Before we begin, let me review some myths I encountered or even had myself, and found to be either false or misguided during my own build.

This one is my favorite because it's just so appallingly wrong.

A typical Lycoming has basically 8 parts for each cylinder: the piston head, the rod, two oil rings, two valves, and two valve springs. Add in a crankshaft, heads, cam shafts, and a timing belt on the front cover, and you have a “core” parts count in the “low dozens” range. Final installation parts count isn't all that much higher: you're only adding some baffling, fuel plumbing, and a few control mechanisms.

The rotary supposedly only has “three moving parts”: two rotors and an eccentric shaft (compared to four piston heads, four rods, etc). It should be simpler, right? It has no valves!

Oh, it has valves all right, they're just called “side seals” and they do exactly the same job. In fact, a single rotor assembly has more parts than an ENTIRE 4-cylinder piston engine. Count the parts: Each rotor has 9 apex seal pieces, 6 apex seal springs, 6 corner seals, 6 corner seal springs, 6 corner seal inserts, 6 side seals, 6 side seal springs, a bearing, and 12 oil seal pieces. Oh, and the rotor. That's 58 pieces. It's a good thing we “only” have two rotors!

Add in the housings, eccentric shaft, the front cover, the stationary gears, the water jacket O-rings, the oil control O-rings, and the various bearings and bushings in the “stack”, plus a few dozen miscellaneous parts with names like “thrust needle bearing” and “end play spacer” and “eccentric shaft bypass valve” and you have an absolutely astonishing parts count. It takes 18 bolts (and 18 seals) just to hold the BLOCK together, to say nothing of all the accessories you bolt onto it. Add in the need for water and oil cooling, a turbo and intercooler (in our case), and a variety of accessories like spark plug coils and fuel rail return solenoids and you have a huge parts count.

This myth is usually accompanied with photos of the damage done when a Lycoming throws a rod, next to one of a rotary that “kept running” even though a nut went through the rotor housing and punched a hole in the face of the rotor.

I do believe that it's true that rotaries have more benign failure modes. Many of the the things famous for killing Lycomings, such as throwing a rod or having a valve develop an intimate relationship with a piston head, both kill the engine instantly and lead to an expensive rebuild process. In a nutshell, you're a glider AND you have a $10k-$15k+ bill waiting for you – if you make it.

In a rotary, there are documented cases of pilots “limping home” with partial power on failed engines in circumstances that would have resulted in a total loss for a piston engine. It is also true that Mazda engineered this engine for low failure rates. One reason I decided to stay relatively close to “stock” is because I wanted to count on this principle.

You also have a host of little “features” that add to safety. With a Lycosaur, you have a gravity-fed fuel system susceptible to vapor lock and carb icing. The rotary is more or less immune to these risks. It's almost impossible to shock-cool a water-cooled engine, and we can use tricks like water sprays on the radiator to provide a temporary boost to cooling (to bring down coolant temps after a long taxi/runup, or an extended climbout). You'd never spray water on a Lycoming cylinder head at 350F!

Just never forget that you're going to be innovating. Suppose you go with an unknown oil cooler and it blows out? Sudden loss of oil pressure is bad for ANY engine - both the rotary and Lycoming could seize faster than a pilot could react - and what would you do about it? It's not as if you have a “secondary” oil system. What about making a fuel rail? Ever watch a piece of MGS-335 impregnated fiberglass burn? I did, once: it burned like a candle. Will your fuel rail leak?

Don't ever assume that you have fewer failure modes. If you're reading this guide this is probably your first engine build, so if anything you're going to have more failure modes to contend with.

Safety is almost linearly related to vigilence and good decision-making. John Slade has blown several turbos and is still with us because he was prepared for it. I've reviewed nearly every experimental aircraft crash report that came out, watching for trends. The only trend I've seen is pilot preparedness.

No matter how well you build it, no matter what engine you use, always assume it can fail, WILL fail, maybe even catastrophically. The better, more prepared pilot will have a higher survival rate with a failure-prone engine than the less-prepared pilot with a better engine. Don't fly beyond your glide distance from a runway until you and your engine are best buddies.

It only takes one failure to test your limits. “First, fly the plane.”

On the plus side, in addition to the opportunity to “limp home” on a damaged engine, the horsepower available from the installation described on this site has a hidden benefit. As John Slade so eloquently put it:

Its ability to get you to altitude and out of the 'oh shit zone' very quickly. On a long runway
you can have enough altitude to turn back before you run out of 'land ahead' room. The first 3 
minutes of a Lycoming flight are a nail biter for me. With a rotary its more like 20 seconds.
- John Slade

Let me finish with an anecdote. One day, John let me take the stick for a takeoff in Slick Kitten. I did what pilots everywhere are trained to do: I floored it (smoothly). Imagine my surprise when John pulled back the power! His engine has so much power that you don't “need” all of it to take off, even with a healthy margin. Imagine a situation where you regularly take off on 75% power unless you “need” it, and you have the sense of it.

OK, this one is actually true, but it's misleading. If you think it's cheaper than a brand-new Lycoming, you're probably right. But if you can get your hands on one that's been recently overhauled, and it's in the $10K-$12K range… if cost was your only criteria I'd steer you that way.

A precise price list is hard to pinpoint, but consider a potential example:

You can skew either table whichever way you want, and supporters of both camps do it all the time in their forum posts. Heck, shipping options alone can modify these values 5-10% each direction. But so far, the rotary looks pretty good. You can trade cost for time up front, and you enjoy higher power, lower ongoing costs, and more features for the rest of the ownership cycle.

