Firefly 46 glow!

Until now, plans for DIY engines have tended to be of smaller diesels and maybe the odd 2 or 3cc non R/C glow. However, the Firefly .46 is the first forty class, two-stroke, R/C self-build engine to be published in a generation. In the UK, I can only remember the Jones, some decades ago, and that was a sixty, and even then you had to buy the plan. However, to please my hippy heart, and as an antidote to these ruinously expensive times, just mark that the Firefly .46 is presented entirely free. Also, due to the good offices of my mate and CAD man Glenn Royds, Firefly is also the first such home-brew engine to have 20 independently proven, 3D-rendered drawings to accompany the main CAD plan.

Now, one of the great things about starting to design your own glow engine is that you can make your operational criteria as easy or as hard as you like. So, from the outset, I wanted an engine that any competent traditional aeromodeller could build on a typical 3.5” centre hobby lathe. Like a model survivalist, I wanted to make (and maintain) every bit of this new engine from scratch in my shed, from basic materials. You know, metal stock in at one end of the shed, and a working engine out of the other. Straight away, this ruled out bought-in ball races. So, when it came to performance, my benchmark was the vague traditional sports forty target of, ‘ten thousand on a ten by six’. 

Article continues below…

Enjoy more RCM&E reading in the monthly magazine.
Click here to subscribe & save.


The Firefly had to be a radio control sport engine that I could fly week-in / week-out at the field. It had to be a practical, reliable item, not some over-designed display case wonder. Most importantly, to encourage new builders, it had to look, and be, very simple to build. When you think about it, the problem is that an aero engine is the summit of internal combustion evolution. It takes no prisoners. Whereas a dodgy home-brew model boat engine might putter away on very feeble power, an aero engine has to pass that killer binary test, i.e. will it fly my model aircraft or won’t it? For the would-be designer, this focuses the mind wonderfully. In the end, I reckoned that if the Chinoise copyists could knock out a glow engine, then so could I. In fact, the prototype was built without drawing any overall plan as such. I relied on puzzling out the layout on long walks with Ollie, which I found was the best method. However, I did sketch out, and carefully dimension, each and every part before I started cutting metal. I don’t have an engineer’s understanding of tolerances, so I relied on patiently fitting each part to each other, just like our aeromodelling ancestors.

Article continues below…

Being an aeromodeller (and not a model engineer) I immediately fretted about the practical constraints of projected model aircraft weights, spans and wing loadings. I also considered handling, torque and propeller size. Finally, I fussed about the carburettor. Rightly or wrongly, I had decided that a forty class engine was the best bet. Big enough for the internal parts not to require the skills of a gynaecologist, yet small enough to make on a modest lathe. Without being defeatist, it seemed to me that even if I finally managed to build a slightly under-performing home-brew .40, it would probably still power a plane. I convinced myself that a powerplant of this size represented a lower fraction of the overall flying weight of a forty-size airframe than it might with, say, a .15 or .25 size model. Taking this a bit further, I reasoned that if I didn’t quite get everything right, but upped the nominal capacity to say .45 or .46 (7.5cc), then I would be giving myself a bit of an edge in that the spare capacity might just cushion the inevitable efficiency losses induced by my amateur construction. 

Article continues below…


Article continues below…

To give you some practical idea of the biggest bits of my engine, the largest diameter material I had to grip in my three-jaw chuck was some 64mm round aluminium bar. This was for the engine’s backplate cum radial mount. The biggest item that needed to be held in my four jaw chuck was the crankcase, a block of aluminium 37 x 37 x 62mm, which naturally started out a few millimetres bigger. Incidentally, all of Firefly’s internal chambers and passageways are easily accessible and well within the compass of a small lathe or a basic drilling machine.


Firefly is a traditional ‘fitted’ engine. So, for example, I made and finished one item, like, say, a cylinder liner, then carefully machined the piston to exactly fit it. The photos show much of the basic engine building process. Unfortunately, a number of other snaps were lost in a storage unit fire, so I no longer have a complete photographic record of the entire build. Since I’m not a natural mathematician or machinist, I tend to make small lathe cuts and take my time, double-checking every single number and dimension twice as I go. I wrote out all my numbers on a pad by the machine, and ticked off the material as I machined it away. I routinely did trial dummy cuts of all operations with the lathe power off and the tool just missing the work, just to check there were no foul-ups. These simple habits have saved me many a job, not to mention the lathe and its tooling.


