Air Intake mods - More Sawsall Art

I continue to plug away at programming the EC3 and things are getting better. The mixture plot is getting smoother and the engine is running better.


I have a fifth fuel injector, which is used as a back-up, but it barely ran. I found that it was over-sized, capable of delivering 87 lbs. of fuel/hour and the mixture was extremely rich. This seems to be out of the adjustment range of the EC3 so I ordered a 55 lb./hr. injector. It will be here next week, so in the meantime I am holding off on tuning the engine and will tend to a few other items on the list.

All along, I have been planning a radical change to my combustion air intake. This was successfully accomplished by Harry Hinckley and Greg Zimmerman in Greg's FAST Swearingen SX300. There will be a little scoop in each plenum snout. This air will be directed, through two hard tubes, to the throttle body. Here is a photo of the mock-up I made last year. I have since made the pipes, now I just need to get the scoops placed in the plenum.


As usual, I start with pouring in 2-part foam, held in place with paper dams & tape.


Then I get out my favorite shop tool; Mr. Sawsall:

I use the 14" saw blade to gently remove some of the foam I didn't want.

After rough trimming, I attached a couple of pieces of hard foam, using 2-part foam as glue. The hard foam will make a better foundation for the lip of the scoop. Some wire pins hold the hard foam in place while the 2-part expands.

Since it's cool in the hangar, I'll wait a day before I get into the detail carving.

This may all seem hare-brained, but what the heck. That's what experimental aviation is all about.

More forwards and backwards

The closer I get to the goal, the farther away it gets - at least, sometimes it seems that way.

My current challenge is getting the electronic timing correct, then to program in the mixture "map". To map the mixture settings with the RWS EC3 Electronic Control Unit (ECU), you will input a setting for each 1/2" step of manifold pressure; from idle, all the way to full throttle. You use the fuel/air gauge to determine what the mixture is, at that throttle setting.

This ultimately provides a look-up table for the ECU for any given throttle input.

Before I do that mapping process, I need to get the base ignition timing right. This consists of making a sole timing setting, the "base" timing. In my case, since I am using 9:1 pistons, it will be 25 degrees BTDC, with the engine at full throttle. By simply pushing a button, you use the ECU programming panel to advance or retard the timing.

So that's the theory. The reality of it all becomes more difficult, at least, for me.

First thing is the working conditions. While southern Illinois is hardly the coldest place in the nation, I am getting tired of winter.

I first tie the tail down to a front wheel of my car to keep it from going anywhere. In order to perform the timing, I have to beg a volunteer to come out and run the throttle for me.

In the above photo, (click to see it bigger) you can just see the toothed timing wheel, just aft of the propeller spinner flange. I mark some of the teeth for timing reference. Then I have to hook up a timing light and have the volunteer run it up to full throttle. I lean over the engine and hold the timing light within two inches of the spinning buzz saw. Since it's pretty bright outside I even have to shade the timing marks with my other hand so I can see the strobe flash. I actually kind of rest the side of my hand on the aft end of the spinning prop extension. I'm not sure that OSHA would approve.

I have to wear two jackets, goggles, hearing muffs and a hoodie to keep myself together. I calculated that at 35 degrees and 100 mph of prop blast, the windchill is 14 degrees. But that is really only half the fun. The other, is the sensory kick I get from the combination of the prop blast, the major engine vibration and noise, the snot blowing out of my nose fogging the goggles, and the inherent danger of working around those spinning parts.

So, after repeatedly (10 times?) doing all that, we found a few discrepancies. One, I made the sensor mounting bracket wrong and will have to re-do it. Two, for some unknown reason the engine runs worse when the timing is in the right place. Three, my digital tach bounces around so much I'm not really sure what it is telling me.

So whine, bitch and groan.

But, at least I keep working on it, and I guess that's progress.

Thanks, today, to John Eynon for patiently running the throttle for me.

Closer to the Goal

The recent effort has been directed towards getting the engine to run properly, with the RWS EC3 system. There are all sorts of variables to work through; new throttle body, injectors, homemade induction pipes, the list is long.

Right now, I'm working through getting the mixture and ignition timing properly adjusted. This is a pretty big deal and will take a while.

To complicate things, the readings from the mixture meter have been unreliable. Here is the one I'm using.

It turns out, I mounted the sensor too close to the end of the exhaust pipe. The O2 sensor is very sensitive and I guess that ambient air is getting sucked back into the exhaust pipe, between pulses, particularly at lower RPMs. This winds up showing the mixture to be too lean. As suggested, I clamped on a piece of pipe I had laying around and the mixture meter reading instantly got better.


Again, the use of this meter and sensor is only temporary, until I get everything calibrated.

Right now, I'm running the engine without the canopy installed, which is giving me a real wind-blown hairstyle. The best I have seen, at full throttle is about 2,200 RPM and 29" hg. I have old data that I had attained 2,340 RPM and 29", so I have little way to go.

I'm getting some of the wiring tidied up and pretty soon it will be time to, hook up the rudder cables, install the back-up battery and install the canopy.

On another successful note, my custom designed and built fuel flow circuit board works just great. The Dynon D180 allows you to connect a flow transducer. These transducers typically have a little fan in it, that rotates as fuel flows. A sensor picks up the pulse count and sends it to the D180 and indicates flow in gph. The problem with my system is that fuel passes through an open loop, starting at the header tank, running through pumps and filters, past the fuel injectors, and then back to the header tank. Because of this, any fuel flow transducer will sense ALL the fuel flowing and not just the fuel consumed.

Note: the flow out of the injectors is controlled by the duration of time that the injectors open. More fuel demand - the injector pulse width increases. Less demand - narrower pulse width. The number of pulses per engine revoultion is fixed.

Mike Dwyer to the rescue. Mike is a Dynon dealer and an electronics whiz. Mike designed a circuit that takes a signal from an injector. It converts an increase in pulse width to an increase in pulse frequency. It's all black magic, to me.
I got with Dennis Hannon, the avionics instructor at SIU, and an avionics student, and they fabricated and bench-tested the converter. I named the piece the Dwyer-Hannon Frequency Converter.
And what do you know? I works! Fantastic!


Mike designed in a calibration potentiometer, but the Dynon also has an adjustment multiplier. There is a little formula to dial in the calibration, so in a short while my fuel flow indicator will be correct.

Sweet!