February 5, 1996 Licking Alternator Whine

Email this article |Print this article

by	John Schwaner
	This article is Copyright © 1995 by Sacramento Sky Ranch Inc. All rights reserved.

About the Author ... John Schwaner is AVweb's powerplant expert. John is a world-class authority on piston aircraft engines, and a specialist in the engineering analysis of engine failures. John runs Sacramento Sky Ranch, Inc., a leading distributor of aircraft and engine parts, and probably the foremost aircraft hose shop and magneto overhaul facility in the U.S. John and his wife live in Sacramento, California. John has also written two superb technical books: Sky Ranch Engineering Manual and The Magneto Ignition System. Both can be previewed in and ordered from the AVweb Online Bookstore.

Identifying the problem

Alternator induced radio noise is a high pitched whine whose pitch and intensity increases and decreases with changes in engine speed. Turning the alternator master switch off also turns off the radio noise.

Solid state regulators that use a pulse-width-modulated field control system can also create a whine in the radios. Regulator-caused whine can be distinguished from the alternator-caused whine in that the intensity and pitch of regulator-induced noise changes with changing current load at a constant engine speed. Thus, turning on the landing lights won't increase alternator whine but will increase regulator whine.

How the alternator works

Current generated in the alternator stator windings is three-phase alternating current, but diodes convert it from AC to DC before it leaves the alternator. Six diodes are required to rectify the three stator phases. Each of the three stator windings is connected to a pair of diodes. Three diodes are connected to the positive output terminal of the alternator, and the other three are connected to the negative (ground) terminal.

As the voltage of each stator winding increases, the corresponding pair of diodes becomes forward biased and allows alternator current to pass. Which stator winding and diode pair is conducting at any moment depends upon rotor position. After the diodes rectify the three AC phases and sum them all together, the combined result is a DC voltage with only a slight amount of AC ripple voltage remaining.

The best way to detect ripple voltage on the electrical bus is with an oscilloscope. Another method is to use an ordinary volt-ohmmeter (VOM) set to measure AC volts. You may have to connect a capacitor in series with the positive meter lead to block out the DC voltage so that only the ripple voltage gets to your meter. (Some meters do this automatically when you select AC volts.) The capacitor is an open circuit to DC but passes AC, so the voltmeter reading you see is the amount of AC ripple voltage on the bus. You will need to do comparison readings with other aircraft to determine what AC voltage level is normal.

What causes alternator whine?

Normally, there is not enough ripple voltage to cause radio noise. But, there are two conditions that can cause an increase in ripple voltage sufficient to create radio noise. These are diode failure and increased circuit impedance.

If an alternator diode fails, the amount of ripple voltage increases markedly. Alternator whine can be a symptom of a bad alternator diode. Two test methods can be used to test the alternator without disassembly.

There is a hand held unit with a probe that clamps over the alternator output wire. A bad diode will show up on the meter. These meters were originally sold as the Ward Aero Alternator Tester model 647. They are currently sold by Support Systems Inc. as model 10-647-01.

The second test method is to use an oscilloscope to check the alternator output for excessive voltage ripple or rectifier spikes caused by a bad diode.

Checking the diodes

With the alternator apart, the diodes can be checked with a VOM set to measure ohms. This test makes sure that each diode conducts in only one direction. You need to unsolder the stator leads from the each diode. Calibrate the VOM on the R x 1 multiplier range scale so that there is zero reading with the VOM leads shorted together.

Connect one test probe to the alternator's positive output terminal and touch the other test probe to each of the three solder terminals of the diodes mounted to the positive rectifier plate. Note the three ohmmeter readings: they should be identical. Now reverse the test probes and repeat the test. Note the three ohmmeter readings: again they should be identical to each other, but not the same as in the previous step. Three of the ohmmeter readings should show a low resistance reading of approximately 6 to 20 ohms and three should show an infinite reading (no meter movement).

Repeat the same test procedure for the three diodes on the negative rectifier plate, connecting one test probe to the negative output terminal and checking all three diodes with the other probe. Then reverse the leads and check again. The diodes should show low resistance in one direction, and infinite resistance in the opposite direction.

Circuit causes

Alternator whine can also be caused by poor electrical connections, especially at the battery. Normally, the low impedance of the battery keeps the aircraft's electrical circuits at a DC potential. (Impedence is simply resistance to an AC current.) Any AC ripple voltage in the aircraft bus is absorbed by the battery. Thus, the aircraft battery acts as a big ripple absorber.

If the battery provided zero impedence (i.e., a short-circuit for AC current), alternator noise could not occur. In the real world, there will always be some impedance. But the lower it is, the less ripple voltage there will be.

