In on 3s
8x6 SF 8525 175W
9x6 SF 7250 227W

On 2s and the 10x7SF in-flight motor temperatures never got above about 20 degrees over ambient the motor was well within its “comfort zone” when used this way.

The CSRC Z-20 controller is physically very similar to the Atlas unit, and based on the similarities I see there, plus the identical programming instructions (and programming cards that have exactly the same choices in exactly the same arrangement) I have concluded that it comes from the same factory as the Atlas. However, there are some detail differences. First of all, the CSRC unit does not have a switch (which is fine with me). Second, and perhaps more important to flying the SHAFT, the throttle response is not as strongly damped. Never when flying it did I feel there was too much lag time between the time I advanced the throttle stick and the time the motor responded as I did when flying the Atlas controller.

On the test stand and in the air the CSRC controller did just as you’d expect it to with no fuss and no hiccups. For a controller for a sport flying airplane that’s really all you need.

The low voltage cutoff is a little over 3V per cell and this was verified by in-flight logged data.

I also tried the CSRC E5-L-10 on the same 3s batteries and 9x6SF as I used with the Atlas 2312/26. Performance here was similar to that with the Atlas – that is, easy unlimited vertical - but since I was flying on a really cold afternoon (ambient temperatures in the upper 30s Fahrenheit) I don’t think the batteries were giving their best. The on-ground and in-flight measured powers were much below what I got on the test stand at room temperature (and with a much bigger battery). In flight ESC and motor temperatures never got much more than 20 degrees above ambient. The speed control under these circumstances got to about 30 degrees above ambient.

Overall the CSRC power system works very well with the SHAFT and is definitely worth consideration to power yours.


I am a long-time “sport pilot”. I’ve mostly flown trainer or intermediate type sport planes for the many years I’ve been flying R/C. So, even though I’m a relatively high-time pilot, I’m right in the target audience for the Super High-performance Aerobatic Flight Trainer – a plane that can be used to learn how to sharpen and make more precise one’s flying skills, then expand them greatly.

As such, the SHAFT has the ability to do, and do easily, everything I know how to do, and it clearly has the capability to teach me much. And as I learned by part way through the second flight, when I threw the dual-rate switch to the “full rate” position, more control throw does not make it nervous or twitchy, it just lets you do things faster and more crisply.

Rolls are axial with only a touch of elevator, and I’m sure the airplane will be able to show me how to (and how not to) add the rudder in for nice slow rolls. Response to the ailerons is crisp and about as precise as it can get given that these are flat plate ailerons. Because the wing is light it doesn’t have a great deal of roll inertia. Rolls stop immediately when the ailerons are neutralized. There is a touch of adverse yaw – just enough to let you know to use the rudder as well.

I have so far not experimented with any aileron differential – there is none in the basic setup of the airplane. This is something (along with flaperons) to try when I put the Berg 7P in the airplane instead of the 4L that’s in it now.

Inverted flight takes a small amount of forward stick, but only a small amount. And unlike many sport/trainer types that have other than symmetrical airfoils, the SHAFT does not get much more sensitive to elevator inputs while inverted. I’m sure that will lead many of you to do confident low inverted passes. Maybe someday I’ll get there, too.

If you move the balance point aft a quarter of an inch and trim the elevator so that advancing the throttle does not cause the airplane to climb then there is no forward stick needed to maintain inverted flight. This is a new experience for me and initially messed up my reflexes with regard to which way to move the elevator stick to climb.

There is enough side area and enough rudder to fly knife edge. Doing so is currently beyond my skills – but this is one of the skills I expect this airplane will be able to teach me. I’m working on it.

High speed flight

The SHAFT is NOT intended to be a speed demon. I mentioned before that Bill Stevens warns against overspeeding the airplane in the discussion of propeller choices in the beginning of the manual. The chief concern is that with sloppy aileron servos, flutter might ensue. To that end full power dives are probably not a good idea. I have NOT tried this out as that’s not my sort of flying anyway.

That said, the SHAFT moves right out when you go to full power, which is a good capability to have handy especially on a breezy day. On the second power system I tried, which is less than 100W peak (into a 10x7SF prop) the top level-flight speed (again from GPS data) is about 45 mph. Going to 3s and a 9x6SF on the same motor took that top speed up to a little over 50. As I said, the SHAFT is not intended to be fast, but smooth flying and highly aerobatic.

