Subaru EG33 Conversion
EG 33 Specification
Engine Mount
Cooling System
Fuel System

Auto Engines in Aviation

Why would anyone want to incorporate an auto engine in an aircraft when there are so many well proven engines designed solely for aircraft use? We have all heard this statement a thousand times yet we persist in trying to achieve this goal. To the right is my auto conversion in its infancy.

The Reagan administration drove the automotive industry to produce automotive power plants with low emissions and high efficiencies. To achieve this end aluminium was chosen to provide lightweight engines that provided generous amounts of power without suffering detonation commonly found with steel cylinder blocks. The high thermal conductivity of the aluminium head and cylinder block reduced the localization of high combustion temperatures and encouraged the use of high compression designs. It was these high compression designs that afforded the trend to high power output with improved efficiency. This change in direction, by the automotive industry, was fortuitous to the experimental aircraft community since there became a number of new engine designs that were light in weight, and of sufficient power to meet the demands of even the larger four seat experimental aircraft.

History demonstrated that the use of automotive conversions in aircraft results in a higher risk of engine failure than when using certified engines or engines designed purposefully for aircraft use. The primary cause for the lower reliability resulted from the lack of adequate communication between fellow experimentalists. If one viewed the installation methods used from the early Cessna 140 to the latest Mooney's, there is a common thread of proven installation technique. These techniques are reflected in the experimental community. The homebuilder who chooses to incorporate an alternative power plant is on his own when it comes to installation technique and one only has to look along "Auto Row" at Oshkosh to view the variety of techniques employed to incorporate an auto engine. It is the exception rather than the rule to find a common design element between one installation and the next.

Is the auto engine less reliable than the more typical Lycoming or Continental? I do not believe that it is less reliable, and in many cases I believe that it can be made more reliable since it is constructed from more modern alloys, employ's liquid cooling, and in many cases the crankshaft is supported by more bearings than the certified counterparts. The piston speeds of the automobile engines are often less than those of the certified counterparts even though the automotive designs are operating at higher RPM. The EG33 used in my installation is a six cylinder engine incorporating a seven bearing forged crankshaft. If one investigates the failures within both auto conversions and certified engines it becomes very evident that most failures result from the systems that support the fundamental engines operation and it is not the basic engine mechanism that fails. In most cases it is the fuel system, ignition system, or PSRU (Propeller Speed Reduction Unit) that failed resulting in a dead stick landing. Why then, do auto conversions receive such bad press? In many cases it's a question of denial. By identifying a failure as resulting from an inadequate automotive engines pilots can put their head in the sand and believe that in using a certified engine they are quite safe. Since many supposedly certified engines are supported by non certified systems this may not be a very smart concept. As an example, the Lycoming 0-360 incorporated in the Cozy MK IV design uses a Elision Carburetor. This carburetor is a fine piece of workmanship but is not certified. Many pilots/builders have had difficulty setting up this equipment posing a risk to the pilot and his/her passengers, but by all accounts the pilots believe that they are flying a certified power plant.

OK enough of bashing the certified aircraft engines of Lycoming and Continental. The Elision carburetor is a fine design and is mealy used to illustrate a point. The concept of adding uncertified systems to supposedly certified power plants extends well beyond the use of the Elision Carburetor. We should all recognize that we are all flying with something less than certified and that those condemning fingers, pointing at the auto conversion, are ill founded.

Does this mean that auto conversions are more reliable? Certainly not, but there is also no reason to suppose that they should be less reliable. The reliability is controlled by the builder. In utilizing an alternate power plant the builder must recognize that he/she is on new soil and that each element of the conversion process should be viewed as potential source of failure. With this knowledge the builder should build appropriate failure tolerant systems or have some plan in place to mitigate such a failure. Clearly there are some components that cannot be duplicated, crankshafts and PSRU's. In these circumstances the builder must satisfy himself that the probability of failure is at an acceptable level. This also applies to certified devices otherwise fly a Microsoft flight simulator instead to remove all physical risk.

