Noise fault diagnosis of car audio mid-range/video system

Figure 1: Simulator of unbalanced interface.

Emulators are wired using standard connectors as shown in Figure 1. They do not carry signals, so they must be clearly labeled so they are not inadvertently permanently installed into the system. Be extremely careful not to damage the speakers or headphones.

four step testing method

Each signal interface is tested using the following four-step method. Multiple cables in the same chassis (e.g. stereo pairs) are to be tested simultaneously:

Is the output quiet?

No—The problem is either in chassis B or in the downstream stage;

Yes - go to the next step;

Is the output quiet?

No - there is a "single pin connection problem" inside chassis B, which can be confirmed by the "Hummer" buzzer test.

Yes - go to the next step;

Is the output quiet?

No—Noise is being induced into the cable, relocate the cable to avoid the interfering field;

Yes - go to the next step;

Is the output quiet?

No - the problem is common impedance coupling and an isolator needs to be installed in the signal path;

Yes—The noise is coming from the output of Chassis A. Perform the test sequence on the next upstream interface. Repeat as necessary until the problem is found.

Found issues with balanced interfaces

For balanced interfaces, testing specifically identifies the following issues:

Figure 2: Simulator of balanced interface.

The emulators are wired using the standard connectors shown in the figure. They do not carry signals, so they must be clearly labeled so they are not inadvertently permanently installed into the system. Be extremely careful not to damage the speakers or headphones.

Each signal interface is tested using the following four-step method. Multiple cables in the same chassis (e.g. stereo pairs) are to be tested simultaneously:

Is the output quiet?

No—The problem is either in chassis B or in the downstream stage;

Yes - go to the next step;

Is the output quiet?

No - there is a "single pin connection problem" inside Chassis B, which can be confirmed by Hummer testing.

Yes - go to the next step;

Is the output quiet?

No—Noise is being induced into the cable, relocate the cable to avoid the interfering field;

Yes - go to the next step;

Is the output quiet?

No - The problem is shield current induced noise (SCIN), use a different type of cable or take steps to reduce shield current.

Yes—The noise originates from the output of Chassis A, and the test sequence is performed on the next upstream interface.

Figure 3: Three chassis with internal pinout issues.

Figure 4: A "Hummer" used to test shields.

"Pin single connection problem" and "Hummer"

The "single-pin problem," coined by Neil Muncy, refers to the phenomenon of common impedance coupling that has been inadvertently engineered into a significant number of products. As Neil says, "Balanced design gets a bad rap because of this, and it's a reputation it doesn't deserve. It's really the novelty issue. Balanced wire interconnects were originally expected to ensure noiseless system performance, but, in fact, they Often this is not done." This problem effectively converts the shield connection into a very low impedance signal input, as shown in Figure 3, where shield current is allowed to flow through internal wires or circuit board traces shared by sensitive amplifier circuits. Any small voltage drop that occurs is amplified and appears at the output of the device. When this problem exists in a system, it may interact with other noise coupling mechanisms, making troubleshooting difficult. It may affect unbalanced audio, video and data interfaces. Fortunately: simple testing can expose problems. The "Hummer" test method was proposed by John Windt. The circuit of this simple device is shown in the figure below. It forces a current of about 50mA to flow through a potentially troublesome shield connection in the device under test. In a properly designed device, this can cause additional noise at the device output. When the clip is shorted, the 12V transformer supplies approximately 50mA. Optional LEDs and an IN4001 diode simply show that a connection has been made and current is indeed flowing.

Using the "Hummer" test

1. Disconnect the input and output cables (except the output under test) and all chassis connections (such as rack mounting) from the device under test;

2. Add power to the device under test;

3. Measure (and if possible, listen to) the output of the device under test. The only noise should be white noise or "hiss." Experiment with various control settings to familiarize yourself with the noise characteristics of the device under test without a Hummer connected.

Figure 5: Pattern generator.

Figure 6: Video simulator.

4. Connect one pin of the Hummer to the chassis of the device under test and touch the other pin to the shield contact of each output or output connector. If the device under test is designed correctly, there will be no hum or changes in the noise floor in the output.

5. Test other possible trouble paths, such as from the input shield contact to the output shield contact or from the protective ground pin of the power cord to the chassis. In some devices, the single pin of the XLR connector doesn't necessarily have to be connected directly to ground - hopefully this is just on the input side. In this case, the "Hummer" LED may not glow, and that's right.

Found a problem with the video interface

This is a simple difference from the troubleshooting steps for audio systems. Because many - if not most - monitors display a blue screen when there is no video signal, this test utilizes a portable video source to keep the display functioning. If you do a lot of video processing, you may want to use a video pattern generator similar to the B&K Precision Model 1211E shown in Figure 6. It is important that the generator is battery powered and not grounded. You will also need a test adapter or "emulator" connection as shown in Figure 5. Utilizes RCA and BNC connectors to easily fit it into a small cast aluminum box. By temporarily placing testers and generators at strategic locations in the system, precise information about the nature of the problem can be exposed. The cable between the tester and chassis B should be as short as possible—much shorter than the cable under test.

To test starting from the monitor input and working backwards to the signal source, each signal interface must be tested using the following four steps:

Are the distractions gone?

No—The problem is either with Chassis B or downstream. Connect the cable again and perform the test on the next downstream interface.

Yes—Go to the next step.

Are the distractions gone?

No—The problem may be due to common impedance coupling within Chassis B or to devices further downstream. If the inputs are different (i.e. the shield is not directly connected to ground), then the problem may be due to exceeding the common-mode voltage limit. If the problem is not in Chassis B, connect the cable again and start the test procedure on the next downstream interface.

Yes—Go to the next step.

Are the distractions gone?

No—The interference is being induced by the cable itself due to the strong AC magnetic field near the cable. Reroute cables to avoid strong magnetic fields. Such magnetic fields include high-current power lines, power transformers, and CRT monitors. Electric field coupling is also possible, but is extremely rare in video systems unless the shield itself is destroyed or disconnected.

Yes—Go to the next step.

Are the distractions gone?

No—Ground current couples through the common impedance of the cable shield. Install a suitable ground isolation device.

Yes—Interference is present at the output of Chassis A. Execute the test sequence on the interface between Chassis A and the upstream signal source.

By taking a systematic approach to diagnosing noise problems in your sound system, you can often locate and correct the problematic circuit or connection. With a systematic testing method and simple tools, you can directly and quickly find even the most elusive audio noise problems.