The Engine

The first step in the planning "Firewall Forward" as it were, is choosing the engine. At first I planned on using the Rotax 447, and then switched to the 2si 460F-35, as it was lighter. Due to the unending negativity by the ultralight community at large, I switched back, and chose the Rotax 447, Air Cooled version. The biggest problem with the 477 is the weight. With reduction drive, free air version is still 70 pounds or more, which exceeds the design weight of the motor mount, of 69 pounds. Plus, I would still have to pull start, which I did not want to do. Then I went to a fly-in where a few people had Kawasaki Engines. I called J-Bird in Wisconsin, and plan on using his Kawasaki 440A. The weight is good, the electric start will work, and most important: the price is right. I finally decided on the slightly more costly, but more suitable Hirth F-33. It is a 28hp replacement for the Rotax 277, and weighs less, even with, get this: ELECTRIC START! No cranking and then climbing into the plane for me. 

As Ison's help stops with the completion of the basic airframe, I will describe my work here in detail, leaving you to draw your own conclusions, make changes, or copy my ideas for your own use. I must warn you though, I am an amateur, and this is my first airplane. Therefore my advice and experience may be faulty, and if you steal my ideas and use them in your own plane and subsequently crash, don't come crying to me.

Motor mount in hand, I have put the engine purchasing off until later, when I really need it. For now I will plan the system based on that motor, and I will not have a perfectly good engine gathering rust and dust in the corner.

Step two is the fuel system. I have decided to mount a Westach fuel gauge sender in the fuel tank, and put a gauge into the panel itself. The actual mounting of the sender is rather difficult. I drilled a 5/8" hole in the top center of the tank for the sender probe, and using the sender as a guide, made five 3/16" holes in the tank. Next I cut the probe to 6" and put it into the tank to check the fit. After calibrating the sender, I sat it and the gauge aside and ordered some AN3-11A bolts and stainless steel stop nuts. Actually, I ordered a bunch of different sized nuts as I did not know the correct size, but if you try it, AN3-11A work good. Using some permatex sealer (non-hardening form-a-gasket from the aircraft section of NAPA auto parts), I coated the cork gasket, tank top, and sender bottom. Next I stuck a wrench into the filler hole, with the stop nut taped to the end. Looking into the gas line hole in the bottom of the tank, I was able to thread the nut on the AN3 bolt, and tighten down. Five AN3 bolts hold the sender on tightly. Afterwards, I covered the bolt tops with permatex, along with running a bead around the side of the gasket and sender. It should prevent any leaks. I used the stop nuts without the nylon insert. I do not know how resistant nylon is to gasoline, and the stainless steel should prevent any corrosion. I cut a four inch diameter hole in the top deck over where the sender will go. It will allow access to check the sender, and to run the wires up to the rest of the plane's electrical system. There are four wires coming off the sender. Two of them hook to a terminal block, and the other two go to the fuel gauge cable, which is removable from the rear of the gauge, which will allow me to easily disconnect the sender and tank from the rest of the system if the tank needs to be removed. Sure, a sight gauge is cheaper, but I prefer a electric fuel gauge, than to have a sight gauge. The position of the fuel tank will also allow me to see when the tank gets close to empty, just in case the gauge fails.

After sender installation, I put permatex on the rubber bushing that goes on the bottom of the tank, and also rubbed permatex on the push-in elbow for the tank. It was tough to get in the hole, but finally popped into place, and seems to be sealed well now.

Fitting up the Sender

Step three has been designing the rest of the fuel system. Fuel delivery to the engine needs to be reliable, and as simple as possible. I took into account several things. 1. As few fuel lines as possible 2. Reliable shut-off, not easy to accidentally shut off. 3. Ease of inspection 4. Low point drain to aircraft exterior 5. Meet or eliminate fuel pump weep hole requirements 6. Reliable metal connections - no plastic

I know some detractors will say, "What about your pulse-line driven fuel pump?" I had decided against it, but at an engine company's persuasion, I am going to put one in. I am rerouting my fuel lines, and added an extra hose connection to fit the fuel line.

My first fuel system design had the lines running all over the place, with the fuel pump mounted on the floor under the engine. After thinking a while, I changed my fuel system design. Sitting in the plane while planning it all helps a LOT. I made a new Starboard side fuel tank support, with the back extending about four inches. The fuel pump was supposed to mount to that but the tabs are 90 degrees in the wrong place. I mounted it on a scrap block of wood, on the starboard side of the plane. Coming off the fuel tank is a short line to the pump. Then there is a brass 90 degree elbow screwed into a brass "street" tee, one side going to a drain valve in the floor, and the other screwed into the inlet side of a facet electric pump. The outlet side of the pump goes to the primer line tee, and then to the fuel on/off valve. The fuel line then runs up, through the top deck, past the primer knob, and forward to the firewall.

My drain port is simpler, as well. I will use a sleeve and the drain valve to clamp a large washer to the outside of the plane, and a thin steel plate to the inside. No screws necessary, and the drain will not interfere with the fabric application later on.

All my fittings and tees are made of BRASS. There is not a nylon or plastic piece in the system, to guard against failure.

The first design is pictured here, drawn in perspective as best I could while leaving out the plane structure. The fuel pump was mounted to the fuselage side, out of the way of where my leg will be. I sat in the plane to figure proper placement. The fuel filter is a purolator replaceable-element type, easily visible. I also mounted the primer and fuel gauge onto the panel, and wired them up. This completes preliminary fuel system installation.  The final installation moved the valve directly after the filter, and adds another tee between the pump and drain. The second tee sends a line to the pulse-pump, hooked in parallel to the electric pump. As the Hirth only has a 50 watt coil, the electric pump will be used for takeoff and landing only. Otherwise I will rely on the pulse pump. If it craps out, I can switch to electric as a backup until I can safely land.

