Identify and fix crackling sound from pc to receiver

[Rubberducky]

I have an AV setup with a Denon X3700H AVR connected to a TV and an Nvidia Shield which works without issues.

The moment I connect my desktop PC to the AVR through HDMI, a crackling sound starts coming through my speakers. This sound comes out regardless of the selected input or volume level on the AVR.

I can reproduce the issue also by just connecting the AVR ground to the PC ground, without a HDMI connection, although the crackling is less intense in that case.

Currently, the desktop is plugged into a non-earthed outlet, but I was also able to reproduce the issue when using a grounded outlet. The AVR is double insulated and has no earth connection.

I've measured the voltage between the PC ground and the AVR ground, and the result can be seen in the image: a high frequency signal with ~1.5Vpp.

Is this signal the source of the noising coming out of the speakers? Is this switch-mode power supply noise somehow "leaking" into the AVR? How do I go about resolving this?

 

[Dunring]

Have you checked the power supply to make sure it has metal to metal contact with your case to provide ground to the outlet? A ground wire attached or even using an uncoated screw on the power supply to case can do it. Then into a grounded outlet of course..

 

[ZolaIII]

Use optical line from PC if you have one and it's good enough for you (PCM up to 192000 Hz 24 bit stereo but only compressed Dolby 5.1 surround). You might try another PSU if you are sure EMI is from it.

 

[JSmith]

PC to the AVR through HDMI

You could use HDMI via Ethernet cable by employing a HDBaseT extender, which should isolate any ground loops etc.

HDBaseT - HDBaseT

Extend fully uncompressed Audio, Video, USB, Gigabit Ethernet, power, and control up to 100 meters/328 feet over a single, standard Category cable.

hdbaset.org

https://www.amazon.com/HDMI-Extender-Over-CAT5-CAT6/dp/B084GF6W53/?th=1

JSmith

 

[Rubberducky]

Thanks for the responses already!

I've done some further investigation:

Have you checked the power supply to make sure it has metal to metal contact with your case to provide ground to the outlet? A ground wire attached or even using an uncoated screw on the power supply to case can do it. Then into a grounded outlet of course..

PSU is properly connected to the metal of the case (PSU earth continuity to metal screws)

Use optical line from PC if you have one and it's good enough for you (PCM up to 192000 Hz 24 bit stereo but only compressed Dolby 5.1 surround). You might try another PSU if you are sure EMI is from it.

I'm using uncompressed surround sound, so optical is not an option.

I've identified a potential ground loop: there are 2 class 1 (grounded casing) devices and a couple class 2 (double insulated, ungrounded casing) devices in the chain.
My subwoofer amplifier and my PC are class 1 and are causing a ground loop. If I disconnect either one of them, the buzzing stops.

However, all of the devices in the chain are plugged into the same outlet, so I can't imagine how an actual potential difference can be caused.

I'm guessing it's EMI noise from the PSU making its way into the signal chain through this ground loop, but I'm unsure how I can verify this.

If I plug everything into a grounded outlet, the noise doesn't go away.

Some questions:
- Can EMI noise from the PSU get into the chain through a ground loop, even if all ground is roughly at the same potential (all same outlet, I measure 0.02VAC with my DMM between the different grounds, which seems like the lowest it can measure)
- Would an RCA isolator be worth it here? With it I can isolate the subwoofer amp ground from the rest of the chain, but I'm afraid it will deteriorate my sound. Also, it seems like hum issues should be fixed a the source rather than with an isolator.

 

[ZolaIII]

So what did you say (as how you classified them) about class 1 part of the chain? It work's with the Shield quite OK as much as I understand.
What are you getting as an out when going through PC GPU HDMI a distortion of broken output/reproduction?
I suggested to try a lossy beatstream (DD, DD+) trough SPDIF for diagnostic purpose.
EMI noise trough both GPU/PSU under significant load can go in I don't believe in such isolators. How bad really is it?
What’s the CPU and is there a HDMI out from MB to try it that way (using discreet GPU)?

 

[Rubberducky]

So what did you say (as how you classified them) about class 1 part of the chain? It work's with the Shield quite OK as much as I understand.
What are you getting as an out when going through PC GPU HDMI a distortion of broken output/reproduction?
I suggested to try a lossy beatstream (DD, DD+) trough SPDIF for diagnostic purpose.
EMI noise trough both GPU/PSU under significant load can go in I don't believe in such isolators. How bad really is it?
What’s the CPU and is there a HDMI out from MB to try it that way (using discreet GPU)?

