"Lean the engine until the EGT needle reaches its maximum temperature," instructors intoned, as student pilots gently pulled knobs or moved levers, "and then move it back until you are running 50 degrees on the rich side of peak temperature." The occasional inquisitive pup might ask why this is done. Instructors would warn of toasted valves, burned spark plug electrodes, holed pistons, and engine failure.
And all pilots learned through experimentation and experience that the Continental and Lycoming engines on their airplanes began to run rough around the point of peak EGT. They ran especially rough if one continued to lean the mixture past peak temperature-the dreaded lean side of peak. Roughness suggests engine failure; passengers get wide-eyed and pilots feel their palms getting moist. No one asked why these four- and six-cylinder air-cooled engines ran rough when leaned. Here be dragons, said the conventional wisdom; just don't go there. But George Braly, who bought a Beech Bonanza in 1991 and shortly thereafter installed an instrument to measure the EGT for each cylinder, noticed that when he pulled the knob that leans the mixture and reduces the fuel flow, the six cylinders of his Continental IO-520 engine reached their peak temperatures at widely scattered points across that range of motion. Why didn't they all peak together? he wondered.
On Compuserve's online aviation forum, pilots of all stripes-and those with none-could debate freely and anonymously the precepts of their training. In 1991 Braly began wondering about engine mixture management in messages to John Deakin. In an e-mail, Deakin recounts that time: "It took Compuserve's AVSIG [AViation Special Interest Group] to bring us all together and serve as a catalyst." Braly led the way, Deakin recalls, "with the rest of us asking questions he could not, at first, answer. Drove him nuts, so he began (in about 1994) the long, long trail that leads to today."
Braly says that the prevailing opinion of the time was that the peak EGT spread he saw on his engine was attributable to the design of Continental's induction system-that there was something wrong with the airflow (it's actually quite good). But mechanics adjusted the fuel injection systems on these engines on the theory that the airflow to each cylinder was equal and perfect. Using four or six containers (often cola bottles, resulting in the coinage "Coke bottle test") to catch the gas and determine the volume delivered, they would carefully tweak the system until it was metering precisely the same amount of gasoline through each injector to its respective cylinder.
Continental engines use continuous-flow fuel injection systems: The injector spritzes fuel in a flow as steady as a garden hose, even when the intake valve to the cylinder has closed. Braly began to suspect that some of the fuel that accumulated when the valve was closed was making its way down the induction system to the adjacent cylinders. If he was right, some cylinders were getting the wrong amounts of fuel, and the variation would prevent all six cylinders from arriving at peak EGT simultaneously. And if the fuel flows that brought the cylinders to peak EGT were different enough, the power outputs from all the cylinders would differ at leaner mixtures, where the power falls off quickly. No wonder the engines ran rough.
Maybe fuel distribution to each cylinder shouldn't be equal. Maybe it should be different.
By 1993, Braly had teamed up with an Ada-based parts manufacturer, Tim Roehl, to form General Aviation Modifications, Inc. He and Roehl began to experiment with injector nozzles calibrated to deliver fuel at a rate precisely matched to the needs of each cylinder. They had help from new microprocessor-based systems that displayed in monkey-simple graphics all the important engine data: exhaust gas temperature-not just for the engine, but for each cylinder-cylinder head temperatures, and, for turbocharged engines, turbine inlet temperature.
Braly started looking for an expert to help with the process of getting a Supplementary Type Certificate from the FAA for GAMI's new injector. Someone recommended a Texan named Carl Goulet. "I had no idea at that time that he was the former head of engineering at Teledyne Continental Motors," Braly recounts. "We had a very short and very remarkable conversation. I was considerably his junior, and he said, 'Now young man, tell me what you plan to do.' I told him, and these were his exact words: 'Hot damn! Somebody's finally gonna fix this problem!' "
GAMI applied for an STC in 1996, and the FAA approved it in the same year. At Goulet's suggestion, Braly submitted a proposal to Teledyne Continental Motors offering to supply fuel injectors and provide customer support. Hearing nothing, GAMI began to market the modification to pilots. Deakin and Atkinson were among the first to install the new injectors in their engines.
Atkinson remembers that after getting the injectors installed and heading home in his airplane, he leaned the mixture by ear, the way he'd been used to. Ignoring the EGT gauge, he liked to pull the leanerator until the engine ran rough, telling him he was just lean of peak, then adjust the mixture from there. "Except this time it wouldn't run rough," he recalls. "I kept pulling it back and it just kept running smoothly."