Subaru 2.2 and 2.5

I am very interested Jonathan. I may need to buy a used Soob engine one of these days and I would really like to know what I should do to it to keep it reliable.
 
Dumping on the Subie

Dumping on the Subie

Seems to me there are over 1000 Sube powered gyros out there from Groen, RAF and converters using Don Parnham's redrives. How many out there using certified type engines? The EJ is the choice of the RAF and Groen factories- not Lycoming or Continental. Think they are idiots?

Someone here was stunned to hear of jug problems with cert engines? Please give me a break. Happens all the time in the real world.

RAF recommends a TBO on the EJs of 1500 hours last time I heard. They probably know a bit about EJs as they have delivered over 600 gyros powered by them.

Eggenfellner has dozens of EJ, EZ Sube engines in around and over 500 hours flight time with zero oil consumption, zero problems, perfect compression and low leakdown. Given a fresh EJ or EZ, these will be far cheaper over the long run compared to cert type engines. By my estimate Subes have accumulated well over 250,000 flight hours to date and rarely have core engine issues if properly cooled and oiled. You can thrash them at 4-5000 for a long time.

I'll agree with the statement that support systems like fuel, ignition and drive failures cause most of the engine outs on automotive conversions but as a professional engine builder for 30 years, I'll take a Subaru over a Lycoming any day.

BTW, poll, who has the highest flight time on a Sube without an overhaul?
 
Ross

Thank you for your post. First I don't know who has the highest flight time without an overhaul. Other than peripherals have you heard of any really bad things happening to them?
 
I'm a very happy suby man, I have a 2.2 EFI motor, come out of a wreck with unknown hours and did a bit over 2,000 hours before it spat a cam belt idler out the front . Put a new lot of idlers and bearings in and still purring away very happily. A friend of mine did 3,800 hours before replacing his 2.2, it was still going well but decided to replace in case it was getting a bit worn.!!

I would recommend replacing cam belt and idlers every 1,500 hours.
Also, it has to be rembered that the RAF belt drive creates a lot of pressure on the crank and a better way to go is get rid of that heavy belt drive and replace it with a New Zealand gear box [I have forgotten the name, sorry] then you can enjoy flying.
 
Within the grand world of auto engine conversions, this is about the pecking order of reliability:

1. Subaru 4-cyl.
2. VW
3. Corvair (William Wynne's)...

6. Subaru 6-cyl...

...The lower entries on the scale pose a threat to life and limb, although the 3l Soob has promise which it hasn't yet fulfilled.


Eggenfellner is using only the Subaru 3.0L H6 now. They seem to have a reliable package together, but not cheap. Still less than an IO-540, but not much.
 
Comparing a certified engine to a Subaru (or any other auto conversion) is more like comparing apple and oranges.

As a example RAF claims 1500 TBO time. How many RAFs have anywhere near 1500 hours? How did RAF determine what the TBO was?

Secondly, you speak of how many Subaru powered gyros there are. Compare that number to how many certified engines there are of the same power range. I'd take a guess and say there are more Cessna 150/152 powered units than there are all the gyros combined.

Now I'm not saying a Subaru is a bad engine all I'm saying is certified engines have a history that Subaru or any other auto conversion engine just doesn't have.

One other point about comparing certified engine to auto conversion engines is how they are run. The certified engine is operating at about 1/2 it's rated output power. While the auto conversion engine is operating at closer to 80-85% output power.

The biggest advantage the auto conversion engine has is the improvement in design. Many of the certified engines were designed over 50+ years ago.

The second advantage is cost. There is no question an auto conversion engine can be purchased and maintained for less than a certified engine.

As I said in the beginning comparing an auto conversion engine to a certified engine is apples to oranges.

Leon
(kc0iv)
 
Ross,

No one is "dumping" on the Soobie. Thom took the original post out of another thread to start some controversy.

RAF factory has said alot of things.... Ron made a good post to this effect.

For the price you can't beat a soob but for total package reliability and endurance and fault-tolerance the numbers on certified aircraft engine performance is clear.

The reason GBA, RAF and others use the soob is price not because they out perform or outlast certifed engines.

Eggenfellner is using only the Subaru 3.0L H6 now. They seem to have a reliable package together, but not cheap. Still less than an IO-540, but not much.

Not cheap is right!

A professionally rebuilt and setup soob is MORE expensive then a used aircraft engine and about the same as a used certified aircraft engine.