Let's assume that the battery positive terminal is corroded. Although DC resistance as measured with an ohmmeter may still be low, the high-frequency resistance (i.e., impedence) may be very high. The higher this impedence, the greater the ripple voltage on the bus and the more whine you hear in your radios.

Circuit impedance can be lowered by making sure the battery posts are clean and making good contact. DC resistance should be less than 0.01 ohm...virtually zero. Also check the alternator ground connections, including the engine grounding strap. DC resistance between the alternator and the negative post of the battery terminal should be as low as possible.

The ideal low-noise circuit would have the alternator power output going directly to the battery's positive terminal. This dumps ripple voltage into the battery, where it is absorbed. The radio power lead would also go directly to a pure DC source, the battery.

If the alternator power lead and the radio power lead connects to a bus, then voltage ripple can go from the alternator to the radio power lead. The amount of voltage ripple at the bus depends upon the impedance between the bus and the battery. This impedance is higher than at the battery.

The return path is from the alternator to the engine, engine mount, firewall, and through the fuselage to the battery. These connections should have low resistance. Flat braided ground straps are ideal for grounding the airframe to the engine mount. Flat braided straps are used because impedance is less with a braided, flat conductor than a round wire conductor.

Filter capacitors

There are two methods of filtering ripple voltage: bypassing the ripple voltage back to the source, or blocking the voltage ripple so that it cannot pass. Capacitors are used to bypass ripple voltage, whereas inductors are used to block noise currents. The most effective approach depends primarily on the circuit impedance.

Capacitors bypass noise currents back to the alternator return path (commonly referred to as ground). To be effective, a capacitor must have a low impedance path back to the alternator. Consequently, a filter capacitor must be mounted as close to possible to the alternator. The capacitor is installed with one lead connected to the power output and the other lead to ground, so that it is in parallel with the circuit.

For DC voltages the capacitor forms an open circuit (high impedance) and doesn't allow any current to pass. At noise frequencies the capacitor forms a short circuit (low impedance) and bypasses noise currents back to the alternator. In this manner we have formed a low-pass filter. The effectiveness of using a capacitor as a noise filter depends upon matching the capacitance rating of the capacitor to the frequency of the noise currents.

The frequency at which the capacitor's capacitance and inductance are equal is where it has the lowest impedance and the best filtering. This is the resonant frequency. The correct size capacitor is one where the frequency we wish to bypass is the same or less than the resonant frequency.

Smaller size capacitors (picofarad range) are effective at high frequencies, while larger size capacitors (microfarad range) are effective at lower frequencies. If you're filtering conducted interference (as you are in an alternator), then this is low-frequency and the capacitor should be in the microfarad range. If you're filtering radiated interference (where the conductor is acting as an antenna), this is high-frequency and the capacitor should be in the picofarad range.

Typically, an alternator filter uses a .5 to 50 microfarad capacitor. Cessna has a 5.72 microfarad capacitor filter available as part number S1915-1.

The best types of capacitors for filtering are ceramic and tantalum capacitors, ceramic for the picofarad range and tantalum for the microfarad range. Electrolytic capacitors are relatively poor noise filters, and also have a short life.

Capacitor resonance can be approximated with the following formula: resonant frequency (in MHz) equals 1/2 pi times the square root of lead length times capacitance. Notice that lead length has a significant effect on the capacitor's resonant frequency. For example, a 500 pf capacitor with 1/4 inch leads resonates at 100 MHz. But with 1 inch leads, it resonates at 50 MHz. So capacitor lead lengths used in filter circuits should be kept as short as possible.

Inductive filters

The other way to filter radio noise is to block the ripple with a series inductor. The most common style of inductor for noise filtering is a ferrite core. These come in many different styles but typically the wire with the noise currents is wrapped around the core, creating an inductor in series with the circuit. DC current passes through the core but high frequency currents induce a magnetic field in the ferromagnetic material of the core. This magnetic field raises the impedance and effectively blocks noise currents. Ferrites are effective on radio power input leads and strobe power input leads. In the first case they prevent noise currents from entering the radio, and in the second case they prevent noise currents from exiting the strobe.

To be effective, ferrite impedance must be larger than circuit impedance. To filter the output of an alternator would required an impractically huge ferrite core. So alternator voltage ripple is usually bypassed to ground by use of a capacitor. However, ferrites are simple to use and have an amazing filtering ability.

Ferrites are best used in low impedance circuits whereas capacitors are best used in high impedance circuits. It is best to install ferrites on the radio power input leads, and to use a filter capacitor at the alternator output terminal.