Humming due to common ground

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello Members,

Here is a Humming Problem that i am currently facing in my Class A amplifier i recently built for my self. I am sure some of Expert members have a simple solution for this problem.

Here is the nature of the problem, I built a stereo Class A amp, each channel has dedicated power supply, and individual amplification boards. The Humming only occurs when i make the GROUND common in both the Channel, When the Grounds are not common the sound is crystal clear.

I have attached a diagram of my current amplifier arrangement, this is the arrangement when there is Humming. If i make the Ground Uncommon between both the sides, The Humming stops instantly.

I'll deeply appreciate kind advises from members, Thank you.

Best Regards.
 

Attachments

  • attachment.jpg
    attachment.jpg
    256.7 KB · Views: 283
Last edited:
Earth the chassis.

Create a Main Audio Ground (MAG) for left channel and another separate one for right channel.

Try connecting one MAG via a Disconnecting Network to the chassis.
Is it quiet?

Try connecting the other MAG via it's own Disconnecting Network to the chassis.
Is it quiet?

The problem you have with your dual mono is that there are two inputs and both are common grounded at their Source end. When these interconnects arrive at the dual mono they get connected again. That creates an enormous ground loop. There are voltage differences around that loop and the two inputs "hear" that AC interference and amplify it to come out of the speakers. You can measure the hum and/or buzz with a DMM set to 200.0mVac attached to the output of the amp.
 
Last edited:
Earth the chassis.

Create a Main Audio Ground (MAG) for left channel and another separate one for right channel.

The problem you have with your dual mono is that there are two inputs and both are common grounded at their Source end. When these interconnects arrive at the dual mono they get connected again. That creates an enormous ground loop. There are voltage differences around that loop and the two inputs "hear" that AC interference and amplify it to come out of the speakers. You can measure the hum and/or buzz with a DMM set to 200.0mVac attached to the output of the amp.

Try connecting one MAG via a Disconnecting Network to the chassis.
Is it quiet?
Yes, It is Quiet

Try connecting the other MAG via it's own Disconnecting Network to the chassis.
Is it quiet?
No, The humming starts when i do this.
 
connect transformer 35V "GND" directly to Filter "GND" (wire is missed in your drawing) - and NOT to common Ground or Chassis!
Do this for left and right channel separately. This transformer "GND" carries high current pikes for charging the capacitors. These should not interfere with signal wires, and also not with "Signal GND".

Same for connecting Filter's output "GND" directly to amp supply GND (the wire which is missed in your drawing). Not to common GND / chassis!
Question: On the amp board, is signal input GND and speaker output GND internally connected to supply GND? If yes, simply connect signal input GND of each amplifier to a single common chassis GND. Thats the GND star point. No supply wires, also no "supply GND" should directly be connected to that star GND.

Regards,
Guthorst
 
Last edited:
connect transformer 35V "GND" directly to Filter "GND" (wire is missed in your drawing) - and NOT to common Ground or Chassis!
Do this for left and right channel separately. This transformer "GND" carries high current pikes for charging the capacitors. These should not interfere with signal wires, and also not with "Signal GND".

Same for connecting Filter's output "GND" directly to amp supply GND (the wire which is missed in your drawing). Not to common GND / chassis!
Question: On the amp board, is signal input GND and speaker output GND internally connected to supply GND? If yes, simply connect signal input GND of each amplifier to a single common chassis GND. Thats the GND star point. No supply wires, also no "supply GND" should directly be connected to that star GND.

Regards,
Guthorst

Thanks for the quote, Will try and update.

On the amp board, is signal input GND and speaker output GND internally connected to supply GND?
Ans) Yes.
 
Thanks for the quote, Will try and update.

On the amp board, is signal input GND and speaker output GND internally connected to supply GND?
Ans) Yes.

That's no good idea. But follow my description and check if it works.

I think, there are some reservoir capacitors on amp board, connected between suppl (+/-12V and +/-35V) and GND. So, every charge and discharge pulse currents for these capacitors travel via dedicated GND wires. This will result in a noisy/humming GND wire. So, you should not use this supply/reservoir ground wire as signal return/signal GND.

The complete GND network should not have any loop. Use only one star point, which is connected to signal inputs (the most sensitive signals). This star point can be the metal enclosure of your amp.

If you had an additional connection from transformer GND or filter GND to that star point, this would result in a ground loop. Because the internally connection inside amp board will form that loop. But if there was not internally GND connection at the amp board, then it was required to connect the filter reservoir GND to your single common star point GND.

Regards,
Guthorst
 
Hello,

I had exactly the same problem with you and did what someone else suggested in this thread and it worked perfectly. I disconnected the input GNDs from left & right star ground points and connected them both straight to the chassis GND, where left and right star grounds meet. I think it will work for you, too.
 
Hello,

I had exactly the same problem with you and did what someone else suggested in this thread and worked perfectly. I disconnected left and right i/p GNDs from left & right star ground and connected them both to chassis ground, where left & right star GNDs meet. Now there's no hum when I short left & right i/p GNDs. I think it will work for you, too.
 
That's no good idea. But follow my description and check if it works.

I think, there are some reservoir capacitors on amp board, connected between suppl (+/-12V and +/-35V) and GND. So, every charge and discharge pulse currents for these capacitors travel via dedicated GND wires. This will result in a noisy/humming GND wire. So, you should not use this supply/reservoir ground wire as signal return/signal GND.

The complete GND network should not have any loop. Use only one star point, which is connected to signal inputs (the most sensitive signals). This star point can be the metal enclosure of your amp.

If you had an additional connection from transformer GND or filter GND to that star point, this would result in a ground loop. Because the internally connection inside amp board will form that loop. But if there was not internally GND connection at the amp board, then it was required to connect the filter reservoir GND to your single common star point GND.

Regards,
Guthorst
Thank you for the useful quote, will try and Update.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.