Unless left alone, GXRP’s magneto system has hickups sometimes. It tends to confuse mechanics who work on it.

As is typical of aircarft engines, each one on GXRP has a dual ignition system consisting of two magnetos. Each magneto is an electrical generator, attached to opposing sides of the accessory gear at the back of the engine, and to six spark plugs of the twelve per engine. During normal operation, both left and right magnetos are on, providing concurrent and redundant sparks in the combustion chamber of each cylinder.

During engine start, for arcane reasons, only one magneto per engine is turned on, but that is enough to do the job. During the pre-takeoff engine runup tests, we test each magneto individually. Because of the loss of one half of ignition sparks, the engine RPM drops somewhat. How much the RPM drops is an indication of whether the remaining magneto and its spark plugs are operating at all, and whether the magneto is timed to provide the spark at the right moment.

The right moment is defined as a fixed number of degrees before top-dead-centre of the crankshaft rotation cycle – 25 degrees in the case of GXRP’s Lycoming IO-540-C4B5 engines. If the sparks are too early, then the fuel/air mixture may start fully burning and expanding before the piston reaches top-dead-centre. This could be as destructive are preignition. If the sparks are too late, then the burning mixture will produce power too late, and vent out the exhaust side. I don’t know how wide the margin of error is, but clearly to maximize power and avoid damage, the timing must be correct.

Setting the timing involves unbolting the device, rotating the prop by hand to the 25-degree position (visibly marked on some large gears up front), then bolting the magneto back on in exactly that gear alignment. This is easy to do incorrectly. Plus, the magneto has internal gears too which have to be in correct relative alignment. During overhaul, the magneto is disassembled and these gears are set correctly. If any of these tightly set steel gears have been found to slip since their last timing alignment, then something bad must have happened. So, once the timing is set properly, there is little need to change anything.

Turning off a magneto is also bit tricky. The devices are designed to keep on providing high-voltage sparks, even if the on/off control circuit (one wire actually) is severed. This is yet another safety feature, trying to keep the engines going even if nasty things happen elsewhere in the aircraft in flight. So, the magneto is always “hot” unless its control line (“P lead”) is grounded. There are two types of cockpit controls for these control wires, and this can mess with people who are used to working with one style or the other.

The style in GXRP consists of a single on-off switch per magneto, so there are four switches total. When a switch is turned off, the matching magneto’s P lead is grounded, disabling that magneto. That means that the other magneto, whose switch is still “on”, would be running the engine. In contrast, the control style in most single-engine airplanes is a rotary dial marked “off”, “left”, “right”, “both”, and a spring-loaded “start” position. The “left” position grounds the right magneto, and vice versa.

So, if during GXRP’s runup magneto check, the right magneto’s switch is flipped to “off”, which magneto is running the engine? The left one. If the consequent engine RPM drop is too high or too low, which magneto’s half of the ignition system is likely to have the problem? The same, left one.

That’s it for the preliminaries. You may guess what pair of problems led me to writing this article. A month or two ago, at GXRP’s last regular maintenance in Brantford, the timing of one engine’s left magneto was changed. It was working well before, and not so well afterward. (The RPM drop was just around the limits.) While visiting Ottawa yesterday, I asked the local mechanics to fix it “back”. Unfortunately they got confused by the different ignition switch style, and spent several hours proving that the right magneto was perfectly timed. In the end, across much forehead-slapping and joking, the problem was finally understood and very quickly corrected.

All GXRP’s magnetos are again happy. I will put a “do not disturb” sign on them to keep them that way.

UPDATE 2005-07-31: Not again … either some midnight saboteur has visited, or else – more likely – there is something seriously wrong with that magneto. It has started acting up again in new and unpleasant ways. There may be no one to pin in it on this time.