Another soob re-builder: http://ramengines.com/_wsn/page2.html

Little EA-81's are 4-15 GRAND not counting the redrive!

A great engine but now you are greatly reducing the advantage over an aircraft engine which is mostly price.

I lost a soob because a gear in the redrive was not hardened properly and spewed metal filings into my engine.

I lost a soob because it was a junkyard engine and had bad internals that showed up after about 10 hours.

I lost the heads on another soob because the carb I was given for the conversion had small jets and it went lean in flight even though it did not get to hot testing while tied down on the ground.


Like many things you will hear people talk about there great engine but not about the ones they had to toss in the bin.


The RAF co that "knows a bit" about EJ-22s has had plenty of them die from bearing and crank-shaft failure because of the belt redrives putting pressure on the rear bearing ( and belts do break in flight).

If you go to the AirSoob forum, which I have been a member of for many years you will find plenty of soob problems but the number one problem is cooling. Soobs especially EJ's are extremely intolerant to air in the cooling system, bad thermostats and other cooling issues.

Now, because I have experience with soobs and I don't produce a glossy brochure claiming they are the greatest thing since pre-rotators mean I am 'bashing' soobs? NO. No engine is perfect and that includes soobs and certified aircraft engines.

What is important is that you know what the issues are and make the best decsion for your application and how you will fly.

The Experimental aircraft community is VERY lucky that we have Soobs as a resource. If not our hobby would be a great deal more expensive for those of us that want reliable self maintained engines.

From my experience I would suggest going with fuel injection over carburation, buying a new engine from subaru or one that has already been converted and has at least 40 hours on it. Go with a gear reduction drive rather then a belt drive and design your cooling system to remove air, stay cool during long taxi operations as well as long flight operations in hot humid environments.

If you have the talent or access to cheap talent for rebuilding engines, rebuild that junkyard soob rather then play the used engine lottery. Sure the engine ran fine in the car but what happened to it when it was stored? shipped? Soobs are popular cars with young mail drivers that beat the $&!T out of them. Do you want to fly an engine that had the clutch popped on it 200 times and had the rev-limiter disabled?

If I had the money to buy a professionally rebuilt soob I would also seriously consider non-certified aircraft engines or used aircraft engines in my decision making process.


Good luck, do research, fly safe and know that all engines can fail!


.
 
Last edited:
Comparing a certified engine to a Subaru (or any other auto conversion) is more like comparing apple and oranges.

Leon
(kc0iv)

Excellent post Leon,

I did not see it while I was writing mine up.

The point about RPM and percent of power usage is important.

Elaborating on a point that Ron Awad made...

If something goes amis in my aircraft engine it still has 50% of spare power range to limp me home. When an engine has a problem and it is already at 80-85% power (or more) and doing far more RPM the engine has no reserve or fault-tolerance to limp you home and is much more likely to just self-destruct.

Also, Gyros are FAR more 'draggy devices' then fixed wing aircraft so even comparing soobs in fixed wing experimental is to a certain extent apples and oranges comparison to soobs in gyros.

Gyros, stress engines much more then FW's. FW's operate at much lower cruse RPMs then gyros and tend not to vary the RPM as much nor Yank-n-bank as much. All of this adds stress to the engine package.

Many people I have talked to about auto conversions have said "oh I got 500 hours on it with out a rebuild" but then 5 minutes later in the conversation you find out they had a redrive or belt failure or a problem in the cooling system crop up or a muffler rust out.... That IS part of the engine PACKAGE.

Soobs are great engines and the CORE is HIGHTLY reliable if you are starting with a virgin, rebuilt or cared for CORE but you must consider the whole package when making decisions about reliability and fault tolerance.

.
 
Yes Tim you are right I pulled this from a thread you had stated the 5 cats of aviation engines.

It was because you had put the Rotax which turns at 6500RPM and most are 2 cycle engines ahead of the Soob 4 cycle engine. All of these engines have redrives and take this higher RPM to turn a prop at around 2700 RPM. I have heard several different horror stories about the Rotax and was just wanting to get to how they compared to the Soob engine. The certified engines are set up to run full out at around 2700 RPM and still be in the proper max hp to those engines and at 50% of actual power capacity. I know that my Soob redrive is a 2.1 to 1 ration and when I am at 80% I am at 4600 RPM.

