The Cessna 340 had just leveled off at its requested cruising altitude of FL240 when all of a sudden the left engine started running rough. The engine was rough enough to really get the attention of the airplane's seasoned owner-pilot.
Being a maintenance-savvy sort, the pilot attempted to troubleshoot the problem. His first thought was that it had to be either a fuel-related problem or an ignition-related problem, so he attempted to determine which one of these possibilities it was.
First, he tried enriching the mixture, but the roughness continued. That suggested it might be an ignition problem.
Next, he performed an in-flight mag check, but found that the engine continued to run rough on each magneto individually as well as on both mags together.
Finally, the pilot called ATC, reported that he had a rough-running left engine, and requested a lower altitude. The center controller was happy to accommodate, and the pilot started his descent. As the airplane approached FL220, the engine roughness seemed to disappear, so the pilot leveled off and asked ATC for FL220 as a final cruise altitude.
The rest of the flight proved uneventful, except that the pilot discovered that the left engine started running rough again when he attempted to adjust the mixture to lean-of-peak (LOP). So he continued at FL220 with the left engine rich and the right engine lean, and arrived at his destination with less reserve fuel than planned.
After landing, the pilot downloaded the data from his JPI engine analyzer and emailed it to me, asking if I would look at it and help him figure out what was wrong with his left engine. The engine analyzer data clearly showed the rough-running episode in the form of erratic EGT spikes that smoothed out considerably after the aircraft descended from FL240 to FL220. The EGT spikes were visible in all six cylinders, strongly suggesting that whatever was wrong affected all cylinders, not just one or two.
Sure looked to me like a classic case of high-altitude misfire.
Flying at high altitudes places extraordinary demands on our ignition system, particularly up in the flight levels. To understand why, we need to brush up on the physics of electrical sparks.
Our engines employ magnetos that generate high-voltage pulses. Each pulse is directed to a particular cylinder by a mechanically driven distributor in the magneto. The pulse is conducted through a wire in the ignition harness to a spark plug. It then jumps the air gap between the spark plug's center and ground electrodes to produce a spark that ignites the fuel-air mixture in the cylinder's combustion chamber.
How much voltage does it take to jump the spark plug's air gap? Well, that depends on two things: the size of the gap and the pressure of the air.
The relationship between dielectric breakdown voltage, gap size and pressure is described by Paschen's Law (after a German scientist F. Paschen who discovered it in 1889). For small air gaps in the millimeter range, the relationship can be approximated as:
where V is the breakdown voltage (in volts), P is air pressure (in atmospheres) and D is the air gap distance (in millimeters).
If you were to test a aircraft spark plug on a bench at sea level, and if the spark plug at a normal gap size of 0.016 to 0.021 inch (0.4 to 0.5 mm), you'd find that it takes about 2000 to 3000 volts to fire the plug.
Everything changes when we install that same spark plug in an aircraft engine running at high power. That's because the air pressure inside the combustion chamber at the time that the spark plug fires is on the order of four times as much as sea-level atmospheric pressure, thanks to compression by the turbocharger and the piston. To fire the plug in this higher-pressure environment can require 7000 to 8000 volts. That's why our magnetos are designed to produce up to 20,000 volts at maximum RPM.
When a magneto generates a high-voltage pulse, we want that pulse to jump the air gap between the electrodes of the spark plug. What we don't want to happen is for the spark to occur anywhere else in the system -- such as inside the magneto. Such an undesirable spark is called an "arc-over" and results in what we call "misfire" (see graphic at right). To ensure that the spark occurs where we want it to occur, we must make sure that the spark plug's gap represents "the path of least resistance" for the high-voltage pulse generated by the magneto.
Here's the problem: As we climb to higher altitudes in a turbocharged aircraft, the air pressure in the combustion chamber remains relatively constant (thanks to the turbocharger), but the air pressure in the magneto decreases. Paschen's Law tells us that as the air pressure in the magneto decreases, it becomes easier and easier for the high-voltage pulse to arc-over inside the magneto rather than across the spark plug gap. If we climb high enough and the air pressure in the magneto becomes low enough, we'll eventually reach the point where arc-over will occur inside the magneto, causing high-altitude misfire.
Trust me, if this occurs, it will really get your attention! (Been there, done that, got the t-shirt.)
