Chapter 23 - Engine - it's much better to follow than to lead

Greg Richter's turbo 13B Cozy III What follows is my personal impression. See Greg's Richter's web site to read the owner's own description and see pictures now.

These notes are for my own use and may help others considering or pursuing this path. But....

BEWARE. As is common in my scribblings, there is at least one fact below which is totally incorrect. I don't know which one it is, and neither do you - so take everything you read here with a pinch of salt...........

A trip to Merrit Island

In late May '01 I contacted Greg Richter, arranged to see his plane and pick his brains. The two hour drive to Merritt Island was well rewarded. Greg's turbo 13B powered Cozy III had about 18 hours on the hobbs and was waiting for a prop repair. Unfortunately they'd just applied some micro around the canopy, so we couldn't open it. No big deal. I'd come to see the engine. I've seen quite a few 13B airplane installations, but this one takes the prize, and then some. In short it's done right. No amateur welding, no rubber hoses - well one, but Greg plans to replace it. Braided stainless Aeroquip hoses and AN fittings throughout. With respects and apologies to the pilot we all know and love who's currently flying with a plywood plenum, I think it's safe to say that Greg's installation is the other end of the scale. As good as it gets. In Greg's words, "This is not an automotive conversion, it's an experimental aviation engine." The plane will be at OSH, booth 40 I believe, showing off the EFIS which Greg's company has developed. I'd be very surprised if N722 doesn't win a prize or two.

Overall size considerations

Greg's plane is a Cozy III, not a IV. I hadn't realized how much difference there is in the width at the back. The distance between the wing roots on the III is 46 inches. The distance on the IV is a full 18 inches more at 62. The distance from firewall to prop flange on Greg's installation is 36 inches. Mine is 42.5. The net result of this is that I'm going to have all kinds of room to fit the oil coolers, radiator, intercooler, AC evaporator etc. Greg had to jiggle quite a bit to get everything under his cowl (including jiggling the cowl itself). I don't think I'm going to have that problem.

Engine, computer control & mounting

13B (two rotor) '93 REW R8 twin turbo model with standard turbo (low compression) rotors. Computer control is via a Tracy Crook, Real World Solutions EC2. Apparently Greg and Tracy went through a few programming iterations to get the software right. We didn't discuss any internal mods like 3mm rotor seals, teflon oil seals, porting etc. The front housing (and the crank angle sensor) is from an '86 model 13B which has hard points. Mounting is via a sump plate for the rear points, and the front housing hard points to dynafocal type rubber mounts and then via a steel angle to custom lower hard points in front of the engine, and the plans top hard points on the firewall. The lower plans firewall hard points are not used.

Fuel System

Direct 1/2 inch feed from each tank to 1 gallon fiberglass header tank in front of firewall. Two Tracy Crook type EFI fuel pumps on the firewall plumbed in parallel via a custom milled aluminum rail, through a nice looking aluminum filter and on to the stock return valve. Stock rails (drilled & tapped for AN fittings) and high capacitance injectors. Fuel return to the header tank. No fuel off switch. Stock intake manifold cut and welded. Air intake to turbo via a small air filter on the firewall - too small - Greg plans to enlarge it.

Water Cooling System

Roughly 10 * 18 * 4 all aluminum radiator (from HKS?) by the port side wing root. Aeroquip braided stainless 2 inch 601 hoses with dash 24 AN 816 fittings. The rad is fed by a large armpit scoop. Greg feels the scoop is too large and hopes to reduce it considerably. The system holds only 1 gallon of coolant. A stock thermostat is used with a 16 lb pressure cap. A small plastic overflow bottle is attached high on the firewall, slightly to the port side. Greg says he will report performance details on his web site when he has enough hours to make them meaningful. In summary - it works well in normal operation, but isn't sufficient for excessive taxiing on a hot day. Greg has tried using an automotive type radiator fan, but found that cruise cooling was impaired and removed it. Stock cut off water pump.

Oil System

One stock Mazda oil cooler near the starboard wing root. All Aeroquip 601 braided stainless pipes with dash 12 816 AN fittings. To reduce pipe lengths to both, Greg would have preferred to put the oil cooler on the port side and the rad on the starboard side, but the turbo on the starboard side and room in the Cozy III cowl precluded this. High pressure oil feeds to turbo and redrive are Aeroquip. The lines are T'd at the turbo and run on to dual oil pressure sensors, one for the EFIS, and one for a standard backup gauge. Stock oil filter and housing. The oil cooler uses an armpit scoop on the starboard side. This scoop is, obviously, the same size as the one on the port side and will eventually be reduced in size.


Be really careful when reading this - I have very little idea what I'm talking about when it comes to turbos. But I'm learning....

