Advice Needed: Can't kill the buzz, but the buzz is killing me.

[wwenze]

I agree with McFly’s assessment. MUST ground AVR chassis! The importance of solid ground connections for audio components is extremely vital.

1) This is up to the designer as well as the government examiner to decide
2) Grounding a Class II appliance when it should not be grounded can transform it from a safe chassis to a dangerous chassis: In the event there is a fault on the secondary side and the chassis is connect to the secondary, earthing the chassis will connect the secondary side to earth and hence negating the safety benefit of an isolation transformer and allows a closed loop to form on the secondary side without tripping the ELCB/RCD

 

[Pecrobet]

1) This is up to the designer as well as the government examiner to decide
2) Grounding a Class II appliance when it should not be grounded can transform it from a safe chassis to a dangerous chassis: In the event there is a fault on the secondary side and the chassis is connect to the secondary, earthing the chassis will connect the secondary side to earth and hence negating the safety benefit of an isolation transformer and allows a closed loop to form on the secondary side without tripping the ELCB/RCD

 

[Speedskater]

a]Modifying the plastic case of a Class II (square within a square safety symbol) appliance (or hi-fi component) to run a 'ground' wire to an internal chassis is very dangerous.
b] But a Class II appliance with an exposed metal chassis can be 'grounded'.
c] Never 'ground' to a stake in the garden.
d] Always 'ground' to the Safety Ground/Protective Earth circuit powering the audio equipment.
e] The 'ground' path is to the Neutral in the main breaker box.
f] The 'ground' path has almost nothing to do with Planet Earth. (actually other connections to Planet Earth [like cable TV] are often part of the problem.

 

[sam_adams]

2) Grounding a Class II appliance when it should not be grounded can transform it from a safe chassis to a dangerous chassis

When a double-insulated audio device—two-wire AC mains connection with no EGC—is connected using audio interconnects to a device with a 'normal' three-wire AC mains with EGC conductor that connects to a grounded outlet, the internal signal ground of the ungrounded device is connected via the shield of the interconnects to the signal ground of the EGC grounded device. At some internal point in the grounded device there will (should) be a single point of connection for that device's signal ground and the chassis or earth ground. Therefore, connecting an 'ungrounded' device to a 'grounded' device establishes a connection to the EGC in the grounded device—and that's where all the fun starts.

I'll let Bill Whitlock spell it out in better detail:

Bill Whitlock - Interconnection of Balanced and Unbalanced Equipment - Jensen AN-003

Bill Whitlock - Hum And Buzz In Unbalanced Interconnect Systems - Jensen AN-004

Bill Whitlock - Hum Buzz and Ground Loops - New Insights Into An Old Problem AES Toronto 2010

Bill Whitlock - An Overview of Audio System Grounding and Interfacing - Indy-AES-2012

Bill Whitlock - Understanding, Finding, And Eliminating Ground Loops - CEDIA Class EST016

Bill Whitlock - The Myth and Mystery of Analog Signal Interfaces - AES Detroit -2021

 

[terryforsythe]

I had a similar issue. The culprit turned out to be my XFinity/Comcast cable. A power strip/surge suppressor with TV cable input/output solved the problem. Actually I tried two. One worked to solve the issue and one did not. The one that worked is made by Monster Power.

 

[wwenze]

When a double-insulated audio device—two-wire AC mains connection with no EGC—is connected using audio interconnects to a device with a 'normal' three-wire AC mains with EGC conductor that connects to a grounded outlet, the internal signal ground of the ungrounded device is connected via the shield of the interconnects to the signal ground of the EGC grounded device. At some internal point in the grounded device there will (should) be a single point of connection for that device's signal ground and the chassis or earth ground. Therefore, connecting an 'ungrounded' device to a 'grounded' device establishes a connection to the EGC in the grounded device—and that's where all the fun starts.

I'll let Bill Whitlock spell it out in better detail:

Bill Whitlock - Interconnection of Balanced and Unbalanced Equipment - Jensen AN-003

Bill Whitlock - Hum And Buzz In Unbalanced Interconnect Systems - Jensen AN-004

Bill Whitlock - Hum Buzz and Ground Loops - New Insights Into An Old Problem AES Toronto 2010

Bill Whitlock - An Overview of Audio System Grounding and Interfacing - Indy-AES-2012

Bill Whitlock - Understanding, Finding, And Eliminating Ground Loops - CEDIA Class EST016

Bill Whitlock - The Myth and Mystery of Analog Signal Interfaces - AES Detroit -2021

Don't worry, in a correctly-designed appliance be it class I or II, the secondary ground will never be directly connected to earth for the same reason mentioned in previous post. Secondary ground must be decoupled from earth to reap the benefits of an isolation transformer. Connection of secondary ground to earth is done through a low-value class Y capacitor (for coupling RF) and a high value >1M resistor for coupling DC i.e. static. This is also the reason why whether an equipment has an earth pin or not has nothing to do with whether you will get the half-AC-voltage "electricity leak" and noise, despite this myth being repeated often even on the more technical forums.

 

[Danonano56]

So before dragging a cable across the house to connect the AVR up to our refrigerator's chassis ground, I decided to bring my UPS over (which has a TVSS ground screw) and try using that for the grounding point.

View attachment 294265

I plugged the UPS into the wall (not the Furman, though I don't think it matters), ran some small 18AWG from the TVSS ground to the phono ground on the AVR, powered up the AVR, powered up the KEF speakers, and... silence!

Well, relative silence - there was still some hiss if you listened for it - but no more buzz! So hot damn, good call @McFly!

Now, interestingly, when turning the AVR off and leaving the KEF speakers on, I still got a harsh buzzing from the KEFs. Nothing near as bad as it was before (it was so loud prior to running this ground cable that I worried it would damage the speakers over time), but still audible to the point of being disruptive.

I didn't plan on having a UPS as a part of my AV stack, but if it keeps the speakers from buzzing I'll do it. Are there any more permanent solutions now that we seem to have the culprit identified? Seems like I just need to find a way to permanently ground the AVR chassis, right?

Thanks again to everybody that's taken time to read this, weigh in, and offer advice. This is why I love communities like this! I seriously can't thank y'all enough - it's been a stressful past couple months, and having something that's supposed to be enjoyable (listening to music / watching shows to unwind) turn into a stressful, frustrating, seemingly unsolvable problem sucks. I'm overjoyed to be making progress on this!

Thanks for posting this, it proved to be a solution for me as well.

 

[restorer-john]

Don't worry, in a correctly-designed appliance be it class I or II, the secondary ground will never be directly connected to earth for the same reason mentioned in previous post. Secondary ground must be decoupled from earth to reap the benefits of an isolation transformer. Connection of secondary ground to earth is done through a low-value class Y capacitor (for coupling RF) and a high value >1M resistor for coupling DC i.e. static. This is also the reason why whether an equipment has an earth pin or not has nothing to do with whether you will get the half-AC-voltage "electricity leak" and noise, despite this myth being repeated often even on the more technical forums.

This is just wrong in so many ways.

 

[AnalogSteph]

At least when it comes to Class I devices, a lot of which actually tend to be more or less hard-grounded to PE (if they aren't generously providing a ground lift feature, at which point the connection tends to shrink to 100R-1k || 100n). It is absolutely possible to make a Class II device with an extra PE connection just for dumping leakage currents into, and even the variety with a grounded chassis but isolated electronics exists, although that is annoying to build in all kinds of ways (once your chassis is not tied to circuit ground and jacks have to be kept isolated, you tend to have all kinds of fun with RFI and shielding).

 

[pablolie]

Can you use Optical? Finding a way to do that is a sure way to decouple.