• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Grounding audio - simple example

IAtaman

Major Contributor
Forum Donor
Joined
Mar 29, 2021
Messages
2,272
Likes
3,836
I don't think I can add much on what solderdude has already said, maybe only one point important to remember again - Earth Ground and circuit ground / signal ground are not necessarily the same thing.

Rule 1 basically should have said: Never connect a 3-prong device to a 2-prong outlet. (again for safety reasons, the safety ground connection is there for a reason)

It just is written down a bit confusing as the [blue] '2 pronged AC powered device' probably should have stated '2 pronged outlet'.
Never connect a 2 pronged AC powered device to a 3 pronged AC powered device might make sense as a rule as a potential way to avoid creating voltage differential between different signal ground potentials maybe?
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
2 pronged devices are double isolated so can safely be used on 3-prong and 2 prong mains sockets.
The fact that small leakage currents can cause hum problems is not a safety concern (when leakage currents do not exceed dangerous levels)
A 3 prong plug devices may not be safe to used in a 2 prong socket as it may have a voltage on components that can be touched or could have such in case of a catastrophic failure.

Of course, whether or not this is physically possible depends on the type of plugs/sockets being used.
 

Zapper

Addicted to Fun and Learning
Joined
Mar 4, 2023
Messages
608
Likes
822
I don't think any of the previous explanations are sufficient.

Ground loops usually refer to inductive ground loops. The earth ground leads and/or signal wire shields form a closed loop. Current is induced in these loops by ambient AC magnetic fields from mains current flowing in the vicinity according to Faraday's law of induction. The magnitude of the voltage around the loop is proportional to the area of the loop and the component of the magnetic field perpendicular to the loop, so bundling grounded power leads and bringing them to a common outlet strip is reduces hum pickup because it minimizes loop area.

However, what the OP is presenting is not an inductive ground loop, because there is no closed current path. The ungrounded SMPS supplies a laptop which drives a USB-powered DAC which drives the headphones. There is no loop involving either ground leads or signal shields.

This is an example of capacitive AC pickup. The output of the SMPS is dielectrically isolated from the mains through a transformer, but there is some capacitive coupling between the mains and the output through parasitic capacitance in the transformer, and frequently by a small bridge capacitor deliberately added between the mains side and the output return to suppress electromagnetic emissions. This couples some mains AC into the laptop, through the USB return, into the headphones, where it couples to the listener through conduction or capacitance. Every object has some capacitance to the environment (i.e. earth ground), including the DAC, the headphones, and the listener, so a small AC current flows along this path. The current will be very small because the mains frequency is low and the capacitance to earth ground is small. The entire current loop consists of earth ground through the utility supply transformer, to the domestic mains, to the SMPS, to the USB return via capacitance, to the headphones and listener, back to the earth ground through capacitance.

There is also the possibility that the AC signal conducted from the SMPS consists of the mains frequency modulating the SMPS switching frequency, which is on the order of 100's of kHz to 1's of MHz. This high frequency would be conducted much more readily through capacitive coupling, and could be rectified into audible hum by the circuitry of the DAC. That would be an example of conducted electromagnetic interference (EMI).

When the DAC is connected to earth ground, this current path is cut short with a direct path to ground, so it won't flow through the headphones and listener. The fact that this stops the hum suggests that the DAC/headphone amp is sensitive to common mode AC current in its headphone output. This seems to me to be a design flaw in the DAC's headphone amplifier.

The way for the OP to test if the above hypothesis is correct is to unplug the laptop from the SMPS and run on battery power. The hum should stop because the capacitive connection to the mains is eliminated.

A USB galvanic isolator should greatly reduce or eliminate the problem by breaking the capacitive path. Such isolators have some input-output capacitance but the amount is very small (2pF for the ADI version). A USB to TOSLINK transmitter/receiver pair would serve the same purpose with even less capacitance.
 