First - why use a turbo when it's better to KISS? Well, the logic is simple and compelling. Its very hard to deal with the sound pulses from a turbo 13B engine [the best choice for our purposes] with a muffler that's close to the engine and inside a cowl. The mufflers tend to fall apart or burn up. Those that don't tend to be heavy. Additionally, the radiation from a inside cowl muffler can be a real problem. The turbo solves this by absorbing the sound pulses, while doing the job it was designed for (; and, incidentally, delivering an extra 100 or so HP ;). Once the turbo is done with the exhaust, Greg dumps it out of the upper side cowl through a three inch J pipe. I didn't hear the engine run, but I'm told that the plane is the quietest on the field. [Later note: Greg started the engine for me on a later visit. I was amazed at how quietly it runs without a muffler. Nowhere near as loud as a Cessna or a motorbike. The sound of a turbocharged 13B is somewhat different. A combination of whine and rumble with variations depending on throttle setting].

Feeling that the stock turbo wasn't up to the task of continuous 6000 RPM use, Greg chose a T04 turbo from Turbonetics. [Note: I later decided to try the stock single turbo - see Chap 23 - Engine]. This unit is NOT cheap. At around $3k it came with an inconel turbine wheel and ceramic bearings. I don't fully understand the science yet, but the turbine wheel design and size of the inlet is carefully specified to fit the engine's precise requirements. The turbo is quite small, doesn't weigh much more than a muffler, and fits nicely against the stock cast iron exhaust manifold. It has a heat shield around it - another part you can just order from Turbonetics. Air from the turbo is channeled via custom welded 2 inch stainless piping to a stock intercooler mounted under the engine and then back to the inlet side. This stainless pipe is connected to the inlet with 2 inch blue inch silicone hose. The intercooler (For those who don't know - an intercooler is simply an air to air radiator) is where you might expect the rad to be (and will probably find it on my plane). This was due to space constraints in the Cozy III cowl. Cooling air is fed to the intercooler by a large center scoop. There is no NACA scoop, but I'm not sure why. Perhaps this was to accomodate the header tank. All three scoops have large carefully designed plenums and rubber baffles to channel the air where needed. Interestingly, the turbo has a one position wastegate. Shut. This solves the problem of how to silence the wasted exhaust, but means that the turbo is running constantly. Considering that most of our engine use is cruise, this doesn't really make much difference, but it does provide another good reason for going with the inconel turbine wheel and overall beefed up turbo unit. a view of the sealed wastegate The pop-off valve is currently mounted at the inlet manifold. Greg plans to move this to the intercooler because (he says) there's no point in cooling the air, then dumping it. Boost at take-off is currently limited to 8 - 10 lb. Greg hopes to increase this to 10 - 12 lb.

A few other tidbits

The alternator came from an '81 diesel Chevette. It has a dry vacuum pump incorporated into it for the vacuum instruments. Greg says he tried the stock smog pump but it didn't produce enough vacuum. He was surprised to hear that other installations have used the smog pump successfully and wonders if he had a bad pump. Battery is a Concorde. Coils, plugs and wires are stock Mazda. Redrive is a Ross, but Greg has an RWS 6 pinion on order.


A 68" three blade creation from Margie Warnke optimized for 300 HP somewhere between take off and cruise. It needs a bit more "bite" because he can overspeed it. Nice looking prop, but not cheap.

Overall performance

Greg tells me his take-off run "feels like a Porche" and he's had the plane up to 190 kt already. He flew a Cozy IV with an O-320 and aborted the take off because he thought something was wrong with the engine. There wasn't. It's early days for N722 with just 20 hours on the hobbs. Greg has promised to give us all some solid figures once he has proven the installation by passing the 100 hour mark.

More poor quality pictures

Where to go from here

Looking at Greg's installation was a bit overwhelming. On the other hand, I'd probably be overwhelmed by a Lycoming installation right now. I know fiberglass, but this engine stuff is still a bit of a mystery. As with all big problems, the best way is to break the task down into smaller parts.

Step 1. Buy a $10 thread gauge
Step 2. Get catalogs from Lewis Marine, Aviall, Turbonetics and Summit Racing.
Step 3. Get the engine mounted.
Step 4. Jury rig the engine and get it running on a false firewall.
Step 5. Position the chosen oil cooler(s), intercooler, rads etc. on the firewall mock-up
Step 6. Compile a list of the AN fittings and pipes needed to connect everything.
Step 7. Spend money
Step 8. Fit it all together
Step 9. Start the sucker up and fly away.

Well, ok. Steps 7 thru 9 are a bit sketchy right now. I'm only up to step 6. I'll fill the rest in as I go along.

My thanks to Greg for taking most of the day to show me his plane and explain his thinking. It's this sort of openness and camaraderie that makes the world of experimental airplanes so much fun. What goes around comes around. One day, when I know what I'm talking about, I'll help someone else.

Regards, and encouragement to all who are thinking of breaking the mold.
John Slade, Hindmost