2
\$\begingroup\$

Here's the schematic for my car:

schematic

simulate this circuit – Schematic created using CircuitLab

The head unit is referenced to car's chassis. My amplifier and DSP are in the trunk. The DSP has an isolated DC/DC converter. And the amplifier internally connects signal ground to the chassis. This form a massive ground loop from the head unit through the RCA cables through the ground connection in the amplifier through the car's body back to the head unit. This has been a source of whine and noise.

I know about the ground loop isolator transformers, but those usually introduce a huge amount of distortion and dirty up the sound.

Other than a transformer in the way of the signal, what are some ways I can break this ground loop?

I am designing the DSP, so I have full control over how it manages the ground. It consists of an ADC daughter board, a DSP board, and 2 DAC boards. All the ground connections in the DSP are connected together, but I have the ability to change that. I thought about opto-coupling the i2s signal coming out of ADC, but I'm not sure if that's a good way.

\$\endgroup\$

1 Answer 1

1
\$\begingroup\$

Grounding

All grounds in a car or system should have a common return point, such as a bolt that is chassis mounted. The battery ground is often right to the metal chassis close by. It is good but not ideal so some noise can exist. In electrical panels all grounds return to the panel that sourced the power, so no ground loops possible. A ground loop is created when a device ground connects to another devices ground before it reach's a common grounding point.

Assuming all grounding is solid and very low resistance:

Alternator

The key is not ground so much as it is the power feed to your head unit and amplifier. The whine comes from your alternator and the only way I have been able to kill the whine was to insert inductors in the power feeds to the head unit (most effective and lower cost) and power amplifiers.

Filters

The head unit seldom pulls more than 5 amps so a 10 amp 1 mH inductor (often in the shape of a toroid) often took care of the problem, because noise in the head unit is fed back to the amplifiers that faithfully amplify the noise as well as your music.

If some residual noise remains install 1 mH inductors to the power feeds (from the battery) to the amplifiers, rated for twice the current the amplifier is rated.

Battery

The battery is assumed to be a good noise filter, but if you test it for AC noise or use a oscilloscope on the terminals while the engine is running you may find several volts of noise that match the alternator whine. Filter capacitors are already in your head unit and amplifiers but like the battery they cannot filter out a strong alternator whine. The inductors complete the filter network by chocking off the whine about 99% in most cases.

Warnings

You could put an inductor at the alternator output but car manufactures warn that if it is too small (current rating) it will not handle the current passing through it and saturate, as in not filter out anything. Any inductor/power choke should be rated for twice the current expected to pass through it.

They also warn of inductors of low value that may actually resonate with the alternator whine and cause back-emf into the alternator, possibly damaging the alternator. Safer to install each inductor at its point of use or off the battery power tap (after the fuse) for the device.

\$\endgroup\$
7
  • \$\begingroup\$ Thanks. I didn't think a series choke in the way of power to the head unit might solve the alternator noise. I'll try that sometime soon. \$\endgroup\$
    – darksky
    Commented Sep 30, 2018 at 23:22
  • 2
    \$\begingroup\$ You want to create LOCAL BATTERIES for the "whine frequencies". That means a 2-element Low Pass Filter. A series inductor and a shunt capacitor are the method. You may need to provide some dampening; A series value would be Sqrt(L / C). \$\endgroup\$ Commented Oct 1, 2018 at 5:02
  • \$\begingroup\$ @analogsystemsrf. With all due respect most of the time an inductor in series with the head unit power got rid of the whine completely. Local Batteries and Supercaps were for systems with 1,000 watts of power-or more. To a certain point very heavy gauge wire was good enough. For the amps it was not so much about whine as it was power starvation from skinny power feeds. \$\endgroup\$
    – user105652
    Commented Oct 1, 2018 at 5:22
  • 1
    \$\begingroup\$ @darksky Have you solved this issue? What ended up being the fix? \$\endgroup\$
    – Taudris
    Commented Nov 4, 2018 at 14:40
  • 1
    \$\begingroup\$ @darksky. Glad to see you making an effort to track down the noise. Car power amps are often tied with heavy gauge wire to the positive battery post, which is NOT quiet. Custom high-current inductors can be expensive, so 'rolling' your own could save you lots of money. Use a large toroid if possible. Only turns that go thru the center of the toroid count as a turn. \$\endgroup\$
    – user105652
    Commented Nov 14, 2018 at 4:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.