My inexpensive Chinese Winkie W*nkie lathe / milling machine is no Myford. You cannot just dial-in a thou and expect exactly a thou to come off the metal. In fact, for identical graduations of the control wheel, I know that my Winkie W*nkie sometimes takes more, sometimes much less. Therefore, I’ve developed a different modus operandi. All the key fits within the engine were achieved by careful machining down on the lathe to around the last thou or so, where I stopped to take stock. In practice, if I was lucky, this left anywhere between three-quarters to one-and-a-half thou of metal to be removed. This was then followed by the careful removal of the excess metal by traditional polishing or lapping. In my case this involved well-used emery cloth, commercial valve grinding paste and DIY wooden laps. These are very simple home-made tools for minute grinding and polishing. They’re made from a split wooden dowel, or a holed wooden block (to hold the emery cloth) and ‘charged’ with very fine grinding paste.

I used Holts (car shop) Valve Grinding Paste, using only the paste from the Fine end of the dual container. Now, a thousandth of an inch is quite a lot to remove when you have to do it by hand, so you must endeavour to get as close as you can ‘off the tool’. That said, with an agricultural machine like mine, discretion sometimes becomes the greater part of valour. You get down to a thou or so and have to be happy with that, rather than risk going ruinously under on the next inaccurate cut. So, to sum up: I now routinely get as close as I dare and then lap or polish in any critical fits. It is slow but sure, and the only way with my particular lathe and home-brew pistons, liners, bearings, carb spigots and back covers. I firmly believe this is why the prototype (utterly astoundingly to me) needed no gaskets, and has no gas or fuel leaks.


Since the engine was turned up using cheap, disposable, triangular carbide-tipped lathe tools from Chronos UK I didn’t need to learn how to grind up my own. It worked well, and I was quite pleased with the cosmetic finish, too. Since I’m a savage and hate changing drive belts, the whole engine was machined at a single lathe speed of 630 rpm. I think I only used the lathe top-slide (compound slide) whilst making the carburettor venturi. Mind you, the inexpensive Winkie W*nkie is an ideal bit of kit for the keen sports aeromodeller, and that integral milling arm is a fantastic bonus. In fact, I did all the milling and drilling on the WWM. I used the milling arm, with a £25 graduated tilting vice, to get the correct ‘angle of dangle’ to drill the carb venturi hole. I also used a rotary table whilst drilling the head and backplate holes, but only because I already had one. Meanwhile, I used digital callipers for almost all measurements, but used a micrometer on key cylindrical items like the piston, the crank pin, gudgeon pin, and the main bearing. I made the crankshaft and the cylinder liner first, reckoning them to be the hardest parts of the engine. These items then acted as gauges for the crankshaft bearing and piston. The prototype’s look is deliberately un-frightening and simple. Its ‘coffee tin meets shoe box’ layout makes marking out and holding the work whilst machining, a complete doddle. In case you’re wondering, although they look a little different, my prototype and the later versions built by the Firefly Proving Group lads are utterly identical inside. They only differ in external cosmetics.


I actually trialled two crankshaft production methods. One was via the traditional all-turned solution, which gives an integral crankpin. Later, on the advice of the Manchester Engine Men, (a group of famed and generous old school control-line Team Race dieseleers, such as John Feeney, Len Moran and John Daly), I tried the pressed-in roll-pin method. This uses a commercial hardened roll pin as the crankpin, simply pressed and secured with Locktite firmly into the crankshaft web. Essentially, it’s an interference fit. In fact, I chilled the pin in the freezer for an hour or two before pressing it into the previously drilled and reamed crankpin hole in the crankshaft web. I now prefer this second method because it’s quicker, requires fewer machining operations, and is much easier to get right. The men from Mancunia also advised me to dispense with my finicky bronze bearing in my prototype dural con rod, which I did with a second example, for comparison. Once again, I found this a better way to proceed compared to my normal traditional method. 


All tapping was metric and done by hand using just a taper tap on each hole and not bothering with second or bottoming taps. That said, I did grind back the sharp taper tap noses a bit to improve their versatility. The only ‘funny’ on the whole engine is the special metric fine M4 x 0.5 tap, used in my carburettor to match the O.S. standard needle valve thread. I sourced this from All fastenings are bread and butter metric modelling sizes, sourced from our local Model Exchange at Greenfield.


Since I wanted to make all the components for my own engine, I also drew up my own carb. For swank, I specifically wanted Firefly to have a remote needle valve, as well as an integral in-flight mixture control. In fact, I spent almost as long designing Firefly’s nifty remote needle valve / carb assembly, as I did working on the whole bloody engine. However, I’m also aware that home-made engines are often handicapped by poor carbs. Consequently, if the prospect of building your own Firefly carb sounds a bit daunting, Firefly is designed to accept a stock O.S. 46LA carburettor. While we’re on the subject, for the builder’s convenience, Firefly is deliberately designed to mate up with O.S. / industry standard / clone parts, such as O.S. fit silencers, prop drivers, and prop driver slip washers. This also allowed Firefly to match the tons of existing O.S. engine bits already residing in my shed. So, although the Firefly drawing shows my own home-brew versions of such standard parts in detail, some builders might wish to acquire the following commercial items to speed their build: l O.S. 46LA prop driver l O.S. 46LA prop driver slip washer l O.S. 46 needle valve l O.S. 46LA standard exhaust l  Cheaper, clone parts from the swapmeet would do, but only if identical in dimension.