Last edited:
OP
M

Music1969

Major Contributor
Joined
Feb 19, 2018
Messages
4,636
Likes
2,809
The entire current loop consists of earth ground through the utility supply transformer, to the domestic mains, to the SMPS, to the USB return via capacitance, to the headphones and listener, back to the earth ground through capacitance.
Thanks for the detailed info.
This is the leakage current pathway (loop) I was interested in knowing about.
Solderdude did explain this too but it's nice to see the explanation in a different form, to reinforce my understanding.

The way for the OP to test if the above hypothesis is correct is to unplug the laptop from the SMPS and run on battery power. The hum should stop because the capacitive connection to the mains is eliminated.
Yes I had already tested this and indeed the him disappears - so I knew exactly where the leakage currents was coming from. But from practical sense I can't run on only battery all the time.

As I mentioned in my opening post, earthing the DAC/headphone amplifier combo, also solves this issue.

So my original question was what happens to these currents when i direct them to earth ground - where exactly do they go.

I didn't realise that when I have the hum in my headphones, these leakage currents are literally going through me, to the earth and all the way back to utility supply transformer ?! Crazy
 

Zapper

Addicted to Fun and Learning
Joined
Mar 4, 2023
Messages
608
Likes
822
As I mentioned in my opening post, earthing the DAC/headphone amplifier combo, also solves this issue.
Right, that's the best and cheapest fix. I mentioned the USB isolator as a more mobile solution should you need it.

Out of curiosity what brand & model is the SMPS?
 
OP
M

Music1969

Major Contributor
Joined
Feb 19, 2018
Messages
4,636
Likes
2,809
Right, that's the best and cheapest fix. I mentioned the USB isolator as a more mobile solution should you need it.

Out of curiosity what brand & model is the SMPS?
There's two SMPS in this simple headphone system

For laptop , it is Apples adapter for Macbook.

For DAC/amp, it is what Chord supplied with Hugo2

Hugo2's USB input is not galvanically isolated so USB isolator may solve it - but there is still leakage from the Hugo2's SMPS that could still come through me and headphones, even if the currents from Macbook are blocked.
 

Groundrod

New Member
Joined
Feb 4, 2023
Messages
4
Likes
0
Most SMPS power bricks have a 3rd pin/prong that does nothing - not continuous with its DC output plug . Exception is medical grade.

So all powerbricks need to be replaced with medical grade?


Why?

I have a 2 pronged laptop (VERY common) connected to a 3 pronged headphone amplifier (with DAC integrated).

When I previously had a 2 pronged headphone amplifier, I would hear buzzing in headphones until I touched metal case.

So without 3 pronged headamp, I would have leakage currents passing through my headphones/body.


Isn't this a recipe for ground loops, especially with RCA interconnects ?

And none of your points said best to use balanced XLR interconnects?


cc @Speedskater @solderdude @IAtaman



The very nature of a SMPS circuit is noise, so all SMPS's use line filters and one of these filters needs to be Earth Grounded.

If this filter is left floating the device it is powering will have half the mains AC voltage (~60Vac) plus the unfiltered switching noise on its DC/Chassis Ground.

Since all the connectors such as USB, HDMI, and Audio... cable shields are connected to the circuit DC chassis Ground, these AC voltages will contaminate the signals passing through them.

A lot of people are using the term ground loop to describe this effect, but the real problem is caused by an ungrounded SMPS in the mix.

Also, regarding leakage currents, this current is not leakage, but in fact the current the SMPS filter draws to do its job and is only a problem if it has no Earth Ground connection.

This is why many people are complaining about getting a shock when connecting too or touching the chassis!

Since all the cables are connected to the chassis, the cables become the only path for ungrounded devices to Earth Ground, and the filter current will pass through them instead of the proper path ( the AC cord), which in turn results in hum and buzz.

As I stated in my original post, it all starts at the AC outlet, so first make you are using a single AC outlet that is wired properly, and then make sure all the devices in your system use a 3 pronged AC power supply and you won't have any problems.

Cheers!
 
Last edited:
Top Bottom