So why isn't this myth busted? It's those features. Sure, the installation costs are similar. But your long-term ownership costs are lower, and you have far more features. Lycoming proponents love to quote the cost of used O-360 engines. But that's an apples-to-oranges comparison against a turbo-boosted 13B. Sure, the Lycoming is simpler, but you get more value from the rotary.

Just do some quick math. Assuming the above table is anywhere close to accurate (and I've already disclaimed accuracy of this material - did you read The Disclaimer?) in a rotary you get 240hp for $13k. That's $54/hp. The Lycoming is closer to $84/hp, and this monkey-math doesn't even allow for long-term maintenance costs. It's like Costco, buying in bulk. You get more horsepower, at a lower cost-per-hp.

Next, add in fuel costs. And although the rotary used to be known for high fuel burn, compared to a Lycoming it's more than acceptable - it's in the same ballpark if you match throttle/power settings, but you can burn cheap automobile gasoline from the pump, not 100LL. When I wrote this guide (July, 2010) autogas was running $2.75/gal and 100LL was $4.83.

Once you've rebuilt your engine, you should have the confidence to do it again whenever required. With that Lycoming you're almost certainly going to need an A&P - it's hard to even get parts for one unless you're “in the system”. But you can do a rotary rebuild in a good weekend or two, and the rebuild kit is about $1200. I've heard quotes for Lycoming “field” overhauls that ran $8k-$15k, or more!

The final advantage, in my book, is the ability to innovate. For some this is just dangerous. You know who you are - consider stamp collecting as a hobby. But if you want to try a 2-pass radiator, or a larger throttle body, or a different brand of turbo, with this engine you can do it, in a way that would never be possible with a Lycoming.

To be fair, I should point out some points of possible concern when installing a rotary.

Although there is an active aftermarket for parts and accessories for the core engine, three important components (the PSRU, ECU, and mount) are each available from only ONE vendor.

If Rotary Aviation ever went out of business, it's possible to work around the EC3 by simply using an aftermarket ECU. Although it wouldn't be optimized for aircraft use, it would probably work fine under most conditions. I would recommend looking into the MegaSquirt. This is an Open Source hardware/software project, which means two things:

1. It's unlikely that it will disappear any time soon.
2. Even if development on it stopped, and you were unable to obtain the units commercially, you could send the provided PCB layouts to a PCB house, purchase the parts from the parts lists, and make your own. It would probably be a day or two's work for an enterprising individual with some soldering skill.

Although it would be disappointing to see the CozyGirrrls stop producing their engine mount (I was thrilled to unpack mine - it's a very well-made product), it could be replicated fairly easily by a competent welder.

The mounting plate is fairly simple: a 1/2” sheet of aluminum cut in the shape shown (the “ears” attach the plate to the mount), with holes match-drilled to the engine core's oil pan holes. Appropriate holes are cut into the face for oil drainage back to the pan. The “ears” attach via Lord mounts to steel bushings welded into a frame.

This is the difficult one. I suppose you could engineer your own. The redrive is a fairly “simple” device in theory: a 6-planet gear set (available as aftermarket “upgrades” for reverse gear for Ford transmissions) is held fixed in a housing. The housing is attached to a mounting plate on the back side of the engine/engine mount.

The sun gear is coupled to the engine via a clutch damper bolted to an RX7 automatic transmission flex plate. As it turns, the planetary gears turn at a reduced rate. (The gears are all turned by the sun gear at a fixed rate - the reason there are six is to transfer all that horsepower.) The outsides of the gears mesh with a receiver that transmits the final output power down a prop extension shaft to the prop. The entire assembly rides in a cast aluminum carrier that is fed with cool oil for lubrication by the engine's oil cooling system.

Actually manufacturing one, though? Wow, that's a hell of a task. You'll need a prop extension and shaft, either something fabricated or purchased that you can machine with the splines to match a ring gear. The ring gear needs helical gear teeth cut into its inside face to match the planetary gear. And you need a shaft to take power from the engine via a damping plate and into the sun gear. Finally, you need to figure out how you're going to manufacture a housing to hold it all.

Once you cross THOSE hurdles, you still need to do all the testing and “debugging” that Tracy did to address wear issues and other problems. Me, I don't like you're chances. If you have the cash to buy the engine core and some parts, consider: there are LOTS of engines. There's only one Tracy. I'd order soon.

No, no, I'm not talking about testing the engine at altitude. There's nothing that's going to happen to a rotary engine at 10,000 feet that Mazda hasn't tested deeper than you ever will. There are quite a few places you can drive to that are that high: Trail Ride Road in the Rockies gets up to over 12,000 in some places. The engine isn't going to do something odd just because you stuck it in an airplane.

What I'm talking about here is all of the parts required to create an entire FWF setup. The radiator. The oil pan plug. The fuel return solenoid. That radiator modification your welder swears by. All of these things create an entire system of components that will never have operated together until YOU fly with them.

All experimental aviation pilots are test-pilots, but those who fly with an untested engine configuration are even more so. If you talk to the “old school” builders, they'll tell you that because of this, you should fly first with an aircraft engine - then, if you want to, do the auto conversion.

Ha. Well, of COURSE you're not going to do THAT. You told your family and friends you went with the auto engine because “They make millions of them, and they're modern technology. They're actually SAFER.” But I know. I know you're secretly a cheapskate that doesn't want to pay more for an engine than you spent to build the plane. And more important, you're secretly drooling over installing a rotary because you can paint it red and black and it would just look amazing under your cowl.

I know.

So take it from me: go into the project with open eyes. Your fuel pumps will clog. Your radiator inlet won't cool well enough, and you'll overheat after some long runups before you've climbed into the pattern. And you'll always wonder if your turbo heat shielding is good enough. Invent ways to test things. Come up with excuses to poke and prod. It'll save your life.