I had thought for some time that the full potential of the Subaru engine would never be realized with the original Gen-1 PSRU. Those few hundred RPM shy of the peak torque seemed to be forever out of reach until the Gen-3 PSRU was developed.
Basic background is that the Gen-1 has a gear ratio of 1.82:1. At maximum prop RPM of 2700, that well defines the engine RPM limit of 4914. At this RPM, the theoretical engine output is around 140 HP. The unit itself, from what I understand, is some kind of transmission out of another application (some car or truck).
The Gen-3 unit works on the same basic principles of the Gen-1, except the unit was developed by Eggenfellner with some beefier parts to make it more suitable to the aircraft application. The gear ratio on this unit is 2.01:1, so again at 2700 prop RPM, the engine can now turn at 5427 RPM; much closer to the peak torque the engine was designed for at its theoretical rating of 165 HP.
I say “theoretical” in the above paragraphs because that is what is being advertised by Subaru for this engine. I really have no interest in paying for a dyno test to find the actual number, and even if I did, there isn’t much of anything I can do to alter the results. What I do know is that I should get a noticeable power increase. That’s enough for me.
When the units started going into production, I was really interested in an upgrade. It turns out that I would have to wait nearly a year before I could fork over a check and get my hands on one. The first couple of batches were small enough that the extras were spoken for by the time I had inquired about them.
So, finally in late 2007 I placed a firm order and got a position for January 2008 delivery. Interestingly, this was about the same time that Eggenfellner began to mandate upgrades for all fielded engines. Apparently two catastrophic Gen-1 failures out of the hundreds that were flying was enough to raise concern with the factory. There were a few maintainence issues also; some older drives had a weaker bearing that was known to fail, and a few units blew out the rear seals and leaked oil.
As for me, my Gen-1 lasted over 200 hours trouble free, and the only reason I upgraded was to get a few extra HP.
When my Gen-3 arrived, I have to say I was impressed. If the internals are half as strong as the outside looks and feels, this thing should last quite a long time. It definitely has a “purpose-built” look to it. Quite the mechanical work of art.
I have to admit that I was pretty nervous about pulling the PSRU. In the nearly three years I have been flying, I never had a reason to pull the prop off, let alone the redrive.
It turned out to be a fairly simple operation. The manuals that the Eggenfellner team produced were well detailed, and were improved fairly quickly once feedback started coming in from the field.
Below are some pictures and a brief description of the upgrade process.
The first task was removal of the propeller and old unit. This is a shot of the mounting plate with the Gen-1 off. The old spline shaft sat in the center of the flywheel and needed a good schwack with a hammer to pop it out.
Next, two bolts had to be removed. A long bolt on one side, and a stud on the other. In order to get the long bold out, a slot was cut into the aluminum mount plate using a variety of cutting tools. Several holes were drilled to remove most of the metal from the slot. Then, a lot of filing. The biggest problem with using the file is that you can’t go very deep without hitting the flywheel. Lots and lots of short strokes to open up a path to extract the bolt. The extra stud was removed in short order with a cutting wheel, then ground down flush with a Dremel tool.
Since the old spline shaft sat in the center of the flywheel, there was no pilot bearing. I special ordered one through the local Subaru dealer. I tapped it in with a hammer and block of wood to distribute the force evenly. It went in and sat nice and flush.
The new spline shaft is a two piece deal. There is an adapter plate, and the actual shaft. Pressing the shaft into the mounting plate was something of a trick. It required some influence with a hammer, which I was all too pleased to exert. The shaft extends rearward through the pilot bearing for proper alignment, and the adapter is bolted down. The brown stuff is some kind of copper based anti-seize compound applied by the factory.
Using two new solid dowel pins for alignment with the mount plate, the new unit slid right on to the original mounting studs and spline shaft. Only a little wiggling near the end was needed to seat everything flush. I did have to remove the starter to allow one of the mounting lugs to clear.
New bushings were purchased from Eggenfellner and installed into the flange. These were super tight. His suggestion was to make a “puller” by using a bolt through the bushing and through a largish socket on the front of the flange. This would allow you to tighten the bolt and pull the bearings which protrude through the front of the flange. After stripping out a couple of nuts and bolts, I used a six inch C-clamp with a stack of large flat washers behind the bushing, and a large socket on the front of the flange and just squeezed them in. Once that was done, the prop was remounted with some new and slightly shorter bolts.
I also purchased the factory brush mount because I was too lazy to make my own. Remounted the coil pack, rerouted the temperature sensor and brush block wires, and that was that! All that is left is a new dynamic prop balance and flight testing which will be written up seperately.