Reasons To Use Automotive Power Plants

The choice to select auto power plant, over the certified rival, should not be taken lightly. It is not as simple as: "I WANT TO SAVE SOME MONEY" If this was my reason then I would not have started the project. My choice of converted auto power results from:

  • Desire for a newer technology,
  • Belief that there is an alternative solution to Lycoming and Continental,
  • Exercising the experimental desires of the individual,
  • Belief that by reducing maintenance costs, the likelihood of completing a more thorough maintenance programme is probable, thus improving the inherent reliability,
  • Belief that there is a power plant that does not have to be molly coddled to keep it alive whilst in operation, and;
  • A more fuel efficient device is out there enabling more flying hours, or more speed, for the same operating cost.

Engine Choice

The Kinda Kozy (Cozy MK IV) requires a Lycoming 0-360 engine as a minimum ( there is at least one flying example using an 0-320 with CS prop ) and a 220 HP Franklin as the alternative engine option. Inconclusive tests done by Nat Puffer indicated little performance gain using the Franklin engine in comparison to the Lycoming 0-360. The Lycoming un-installed weight runs at about 300 lbs and an estimated 350 lb installed weight, whereas the Franklin come in at approximately 100 lbs above this figure. The data below indicates a dry weight without accessories of nominally 300 lbs.
Model 0-360-A
FAA Type Certificate No 286
Rated Horsepower 180 hp @ 2700
Number of Cylinders 4
Compression Ratio 8.5 : 1
Displacement 361 Cu. inches
Bore 5.125 inches
Stroke 4.375 inches
Engine Weight (Dry - With starter and Generator) 289 to 298


Model 6A-350-C1R
FAA Type Certificate No E9EA
Rated Horsepower 220 hp @ 2800RPM
Number of Cylinders 6
Compression Ratio 10.5
Displacement 350 Cu. inches
Bore 4.625 Inches
Stroke 3.5 Inches
Engine Weight (Dry - Without Accessories) 297 lbs.

My goal was to achieve reliable power of at least 180 HP with a weight no greater than the Franklin installation. I am a heavy pilot at 225 lbs so that the heavier power plant configuration served to balance my excessive front seat weight requirement. The Franklin would meet this goal and with the six cylinders would provide a very smooth power source. I have always felt that the Lycoming 0-360 and I0-360 is too powerful for a four cylinder configuration. 

I have investigated Ford and Chevrolet V6 configurations however the installed weight was found to be very high unless an aluminium cylinder block was incorporated. This aluminium option became very expensive. The Mazda two rotor rotary engine became a good candidate but other builders had been experiencing difficulty with cooling furthermore the output power was marginal with little reserve. In hindsight the three rotor Mazda may have been a good option but was not an engine in which I had familiarity. Subaru were offering a very nice range of engines ranging from 70 hp with the EA81 through to 230 hp with the EG33. The table below indicates the stock power from the Subaru range.

Model Power @RPM Displacement
EA81 70 hp 5600 1.8 L
EA82 80 hp 5600 1.8 L
EG22 135 hp 5600 2.2 L
EG25 160 hp 5400 2.5 L
EG33 230 hp 5400 3.3 L

The above table indicated that the EG25 would be marginal in its stock form but had some promise if the engine was tuned for more power. The EG33 however offered more than enough power and even if some power loss was encountered during the conversion, the design would render a safe configuration. Weight became the final issue. The cylinder block, heads and all components were all aluminium so it was likely that the overall weight would be within an acceptable range. With no reliable weight data available, I took a chance. The resulting all up firewall backwards installation is proving to be close to 400 lbs which is similar to that of the Franklin installation.

Model EG33
FAA Type Certificate No N/A
Rated Horsepower 230 hp @ 5400
Number of Cylinders 6
Compression Ratio 10 : 1
Displacement 202 Cu. inches
Bore 3.815 inches
Stroke 2.95 inches
Engine Weight (Dry - Without Accessories) 285 lbs

A full account of the conversion is given within these pages under the heading Subaru Conversion.

Last Updated:

Thursday August 31, 2006