If you ever experience high-altitude misfire in flight, the first thing you should do is reduce manifold pressure. This will reduce the combustion-chamber pressure, and make it easier for the spark to jump the spark plug gap rather than inside the magneto. Your next move should be to descend to a lower altitude, thereby increasing the air pressure inside the magneto and raising the breakdown voltage.
When you get back on the ground, ask your mechanic open up the mags and inspect the inside of the distributor gears and blocks for signs of burning and carbon tracking. Carbon tracking must be cleaned off, and components that exhibit signs of burning or overheating should be replaced.
To prevent high-altitude misfire, we need to make it easier for the high-voltage pulses to occur where they're supposed to occur (at the spark plug), and more difficult for them to occur where they aren't supposed to occur (inside the magneto). How can we accomplish this?
To make it easier for the spark to occur at the spark plug, we need to keep the spark-plug gaps tight. Most aircraft spark plugs have specs that call for a gap between 0.016 and 0.021 inch. Keeping the gaps at the tight end of that range (0.016 inch) reduces the voltage required to fire the plug and provide increased margin against high-altitude misfire. Of course, the spark plug gaps get bigger as the plugs wear in service, so it's important to clean and re-gap the plugs on a regular basis. I suggest you do this at least every 100 hours; if you regularly fly into the mid-20s, you might need to do it every 50 hours.
There are two ways to make it harder for the spark to arc-over inside the magneto. One is to use magnetos that are as physically large as possible, reducing the chance of internal arc-over because the electrodes in the distributor are so widely spaced. For example, the huge TCM/Bendix S-1200 "tractor mags" that I use on my airplane have distributor block electrodes that are spaced 1.2 inches apart, so they're much more resistant to high-altitude misfire than the smaller TCM/Bendix S-20/200 mags and the even-smaller Slick 6300 mags also approved for my engines.
The big S-1200 mags also produce a hotter spark, and seem to be more durable and trouble-free than Slicks. Of course, they're heavy, and physically too large to fit some engine installations.
The other way to minimize the chance of arc-over is to pressurize the mags by pumping bleed air from the turbocharger into them (see figure above right). RAM Aircraft, for example, fits pressurized Slick mags on all its TSIO-520 engines. For really high altitudes, a pressurized version of the big TCM/Bendix S-1200 mag -- the S-1250 -- is available, and used by RAM on their GTSIO-520 engines used on the Cessna 421 that flies up to FL280.
Pressurized mags are a mixed blessing, however. Although the pressurization is an effective way to eliminate the high-altitude misfire problem, it also creates a new problem -- internal contamination of the magneto -- particularly when flying through moisture (rain or clouds). As a result, pressurized mags need to be opened up and cleaned a lot more frequently than do non-pressurized ones. In fact, Slick Service Bulletin SB1-88A recommends a teardown and internal inspection of pressurized mags every 100 hours (compared with 500 hours for non-pressurized mags).
If you do have pressurized mags installed, make sure they receive frequent inspection and maintenance, and change the filter in the magneto pressurization line often. Both TCM and RAM offer improved green in-line filters that are more effective than the older clear filters in keeping moisture and contaminants out of the mags. The newer filters also include a sump and drain line for improved moisture elimination.
I emailed the owner of the Cessna 340, suggesting that he ask his mechanic to check the spark plug gaps on his left engine, and adjust them all down to 0.016 inch. Since the 340's engines were RAM overhauls equipped with pressurized Slick magnetos, I also suggested that the mechanic inspect the pressurization lines to the mags to make sure that they weren't disconnected or leaking.
Two days later, the owner emailed me back to report what his mechanic had found. All spark plugs had gaps in excess of 0.021 inch, and the in-line filter for the magnetos on the left engine had broken, causing both mags to become unpressurized. One of those mags had arced over so badly that the distributor gear was badly burned and had to be replaced (see photo below right).
No wonder the left engine was so unhappy at FL240! Fortunately, the fix was quick and relatively inexpensive. After re-gapping all the plugs to 0.016 inch, replacing the broken in-line filter and replacing the burned magneto distributor gear, the owner reports that the engines are now running smoothly at FL240 and LOP.
See you next month.
Want to read more from Mike Busch? Check out the rest of his Savvy Aviator columns.