My point about there being II class I (I incorrectly used 1 & 2 before) appliances means that there is a "loop" that forms with the grounds: PC ground -> HDMI -> AVR (Class II, unearthed) -> Subwoofer amp (earthed) -> Back to PC through the earth wire.

If this loop is broken, the EMI noise goes away completely. When I use TOSLINK / Optical, the issue does not happen. Because of this, I'm quite sure the RCA isolator would fix it, since it galvanically isolated the subwoofer amp from the AVR through a transformer. But I expect this to screw with the output level somehow?

The GPU I'm using is a discreet GPU currently, but the actual GPU doesn't matter I think, since I can trigger the issue by just touching the HDMI cable to the metal case of the PC.

 

[Groundrod]

My point about there being II class I (I incorrectly used 1 & 2 before) appliances means that there is a "loop" that forms with the grounds: PC ground -> HDMI -> AVR (Class II, unearthed) -> Subwoofer amp (earthed) -> Back to PC through the earth wire.

If this loop is broken, the EMI noise goes away completely. When I use TOSLINK / Optical, the issue does not happen. Because of this, I'm quite sure the RCA isolator would fix it, since it galvanically isolated the subwoofer amp from the AVR through a transformer. But I expect this to screw with the output level somehow?

The GPU I'm using is a discreet GPU currently, but the actual GPU doesn't matter I think, since I can trigger the issue by just touching the HDMI cable to the metal case of the PC.

To solve this issue you first have to confirm with an AC tester that all AC outlets being used are wired correctly.
The problem is that the TV/AVR are not grounded and the PC is! ( that's if the AC outlets are wired correctly).
This noise is usually a result of the AC polarity of one of the outlets being reversed.

cheers!

 

[AnalogSteph]

In a similar scenario (involving active monitors rather than a sub), another member here had pretty good luck with explicitly earthing the AVR. Then the annoying ground currents from the PC are largely diverted (assuming your earth connection has substantially lower resistance than the HDMI cable shield). This may be enough to audibly eliminate the problem, despite a ground loop technically still being there.

 

[Groundrod]

In a similar scenario (involving active monitors rather than a sub), another member here had pretty good luck with explicitly earthing the AVR. Then the annoying ground currents from the PC are largely diverted (assuming your earth connection has substantially lower resistance than the HDMI cable shield). This may be enough to audibly eliminate the problem, despite a ground loop technically still being there.

Technically speaking the term "Ground Loop" is misused a lot in forums, a Ground Loop is a rare occurrence were a system has 2 Grounds from two different transformers, but in a house there is only 1 Ground!

Although, in circuit PCB designs and other sensitive electronics with multiple Ground connections that are not connected to a single point are examples of local Ground loops, these types of loops are experienced in a lab setting and not in the field.

The issue with Ground pertaining to consumer electronics and this issue specifically, has to do with Switch Mode Power Supplies (AKA electronic transformers).

Because SMPS make a lot of RFI/EMI they use line filters and to work properly need to be Grounded.

If the Ground is missing an AC voltage equal to half the mains (~65Vac ) with reference to Ground will exist on its DC/chassis Ground.

So 2 pronged AC powered devices cable connections are carrying 65Vac on there shields until they are connected to a device that is Grounded, but not without a spark!

This is why best practices dictate that when connecting equipment, all power cords should be unplugged first.

Also, if the 2 devices being connected are plugged into different AC outlets and one has a reverse polarity, a current loop will occur because 1 of the RFI/EMI capacitors in the SMPS filter is being shorted out, and results in switching noise being injected into the AC line.

So, for this issue, test both AC outlets and make sure they have the correct polarity and are Grounded. ( AC testers are a dime a dozen at the hardware store )

Then, if the problem still exists, it may be that the 2 pronged plugs are not polarized, in which case, reversing them should be the end of the problem.

PS: I have the same setup, TV/AVR and Shield box!

The TV and AVR need to be Grounded and the best way is to get a power bar with coax connectors on it and run good quality RG6 patch cables to the tuners RF connectors.

I just use the RF output connector of the power bar and use a splitter to get the two connections needed.

Of coarse this should only be done after testing the AC outlets for proper wiring first!

Cheers!

 

[Rubberducky]

Thanks for all the responses!

In the mean time I removed the old wiring and properly wired that outlet with a ground / earth connection. I haven't used the PC very often, but it seems like the noise is almost gone. It's still there, but seemingly much softer.