I have heard or read I really can't remember that tests have been done on the 2.2 where it was run for 500 hours at 100% with out a mishap. This is like autos when on a test track. Done under ideal conditions and with little to no induced stress on these motors. Just a test to determine reliability.

I am NOT trying to be controversial in this matter just curious. Like everyone else on this planet I am trying to have the best for the money I have already spent. If it takes a different motor then I will budget for it. :wave:
 
I'm reading things here like cooling problems, carb jetting problems, redrive problems. None of these are core engine problems. While a drive failure will bring you down as surely as a mechanical engine failure, this is not the fault of the engine. Kinda like saying a Lycoming failed because someone ran it lean on a hot day and melted a piston. Operator error can't be blamed on the engine.

Absolutely the whole package needs to be engineered and operated properly to have reliability but I stand by my statement that if oiled and watered properly and you have a proper drive design, these things are just as good as a Lyco at a fraction of the cost. More thought and testing is certainly required than bolting on a Lyco for sure. It ain't all roses. I know, I've done it and am working on a second one.

I agree, putting a carb on an EJ is a step backwards.

Cost is no doubt a factor in the factory choice of Subarus. A cert engine would price the kit out of reach for many.

I'd be very interested to hear from RAF, Groen and Parnham owners who have had drive or engine problems- or successes.
 
It was because you had put the Rotax which turns at 6500RPM and most are 2 cycle engines ahead of the Soob 4 cycle engine. All of these engines have redrives and take this higher RPM to turn a prop at around 2700 RPM. I have heard several different horror stories about the Rotax and was just wanting to get to how they compared to the Soob engine. :

Thom, here we are confusing TYPE with Application.

Rotax 2 strokes are smaller engines and so far as I know fall BELOW the weight-power applications of even the EA-81-82 engines do they are not directly comparable.

Rotax 4-Strokes however, are valid to compare against auto-conversions.

Rotax 2-Strokes suffer the same problems as any 2-stroke engine and Rotax has put out a lemon or two (Stan's experience for example). That said however, there is a reason that Rotax holds the market for 2-Stroke ultralight aircraft engines and it is Rotax that sets the high-water mark for reliability and performance.

In this size-class the engines to compare to would be 1/2 VW's, Kawasaki snow-mobile engines and similar. NOT EJ-22's and 25s.


.
 
Ross,

I agree completely! Excellent post.

The only statement that I would make a MINOR comment on is this one:

"just as good as a Lyco at a fraction of the cost"

I agree so far as CORE goes but I don't know of a redrive solution that has shown to be as well tested and reliable as the Lycoming direct drive.

Also, in the case of a Certified Lycoming, you have a network that tracks manufacturing errors that can warn you of a bad production run of a part before it becomes a problem. No such infrastructure is present in used or rebuilt soob engines and definitely not the redrives and support components.

But so far as the CORE cores I am right with you, I actually made a statement similar to this about Jonathan's EJ-22.

.
 
Ross is full of it! ;)

You can buy a brand new Lycoming engine in the 150 horsepower range for under 20 thousand dollars. How much does a simular horsepower Subaru engine cost from one of the autoconversion manufactors? As much or more than the Lycoming.

Sure if you want to go to your local junkyard and pay 1000$ for a used EJ-25 and then pay 3-4 grand for a redrive, and then hire the local muffler shop to make you a exhaust.... Yeah, you can have less money than a Lycoming. But what you have is a used engine, a unproven package, and nothing comparible to a Lycoming for dependability.

Trying to debate pros and cons with people like Ross is pointless in my experiences. Anyone willing to spend years and thousand of hours to build a plane like a RV-6, then go against the grain to put in a auto engine over the reccomended Lycoming.... well those guys are fricking Hard Core! Basically you couldn't give these guys a Lycoming, they believe a Lycoming is total junk. We have a famous RV-6 pilot at my airport, he has a Mazda Rotary in his and he thinks anything with Pistons is crap, much less a Lycoming!!!
 
A Rotax two stroke has very few moving parts, and it is designed to be ran at 6000 + rpms continuously. Car engines just weren't designed for this PERIOD. Just because Subaru lays out the cylinders the same way as true aircraft engines does not mean the Subaru is a aircraft engine doing duty pulling Subaru Forresters around. It is a car engine designed to be used in the lower rpms ranges typically seen in the car
 
Ross,

I agree completely! Excellent post.