In the wind

I’ve taken my SHAFT flying in winds measured at 10 mph gusting to 15 mph at the surface and found that it copes with such a wind very well indeed, especially for a plane with such a low a wing loading. The combination of lots of control authority and sufficient pitch speed at full throttle to punch through the wind when needed helps a great deal, of course. But safe, predictable and obedient handling along with low drag are key to making flying in the wind as fun as it can be, and this airplane certainly has those characteristics.

With the low wing loading, it can also be hung in the air at part throttle and slightly nose up, with heading corrections made by rudder to keep the nose into the wind. I call this “stringless kite” flying. For some reason I really enjoy doing that on windy days.

The wide and resilient landing gear means that less-than-smooth arrivals, which happen often when flying – well, when landing – in gusty winds, are generally non-events. So far the worst damage my SHAFT has suffered in these conditions is a broken prop. Consequently, this is a plane that I will not be afraid to bring out at a meet when it’s breezy enough to keep folks from flying.

At greater weights

Just to see what would happen on one flight with the CSRC motor on 2s (not 3) I put in a 6.2 ounce battery (2s Kokam 3200). All up weight for that flight including that battery and the logging stuff aboard was 20.5 ounces. My log entry reads: “Not bad, not bad at all. Amazingly little difference in handling and enough more power from battery [vs. the CSRC 1650 on the prior flight] to offset the weight gain. Slightly faster, better inertia, but otherwise virtually indistinguishable from prior flight. Neat!” Peak power was right at 130W according to the EagleTree MicroPowerV3 aboard for the flight vs. just under 100W for the flight prior on the CSRC 1650 mAh 10C battery (which weighs 3 ounces less).

Flight video

The video that follows is assembled from clips all taken during my fourth flight on the SHAFT - while I was still trying to get really well acquainted with it. Since then I have gotten it to spin (and stop spinning) on command, and am making my way toward learning some of the other lessons it has to teach me.


Is This For A Beginner?

No, it’s NOT for a beginner pilot, though it could be a suitable first balsa kit build for someone who is comfortable flying a four-channel sport plane.

Bottom line


  • Complete kit with quality hardware and wheels supplied
  • Self-aligning interlocking design
  • Virtually mistake-proof construction (if the part doesn't fit, you have the wrong one)
  • Strong and lightweight structure
  • Obedient and responsive handling with no nasty surprises at the corners of the envelope
  • Wide track resilient landing gear
  • Removable horizontal tail and landing gear making for easy packing for travel
  • Light wing loading means high performance on a small flying site (it's comfortable in not much more than a single soccer field) and with moderate power
  • Good wind handling capability
  • Stylish looks - the best looking wing-on-box-fuselage plane I've seen


  • Needed to open up elevator joiner clearance in rudder (fixed now)
  • I don't get to fly mine enough

Simply put I really really like this airplane. Unless you’re looking for an all-out “3D” airplane on the one hand or a basic trainer on the other I expect you will like it, too. What more can I say? I look forward to flying it as often as I can for a long time to come.

Oh, and here's a sample of another neat thing you can do with the Eagle Tree logger and the GPS expander - overlay your flight path on a Google Earth image!


All still and video in-flight photography by Ian Cawley. I took the construction pictures and the static shots of the airplane on the ground in the cul-de-sac.

Breaking News

Bill Stevens is working on both a larger and smaller version of the SHAFT, with the 52 inch span SHAFT 25e going into beta test as this is written.

Last edited by Angela H; Feb 21, 2008 at 08:21 PM..
Thread Tools
Feb 21, 2008, 07:04 PM
War Eagle!
Spackles94's Avatar

Wonderful job with the review. You did a fine job with the build and the covering, and I appreciate you taking your time and testing it with two different power systems. Nice job!

Downloading the video as I type... I'm sure it'll be as good as the photos! And yeah, that's a money shot you got -- definitely worth anyone's desktop image!

Keep up the good work!
Feb 21, 2008, 09:10 PM
Registered User
Great review Bernard!

I wish your review had come out while I was building my Shaft last week and I had read your suggestion on how to keep the tail part square (step 14). I did not do what you suggested and now my vertical stabilizer is not quite at right angle with the horizontal stabilizer :-( It is off by a few degrees and it will probably affect how it flies. Oh well, another lesson learned.

