Noise when Amplifier connected via USB DAC and HDMI - ground loop?

[peak]

Hi team,

I am hoping someone might be able to help solving some noise issues with my setup (diagram below).

I was recently the fortunate beneficiary of a nifty Denon Stereo Amp, and decided I needed to make sure it was getting the best quality input for music listening. After learning from this forum about the lacklustre performance of my AVR, I decided to try a SMSL SU-1 DAC connected to a Raspberry Pi 5. Using this only for stereo listening provided great results. On the Amp I can choose between input from the DAC, or switch to 'pure direct' input from the receiver when watching movies/tv/etc. The thinking being that music performance is maximised, whilst also being able to use the amp and speakers as part of my home theatre. Any general setup suggestions welcome.

Now I am keen to also use the RPi for movie streaming, so I hooked it up the HDMI output of the RPi to the AVR.
Problem: lots of noise from the front speakers whenever the amp is on, regardless of input selection, volume, or if the receiver is even on. Unplugging either the RPi HDMI connection, or the DAC (either USB or RCA connection) resolves the problem.

Is this a ground loop issue? Does the fact that the RPi PSU isn't earthed contribute? Any ideas for solutions?

I have looked at some USB isolators, though concerned that they may limit the current required to power the SU-1.

Many thanks in advance for your ideas.

 

[BDWoody]

Is this a ground loop issue? Does the fact that the RPi PSU isn't earthed contribute? Any ideas for solutions?

Sounds like it at first glance.

I'm not going to be the best at helping you sort it out, but wanted to give it a bump back to the top of the queue so you might get some better responses.

 

[Lambda]

Is this a ground loop issue?

I see at least one loop looking at the picture.

its a nice picture but if your looking for ground loops it would help if you also draw ain all the ground connections

I have looked at some USB isolators, though concerned that they may limit the current required to power the SU-1.

if this is a concern get an USB isolator with external power input or add an active power USB HUB.

Also look into Optical HDMI cables

Or just uses the AVR as DAC

Or uses Optical SPDIF (there are many Pi to SPIDIF breakout boards )

 

[Crook]

I've had a ground loop issues on past equipment. Running a wire from metal case of phono preamp to metal case of stereo amp (there was a passive pre in between at that time) fixed things for me. It was just a case of trial and error and a copper wire. I have no idea whether what I did would be considered normal practice?

 

[peak]

its a nice picture but if your looking for ground loops it would help if you also draw ain all the ground connections

All of the powered equipment is earthed, except the RPi + DAC.

Based on my limited understanding of ground loops, don’t they occur when equipment is earthed, not for unearthed gear? The problem arises when a secondary unearthed path is introduced.

if this is a concern get an USB isolator with external power input or add an active power USB HUB.

Also look into Optical HDMI cables

Or just uses the AVR as DAC

Or uses Optical SPDIF (there are many Pi to SPIDIF breakout boards )

I shall try a basic usb isolator, then add power if need be then.

AFAIK optical hdmi still contain copper? That is, they don’t electrically isolate the components?

The whole point of this endeavour was to avoid using the crappy DAC of the AVR…

Are you suggesting spdif from RPi to the DAC? Assuming a separate PSU for the DAC won’t introduce other similar issues, that might be the go!

 

[Vincent Kars]

don’t they occur when equipment is earthed, not for unearthed gear

To the best of my knowledge it occurs when gear is connected to the ground in more than 1 way and there is a difference in potential between the two.
E.g. your TV is connected to the earth but it is also connected to a antenna also connected to the earth.

 

[Lambda]

I shall try a basic usb isolator, then add power if need be then.

serch for Adum4166 Adum3166 USB 2.0 Isolator 480Mbps
Or uses the topping with external power option.

AFAIK optical hdmi still contain copper? That is, they don’t electrically isolate the components?

But less and for low speed signal. you can uses a longer cable to get higher impedance.

The whole point of this endeavour was to avoid using the crappy DAC of the AVR…

Do you relay hear a difference?

All of the powered equipment is earthed,

Then you have 4 ground loops.

don’t they occur when equipment is earthed, not for unearthed gear? The problem arises when a secondary unearthed path is introduced.

No they arises if you have a loop. the loop can be trough Mains safety ground. but it don’t have to be.

You have 3 loops going to mains earth ground and analog gnd

and one loop at the pi with Analog as well es Digital GND forming a loop.
This is the one causing problems for you.

in general you want to keep
the loop Area as small as possible.
The loop Impedance as hight as possible.
for example with an UBS isolator and separated isolated power supply's or optical Toslink and separated isolated power supply's

 

[vddobrev]

@peak From your diagram, what happens if you unplug the TV HDMI, but all else plugged? I was in a similar situation, where the culprit was the coax cable coming in from my cable TV.
Edit: To be complete, in my situation it was propagated via the HDMI from the TV. Since the cable TV company did not cooperate, I solved the issue by grounding the incoming coax cable myself.

 

[Lambda]

I solved the issue by grounding the incoming coax cable myself.

Well, that’s what you’re supposed to do in most parts of the world anyway. It’s called equipotential bonding, and it should be done with all incoming lines like water, gas, and cable.
To get rid of high-frequency (HF) noise, you can use ferrites. For low-frequency noise, you can use a coax isolator—a very simple and inexpensive device, usually consisting of just two ceramic capacitors.