The only statement that I would make a MINOR comment on is this one:

"just as good as a Lyco at a fraction of the cost"

I agree so far as CORE goes but I don't know of a redrive solution that has shown to be as well tested and reliable as the Lycoming direct drive.

Also, in the case of a Certified Lycoming, you have a network that tracks manufacturing errors that can warn you of a bad production run of a part before it becomes a problem. No such infrastructure is present in used or rebuilt soob engines and definitely not the redrives and support components.

But so far as the CORE cores I am right with you, I actually made a statement similar to this about Jonathan's EJ-22.

.

There is no doubt that there are some pretty crappy drives out there and almost none have had torsional vibration testing done to them. That is a cause for concern.

I don't buy the certified paper trail statement. This is not protection from bad parts getting into customer's engines. This was clearly demonstrated over the last 5-6 years with the crankshaft debacles at Lycoming (3 of them!) where dozens of cranks broke, people were killed and thousands of cranks were recalled plus several lawsuits against Textron. The paper trail only allows Textron to contact certain owners with the bad news after the fact.

I don't think Lycomings are crap, just that their QC has been poor (better now after the lawsuits) and that they have been slow to make changes to their designs where classic repeated failures have happened for the last 20 years. Some of the clones have addressed these issues. Why not Lycoming?

Lycomings work fine for the majority of flyers but to ignore their everyday failures, problems and weaknesses is myopic and unrealistic. Even as a Subaru advocate, I tell many people who are building aircraft who have no knowledge of engines to buy an ECI clone rather than a Sube or factory Lycoming.

Equally, I agree that using a junkyard Subaru can be a risky choice in an aircraft. How lucky do you feel? On the other hand, Subes are not forgiving of having hammer mechanics work on their insides. I am aware of more failures caused by "overhauls" from unqualified people than on a factory virgin engine with low miles, clean oil, clean filter, good compression and good leakdown.

Price wise, the firewall forward Sube kits are more expensive than a comparable Lyco clone these days. I think the competition is good and so are the choices available to buyers.
 
Sorry guys, I need to call you out on this one...

Elaborating on a point that Ron Awad made...If something goes amis in my aircraft engine it still has 50% of spare power range to limp me home...

The certified engines are set up to run full out at around 2700 RPM and still be in the proper max hp to those engines and at 50% of actual power capacity...

Huh?

A certificated engine running at 2700 RPM is running at 100 per cent of its capacity, not 50. Dropping a cylinder on a Lycoming loses the same per cent of power as dropping a cylinder on an EJ 22/25. Either one should have been chosen (sized) to cruise at 75 per cent power or so.

A Lycoming O-320 has about double the displacement of an EJ25, and makes max power at about half the RPM. If both make 150 HP, the Lycoming must create twice the combustion forces on piston pins, connecting rods, etc. The Subaru must spin twice as fast, and that creates more inertial forces on reciprocating parts, but those parts are much lighter to start with, and I'm comfortable that improved engineering and metallurgy can cover that.

The limiting factors in how much power both can make is (a) how much air they can pump and (b) how much waste heat they can dissipate.

Rotax is proving with the 912/914 that a high-RPM, small-displacement, water-cooled four-stroke engine with redrive can set and reach high TBO. There's nothing inherent in higher RPM that makes reliability go down, if the engine is designed for it.

Water cooling should allow more precise tolerances, better efficiency, and longer life of some components.

The blame for short-lived Subarus has to be the improvised crap we hang on them for external systems.

The one great exception would be engines with manufacturing defects. Cars just pull over if they lose a crankshaft, and carmakers plan their quality-vs-cost accordingly.

I know someone locally who just had a Sub 2.5 brand new from the factory fail at under 5 hours due to an incorrectly installed part in the bottom end, from the factory. It happens, but Lycoming hasn't done too well with bottom ends lately, either. This is where guys like Jonathan, who can blueprint a bottom end, have an advantage over most of us.
 
Indeed, excellent post. It is a complete fallacy to assume that high rpm causes high wear rates. This is not supported by any facts. Shorter strokes and smaller bearing journals keep piston and bearing surface speeds only slightly higher than a Lycoming at typical cruise rpms.

Auto engine validation these days generally exceeds what a certified engine needs to demonstrate with most OEs doing at least 400 hours at rated hp rpm and some doing 1200 and 1600 hours at WOT and high rpm for transmission durability validation. Subaru proved their EJs with a 100,000 km world speed record almost 20 years ago. 5400 rpm WOT for 17 days if memory serves me correctly.