Feb 21, 2008, 11:37 PM
Registered User

A very nice review. I especially enjoyed the video. It sure does seem to slow right down but remain very solid.

I bought one of the first kits at NEAT last fall. It's still in the plastic bag I'm hoping your review will get me into gear to take it out and build it! You're review is very thorough and has pointed out a few possible pitfalls as well as provided some good links.

Thanks and hopefully mine will be in the air come spring!

Jim, KK1W
Feb 21, 2008, 11:43 PM
Registered User
BEC's Avatar
Thanks for the kind words, all.

I put about 50 minutes more on mine last Sunday with the logger stuff all taken out (2.7 ounces worth). What fun! This airplane is going to teach me so much, and it's a ball to fly.

Mike, sorry I was too late....hope your SHAFT goes well anyway.
Feb 22, 2008, 06:45 AM
Dr. Dave
Bernard, Wow! I really appreciate seeing projects like this that you have invested so much time into. In other words a full build. Great looking finished product. Excellent. Thanks as well for all the testing.
Feb 22, 2008, 07:45 AM
Registered User
jpwkeeper's Avatar
Could I get you to go here:

Difficulty to fly rating system

And rate this plane? I think it's been rated once, but more ratings increase their accuracy.
Feb 22, 2008, 10:37 AM
Registered User
600ml's Avatar
Work the shaft!
Feb 22, 2008, 01:56 PM
Registered User
BEC's Avatar
Originally Posted by jpwkeeper
Could I get you to go here:

Difficulty to fly rating system

And rate this plane? I think it's been rated once, but more ratings increase their accuracy.
I already did - post 72 in that thread.
Feb 22, 2008, 04:25 PM
Registered User
pda4you's Avatar
BEC - thanks for the great review....

I really need to get one of these.... My wife loves it when I come home with more airplanes....

Feb 22, 2008, 05:11 PM
War Eagle!
Spackles94's Avatar
Originally Posted by pda4you
BEC - thanks for the great review....

I really need to get one of these.... My wife loves it when I come home with more airplanes....

I'm glad I'm not the only one.

Yeah, I really need to get myself one, too. It looks like a winner. And either of Bernard's options for a power system look very suitable, too.
Feb 22, 2008, 07:04 PM
Registered User
BEC's Avatar
I'm sure there are also Scorpion, Hyperion and AXi choices that would be comparable. Get 130-170W into a 9 or 10 inch prop and you'll have PLENTY. Anything over 100W would be good.

Here are some 3s suggestions besides those I've flown:

The AXi 2212/26.

The Scorpion the 2212-26 or 2212-22. The APC 9x6SF in my tests draws almost exactly the same current as the 10x4.7 so you can use the 10x4.7 numbers in the Scorpion tables for reference.

From Hyperion the Z2213-26 or -20.

And, of course, there is the Hacker A20-22L that Bill recommends.....
Feb 23, 2008, 02:45 PM
Dave North
timocharis's Avatar
Bernard, that's an excellent review. Thorough, well-written and charming. You almost had me ready to go to Bill's site and order one, though I'm not normally interested in this kind of plane!

Get that knife edge going ...

Feb 23, 2008, 05:06 PM
JimNM's Avatar
The Scorpion 2215-18 on either GWS 9x7.5HD or a 9x5x3 - flies very well.

The UF Frio 10/10 motor on the same props, or even a 10x5APCE - works well too.

Feb 23, 2008, 10:53 PM
Registered User
BEC's Avatar
I'd thought about Scorpion 2215s and AXi 2217s as well.

The bottom line is that there are MANY choices that would work and work well on this marvelous airplane. Even something really inexpensive like the Suppo 2217-09 would probably do well....

Quick Reply

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Anyone else into high performance paper free flight? Like whitewings style? steelhead Hand Launch 32 Apr 28, 2009 05:38 PM
Discussion My high performance in flight data graphs.. morphs Power Systems 194 Apr 11, 2007 09:51 PM
MEGA BG high performance motor kartman24 Aircraft - Electric - Airplanes (FS/W) 0 Nov 07, 2001 04:38 PM
Cheapest High Performance Possible? 99esq Electric Ducted Fan Jet Talk 5 Aug 07, 2001 05:51 PM