Ground loop problem

Thread Starter

Maverick923

Joined Feb 4, 2019
13
I have a little amplifier and preamp board which share the same 24v dc power supplied from a battery. Power is being converted to 12vdc through a buck converter. Everything works, but of course, I get a ground loop when I set the preamp to bluetooth especially when it connects to my phone it can be heard. I did use both with separate power sources, and the noise was gone. Separate power sources is not really an option. I have seen videos on different methods of fixing this issue, but not one that pertains to the setup I have. How can I go about eliminating the noise caused by the groundloop? Any assistance is appreciated.
 

SteveSh

Joined Nov 5, 2019
109
How do you know it's a ground loop that's causing the problem?
Is it possible you're just "seeing" noise from the SMPS (switch mode power supply)?
A simple block diagram of your setup would be helpful - one that shows how the power and signal returns are connected.
 

MrSoftware

Joined Oct 29, 2013
2,188
If it's a ground issue, one thing you can do is make sure everything is grounded at the same exact point. So the the grounds for the amplifier and pre-amp and anything else related to the audio should all connect to the same single ground location.
 

BobaMosfet

Joined Jul 1, 2009
2,110
I have a little amplifier and preamp board which share the same 24v dc power supplied from a battery. Power is being converted to 12vdc through a buck converter. Everything works, but of course, I get a ground loop when I set the preamp to bluetooth especially when it connects to my phone it can be heard. I did use both with separate power sources, and the noise was gone. Separate power sources is not really an option. I have seen videos on different methods of fixing this issue, but not one that pertains to the setup I have. How can I go about eliminating the noise caused by the groundloop? Any assistance is appreciated.
First of all, it's likely you don't actually understand what a ground-loop is, or you could very specifically test it with a DMM. Don't feel bad, most people in electronics are confused about it, because smart people seem unable to explain simple things, well, simply. Nonetheless, I think this will help--

A ground loop is a condition where a circuit has 2 (or more) different levels of ground. This is exactly, and the ONLY thing that a 'ground loop' is. It isn't actually a loop. once again, smart people say stupid things for some unknown reason and confuse everyone. What it really means is that somewhere in your circuit, assuming it IS a ground loop issue, you have placed a resistance between the ground and one portion of your circuit. Which creates a voltage divider, so that portion gets less voltage, and the ground level is raised, while the other circuit portion is OK (if you only have 1 ground loop).

How to Test:

Put a ground lead of a DMM on your last ground point (power-supply - negative terminal), and then place the positive probe near it on a ground, and check all your closest grounding points. And then extend to the next farthest away and repeat. You should register zero volts difference (aka dropped/lost) between your probe points. If you read anything else, carefully examine the circuit between those probe points, because you are now across that 'resistance'.

If you can't find an elevated ground, you have no ground loop.
 

DickCappels

Joined Aug 21, 2008
10,152
I like your block diagram, especially that you included yourself in the system,

It seems that you have a transmitter (the Bluetooth talking back to the music source) getting into your amplifier which is detecting the RF. An alternate explanation is that the Bluetooth module is drawing large current spikes from the power supply and that is how it is getting into the amplifier. He also added a LC low pass filter between the power supply and the Bluetooth module.

The LC filter would help reduce the amplitude of current spikes from the Bluetooth model getting back into the isolated output power supply and maybe that actually does something. You should try that.

If it is RF getting into the audio amplifier through the power or audio connections you can try to eliminate it with a conducted EMI filter. Commercial filters are generally expected to be used below 100 MHz, so if you try a commercial filter, have a look at the datasheet.
1573915860680.png
You can put this circuit or one like it in series with the audio amplifier, just leave out the Y capacitors (CY1 and CY2) the L and N connections are interchangeable until you hook L to ground or +5. You probably already understand that last part. The inductor with two windings is referred to as a common mode choke.

This kind of filter can also be used for the audio signals.
1573916272384.png
In the above example, the AC wall adapter and audio inputs to a transmitter had to be RF isolated as they entered the transmitter by T1 and T2. T2 has four windings all in the same phase. If you can do this for ground, +5 (3.3?), Left, and Right, you may only need one common mode choke between transmitter and receiver.

When I deal with sneaky signals like the one you are dealing with I sit down and think of as many ways the signal might get from the source to the circuit that is being affected with it, then systematically apply filters and move around ground point s start with the most likely until some improvement appears. A good place to start is to work on isolating the source of the interference rather than protecting the circuit that is interfered with, at least that is my preference when practical.

An oscilloscope is very useful in identifying how the signal gets in. Once you have a system that is satisfactory, start taking fixes out until you cannot take another fix out without performance becoming less than satisfactory. In other words, during the process you are likely to, at some point, have a more than one fix and sometimes you don't really need that many.
 

crutschow

Joined Mar 14, 2008
34,285
It isn't actually a loop
But it is a "loop", that's why smart people call it that. ;)
It's where a circuit has more than one ground path back to the source, making a loop.
This can cause an unintended or circulating (If from the AC mains) ground current, which generates a noise voltage in the ground resistance.

In the TS's case, this is likely the signal ground between the Bluetooth and the amp.
Some of the Bluetooth noisy power supply current goes through the amp ground back to the supply, causing the observed noise.
 
Last edited:

SteveSh

Joined Nov 5, 2019
109
Maverick923 -

Back to the simple sketch you provided, could you add some more detail to that? I want to see exactly how the two power sources, the amp, and the bluetooth module are connected, both in terms of the power rails and the signal returns (grounds). I am interested in the topology of of your circuit - what connects to what and in what order.

I don't think your problem is radiated susceptibility (that is, RF coupling into the amp), because you said that you do not hear the noise if the amp and bluetooth module are powered from separate power supplies.
 

SteveSh

Joined Nov 5, 2019
109
Maverick923-

If what you drew is accurate, the switching noise from the 12V power supply could be flowing through the power and return leads for the amp.

Simplest thing to try would be to "single point" the supply and return lines at the 24V battery. You could also try adding a cap across the +24V and return inputs to the switching power supply. See marked up sketch below.

Do you know what frequency the SMPS switches at?

1573991248933.png
 
Top