The EG33D I just dismantled has an estimated 2100 hours on it. No measurable wear on any parts. I'm not talking within wear limits here, I'm saying not even .0001. I'm guessing this engine spent most of its life between 2000 and 3500 rpm being in a sports coupe. Anyone here think that this thing won't last 1000 hours at 4000 rpm, WOT. This is a beautiful piece of engineering and metallurgy.

A bagged out Lycoming or Subaru is probably gonna treat you badly sometime in the future. Keep 'em fresh and you'll probably have good luck with either.
 
Last edited:
I agree in general Paul, a great post.

I got confused when you quoted 2 sentences in a row and only one was mine though...

As I have repeatedly said, the core of the soobs is fantastic, however, it is still not type designed and light, small parts moving parts have less fault tolerance I think.

The lycoming is a better known system that has seen more duty in our application and there is something to be said for KEEP IT SIMPLE.

And again, the supporting systems are the downfall.

Cooling system fails = application fails.
Redrive fails = application fails.
Battery fails = application fails.
Charging system fails = application fails.

A lycoming is oil cooled, does not depend on water pumps, thermostats, hoses, radiators and coolant.

Direct drive

Magnetos that do not require a working battery/charging system..

And there are the redundant systems found in many aircraft engines, and those 'could' be added to the soob but typically are not such as:

Redundant mags.
Redundant plugs and wires.
Mechanical and electrical fuel pumps
and so on...

.
 
I have been reading all this tit for tat for the past thirty minutes and just have to put in my two cents worth.

Just for the record, I started converting Subaru engines in 1988 and have done well over five hundred systems which are on just about every brand of gyroplane and airplane.

I only know of four engine failures in 17 years with my systems--one crankshaft failure, one alternator failure in which the pilot ran the battery down trying to make it home (no backup battery), one fuel pump failure (no backup pump) and one connector pin corrosion problem (no dielectric grease in connectors).

Two pilots had belt failures after a few hundred hours. One was set too tight and the other one had a prop strike which bent the stub shaft and caused excessive belt misalignment.

BTW, the crankshaft failure was on the dual overhead cam engine which had a 6-7 thousandths runout problem on the crankshaft flange. That problem was fixed when the single overhead cam models came out.

Most of Rotary Air Force Marketing's problems with crankshaft failures were due to two factors: 1) most of the engines were bought in 1994 in the "great midwestflood" and sat storagefor years until grease dried out in the cam bearings, etc.

The second problem was that when RAF bought one of my reduction drives and made a Canadian copy of it they replaced my $35.00 outer support bearing with two $1.87 non-precision bearings made in China. There is nothing wrong with a properly designed and built cog belt reduction drive.

I initially had twelve gyro CFIs flying my systems and they built a lot of time on them. Ron Menzie put around 1400 hours on his RAF 2000 before he converted to the Sparrowhawk configuration.

Jim Logan put many hundreds of hours on his ship with my drive before RAF came out with their EJ22 package as did Curtis Patten and Duane Hunn. Jim is still using the 1991 EJ22 ECU and harness that I built for him. I think he has around 3800 hours on his RAF 2000 now.

CFI Thomas Gage put 1400 hours on his Air Command Tandem powered by an ER27 six cylinder before he quit flight training about five years ago. In fact, Thomas lost a coolant hose after he put a few hours on the engine package and flew for five minutes back to the airport with the temp gage pegged out.
He replaced the head gaskets and never touched the engine again after that incident.

I have modified over 300 EFI harnesses on all models of the fuel injected Subaru engines. The only failures that I know of were minor and only one caused an unplanned landing. That was the one that had the corrosion problem 12 years after I built the package for the pilot.

Subaru engines are reliable (and inexpensive) if you put things together yourself which is pretty easy on gyroplanes. Airplanes with cowled engines is a whole other story. You really have to do your homework, especially in the cooling department.

I have been retired for a couple of years, but my reduction drive design is being revived by a kit aircraft company that will be offering firewall forward EJ25 engine packages.
 
Don,

Thank you for contributing to this thread!

I follow your excellent soob articles in Homebuilt Rotorcraft Magazine.

I will soon be flying one of your conversations. I bought from Terry Eland.

I wish I had gone with your EFI on my first engines.

Can you tell us what company will be producing your redrive?
 
Top