As the fuel valve I ordered mounted backwards from what it should have, the fuel line had to wrap around itself, leaning towards the possibility of kinking. I redesigned the system, moving the valve as close to the tank output as I could. I also replaced the tank push-in fuel thing with a set of brass elbows. A female elbow is inside the tank, with a male elbow extending through the hole in the bottom. This was a tough job, seeing as how I had to look through the filler hole to do everything, and could not stick any tools into the tank itself. Most of the brass parts I used can be found at ACE hardware, a lot cheaper than aircraft spruce sells them.  The mounting block and fuel pump support all the brass hardware from one end. prior to final assembly, rubber washers will be added to the fuel pump base. Hopefully this will silence some of the fuel pump-to-fuselage noise, and reduce stress on the mounting block and screws.  The only disadvantage of this system compared with the other one, is the fuel valve cannot be operated with the seat belt on.  But, with the valve as close as possible to the tank, and not after the pump, it is a compromise. The pulse and electric pump fuel  lines meet after the pulse pump and one line goes to the carb from there. My pumps are paralleled.

The installed fuel system is visible here, before I plugged all the hoses in place. After this picture was taken, I redid the hoses, so they are much simpler, and direct. Pictures are forthcoming.

Here is the revamped system, without some of the new hoses in place. The tank line runs to the fuel valve, and the to a tee. One line runs down to the drain port, the other, up through the bulkhead to the mechanical pump. The tee also leads into the Facet Electric pump, and out the other side is the line to the engine, which has been tapped for a primer line. You can also see the block where the fuel valve used to be. I made a new angle, and will mount a few switches there, as well as the plug for the helmet. The clear line is the static line, which runs under the seat to the tail section. The scraped up wood block in the foreground used to hold the fuel valve. The bracket will be modified most likely to hold the radio, or at least the headset plug-in jack.

After the fuel stuff, it was time for the electrical work. I decided to go with a battery as well as a rectifier. The battery will provide an important back-up to the fuel pump if the rectifier craps out. Redundancy is good.

The first step was picking parts, and deciding mounting locations. The terminal block is already mounted between the fuel sender hole and the panel. The regulator will be mounted beside it. I am using a KW voltage regulator (CPS # 9105), which does not require a minimum load to operate. It will operate the radio, GPS, gauges, trim, and fuel pump, while keeping a battery charged for emergency backup. On the drawing below you can see a diagram of my electrical system. It all starts with the regulator, on the left. The hour meter has been replaced with a Tiny-Tach, which is self powered by a small battery. Next is the Regulator Failure Relay. It is a small relay, whose function is only to alert me if the regulator craps out. The diode at the top of the drawing keeps the relay from drawing battery power. The reason it is 35 amps is that is the closest size stud diode I could find. If you find one closer to the size you need, use it! The Regulator can put out up to 15 amps, so you need a diode that size or better. With the Hirth Engine, I will be limited to about 3.5 amps. The diode should outlast the plane at that much of a difference. The relay will only close if the regulator is working. When the fuel pump is switched on, the engine will be off, and the relay will be open. A LED, hooked to the Normally Open side of the relay, will then get current from the battery, and will glow red. The engine will be started, and as soon as the regulator is producing power, the relay will close, switching the LED off. IF, at any point during the flight the rectifier or coil dies, the relay will open, and the light will come on. It will let me know instantly if the is a problem. Of course, if the wrong section of the coil dies, the engine quit anyway. 

The fuel pump switch controls the electric fuel pump, and the power for the push-button start switch. Therefore, the motor cannot be started without the fuel pump being ON. This prevents embarrassing and expensive "oopses" where one would have the motor die from fuel starvation during takeoff. Instead of directly switching the starter, another relay is used.

A line jumps from the fuel pump side of the fuel pump switch to the radio power plug. This plug is isolated from the main bus by a 3 amp diode. If the fuse blows or for any reason the master switch is cut off, the radio and fuel pump will be the only things running off the battery. This line may be eliminated in the final design. As of now, my GPS runs off its own 2 AA batteries, but will soon have a cable to connect to the plane's electrical system, as the batteries only last ten hours or so. The tiny tach is also self powered, as are the CHT and EGT gauges. The only other thing which would run off power would be a fuel gauge and a voltmeter, which don't use that much power.

The design below is my own, although based on schematics I have seen in both the Sport plane Resource Guide volume 2, and the CPS catalog's "care and feeding of the Rotax motor". Feel free to use any ideas here, and if you have any ideas or need some clarifications, e-mail me.

This is the beginning of the wiring, which is seen above. The black plastic box houses both the regulator failure relay, and the starter relay. The little blue box is the regulator relay, and the white object with the coil is the start relay. The large stud diode is hard to see, but it is beside the blue relay. An AN nut holds one side to the breadboard, and the other side is mostly hidden because of the picture angle and the breadboard. When complete, a plastic lid fits on this box, holding everything in place. The wires look a little messy right now, but will be cleaned up and tied together later. Two terminal strips were placed beside this box, one for ground and one for positive. You can also see the red pushbutton start switch, beside the fuel gauge.

Here you can see further work on the electrical system. The terminal strips are in place, and most of the wires are hooked up. The thin loose wire goes to the push-to-talk switch, with the other wires near the top of the picture going through the deck down to the fuel pump, and on to the battery. The blue fuel line and primer line is visible as well, and the clear line coiled up will be for the static port. In the background are the blue fuel lines, coming through the bulkhead in their 90 degree elbows.