As mentioned, the engine has a number of areas where particular care is needed to avoid leaks and to ensure adequate compression. These key fit locations are:

  •  Piston to liner, to ensure leak-free compression.
  • Liner to barrel, to avoid gas leaks and to enhance good heat transfer from the cylinder to the barrel for correct cooling.
  • Main bearing fit. This main bronze bush supports the crankshaft and must be fuel and gas tight, yet run as freely and as cool as possible.
  • Carburettor spigot to crankcase hole fit – essential to avoid air leaks and ensure reliable running and engine tuning.
  • Throttle barrel within the carburettor – must be slick and air tight in operation to ensure reliable running and throttling.
  • Rear crankcase cover / integral engine mount – needs to be gas tight and secure enough to support the whole engine.
  •  Cylinder head to barrel interface – needs to be gas tight.

10K ON AN 10 X 6 

Firefly is a plain bearing engine, employing a phosphor bronze (or cast iron) main bearing that you deliberately make yourself. This means that unlike a modern CNC-built engine, which will run-in almost immediately, Firefly will take a good bit longer for all her bits to bed down. It will certainly take time for the best fits to develop. I reckon that, providing you can get the piston and liner correct, the main bronze crankshaft bearing is the biggest determinant of Firefly’s ultimate performance. Of the five Firefly 46s built so far, first-run top revs were initially low, in the 6,000 – 7,400 range. However, fear not, this will soon rise with careful running-in! For example, my prototype began at around the 7,400 mark but can now easily turn an APC 10 x 6 prop at well over 10,000 revs on Southern Modelcraft 10% nitro / castor mix. In practice, my prototype flies a 5.5 lb World Models Skyraider ARTF aerobat very nicely, thank you. Firefly is no O.S. AX, but she can loop and roll this 52” span ARTF from level flight, so I reckon that she’s fulfilled her sports engine design brief.


Luckily for me, generous aeromodelling chums like Neville Griffiths, Jimmy Moore and John Feeney gave me massive moral support. They also donated most of the metals I needed for the prototype. Actually, I did weaken and buy some cast-iron for the piston at the Harrogate Model Engineering Show. John Feeney, in particular, suffered many a daylight raid on his workshop, when I was on the prowl for reamers, metal stock and technical sustenance. I used these metals in my prototype and did not harden any: Crankcase: Aluminium alloy – HE30 Crankshaft: Carbon steel – EN8M Cylinder liner: As crankshaft Con rod: Aluminium alloy – H15 dural Bearing: Phosphor bronze – PB1 (or cast iron) Cylinder head: Aluminium – HE30 Piston: Cast iron


Sourcing materials, or even telling one metal from another, can be hard work for the beginner, therefore John Feeney has promised to make a modestly priced materials and fasteners pack available for Firefly builders. Trust me, it’s a very cost-effective and convenient way of getting the various materials together. So, I commend it to you. You can contact John at: [email protected].


Due to a storage unit fire, most of my Firefly bits and pieces, my photos, drawings, portable hard disc and notes were lost. Providentially, I had kept most of the prototype stripped down in a jam jar, in my house. Upset at the fire, I forgot about the Firefly project for a while. Then, I made the acquaintance of Foamie Dave Royds and his brother Glenn, a professional CAD man. Glenn is the hero of the whole Firefly 46 project because he offered to draw up new CAD plans for me, directly from my prototype. When Glenn showed me the finished product, and the fully rendered 3D views I knew we had a winner. They are easily the best model engine drawings I’ve ever seen. This completely re-energised the project. Then it was my turn for a brainwave. In my Weekenders column, I asked RCM&E readers to offer to build a Firefly ‘blind’ from Glenn’s new plans. To my delight and astonishment, four lathe owning modellers immediately stepped forward. Most importantly, they built their prototypes based only on the plan and with no input from me. Since they were thus ‘proving’ the Whittaker Firefly .46 and her CAD plans, naturally I dubbed them the Firefly Proving Group. The good news for any prospective builder is that each engine worked first go! These four very generous readers are: Mike Bundy, Mike Havard, John Payne and Frank Payne. There’s even more good news! The lads have agreed to make their email addresses available to any reader who is building her / his own Firefly. Send me an email and I’ll pass them on.


Over the next few months we’ll be giving you Glen’s superbly detailed component drawings and 3D renderings. However, if you can’t wait that long, you can download them free from or via my own website at


That’s all for now. Remember to buy the March issue to catch the first part of Frank Payne’s step-by-step guide to building his own Firefly .46 from the RCM&E plan available at


See her running on YouTube at:

Subscribe to RCME Magazine Enjoy more RCM&E Magazine reading every month. Click here to subscribe.