single wires with a varying current emit radiation. They are aerials.
Single wires collect radiation. They are aerials.
A connection between two modules is a two wire connection.
That two wire connection is usually a close coupled pair of Flow and Return.
Whatever current comes out of the Source MUST RETURN to the Source.
For least radiated emission the two close coupled wires can be a twisted pair.
For least collected/received interference the two close coupled wires can be a twisted pair.
ALL the connections between modules in your amplifier ARE TWO WIRE connections.
They should be at least close coupled throughout the whole route and better if they are a twisted pair throughout the whole route.
if your problem is due to multiple ground loops between the six channels and the preamp, the grounds of p2/r10/r9/c14/c6 should be commoned and go to ground via a 4.7 to 22R.
Hello ctrlx and Thanks!
This has been tried as reported in two of my preceding posts.
luke
Hello
Here is a recent update of the situation.
The good news is that it does not blow resistor anymore like the description in my post #15. Positive!
But I still have the same buzz problem even though I tried different scenarios.
The only Loop I can see is the Inputs to each amp but I will try to "star" each of them.
I'll be short and will come back later with more observations.
I'll just show the revised Gnd schematic, below:
Luke
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Hello
The only Loop I can see is : (all the shields) to (Inputs of each amp' s circuits) to ( star point).
But I tried to connect directly from each Shields to Star point without connecting to amp's input and it's noisier. Worse.
Anyway all these tries are still giving noise:
1) 15 ohms resistors between each shields and amp circuit inputs Gnd.
2) All shields connected together on each channel at connectors or amp's circuit inputs Gnd.
3) All shields to chassis only.
4) All shields to chassis and amp circuit inputs.
5) Broken connection of Earth to switching circuit.( No-contact at for earth/Gnd circuit).
6) Cap coupled only on earth/Gnd circuit
7) Complete coupling on earth/Gnd circuit
Luke
PS. Still surviving but...
The only Loop I can see is : (all the shields) to (Inputs of each amp' s circuits) to ( star point).
But I tried to connect directly from each Shields to Star point without connecting to amp's input and it's noisier. Worse.
Anyway all these tries are still giving noise:
1) 15 ohms resistors between each shields and amp circuit inputs Gnd.
2) All shields connected together on each channel at connectors or amp's circuit inputs Gnd.
3) All shields to chassis only.
4) All shields to chassis and amp circuit inputs.
5) Broken connection of Earth to switching circuit.( No-contact at for earth/Gnd circuit).
6) Cap coupled only on earth/Gnd circuit
7) Complete coupling on earth/Gnd circuit
Luke
PS. Still surviving but...
An externally hosted image should be here but it was not working when we last tested it.
When You have a humming noise, wrong is that everyone thinks that the hum induced in the hot wire of the signal cable. Hum occurs due to the closure of surface ground in which the mains voltage induces a current that flows through the signal line. Solution is to reduce the surface which creates a ground loop. One of the ways the input signals connectors place together and drag them to a central point mass and then branch to each channel with the ground wire from the power supply. Another way is to make a separate power supply for each channel with non-fused masses among them. There is no third way
If there is no nouse with no inputs, I'd first try tipi onderstand how the loop is forming with the ecpxtetnal I input.
If you plug 1 input and there is little or no noise, fix that first.
Plug a second input in and T&C noise goes up. Is a loop being formed between the two pieces of equipment via the power amp?
Are different input types, when plugged into the amp on their own making the same noise level or more or less?
If you plug 1 input and there is little or no noise, fix that first.
Plug a second input in and T&C noise goes up. Is a loop being formed between the two pieces of equipment via the power amp?
Are different input types, when plugged into the amp on their own making the same noise level or more or less?
You a have a common impedance coupling between where the two amps ground connection join and the connection to the ms in PSU.
Are you 100 % sure the incoming ground connection is not making contact with the chassis where it enters? If it does, you will have a ground loop.
Last edited:
On your latest schematic it appears you have the input grounds and speaker grounds tied together, and then to power supply star, is that correct? If so, you may need to change that.
Mike
Mike
The pic showing the layout automatically generates excessive LOOPs when ever any two inputs are connected to a common ground back at the Source.
This LOOP can be broken by separating the input Signal Return from the Amplifier Power Ground.
That breaks the LOOP.
But you must reference the Signal Return to the amplifier's Reference Ground. This is where the Disconnecting Network comes in.
D.Self describes it. Leach describes it. But they use a sole resistor. For added Safety I now use a pair of inverse parallel power diodes and a resistor.
I repeatedly remind Builders that a multiple Channel amplifiers MUST use separate Signal Return from the Power Ground.
This LOOP can be broken by separating the input Signal Return from the Amplifier Power Ground.
That breaks the LOOP.
But you must reference the Signal Return to the amplifier's Reference Ground. This is where the Disconnecting Network comes in.
D.Self describes it. Leach describes it. But they use a sole resistor. For added Safety I now use a pair of inverse parallel power diodes and a resistor.
I repeatedly remind Builders that a multiple Channel amplifiers MUST use separate Signal Return from the Power Ground.
You can see how I solve problem on diy Hiraga amplifier. Hiraga-brum - Stranica 3 - Solid State - diyAudio.rs
just to clarify that adding a low value R to the shield is not the same as i have detailed in post 43.
The Shield must continue the TWO WIRE connection all the way from the Source to the amp Input.
You must NOT break the Shield connection. It is part of the RETURN route and MUST be maintained.
You must NOT break the Shield connection. It is part of the RETURN route and MUST be maintained.
Hello Impuls and thanks!
Do you mean that I have to connect the six shields to a cable, to send them to the star point and again, take six other wires and connect them back to each inputs?
Confirm please!
Luke
Last edited:
Hello Bonsai
Finally you're suggesting a "step by step" application of a solution.
But this is what I've been doing since beginning.
I connect two inputs and use trial-and-error...
Luke
Hello Bonsai and thanks
I would say yes...100% sure.
For test I removed connection as shown by the arrow on the picture below, to remove Earth of the circuit, Earth being still connected to the chassis
I did not measure any continuity between Chassis and any Gnd point on any circuit.
An externally hosted image should be here but it was not working when we last tested it.
Luke
Hello Mike and thanks!
1) Input Gnd "does NOT share path"... from chassis connectors, through shield, through amp circuit, up to the MainPSU's Gnd connections.
2) From the MainPSU's Gnd connections, the wire for SPKR (-) starts to end at the Speaker out connector.
3) Though one can say InputGND and SPKR GND travels together to the star point. But the common path is a single and independant PCB track 0.25" wide by thickness of a 3oz/F² copper to about 2.5 inches (6.5cm) long!
Could it be "IT" ?...ze problem?
Luke
Last edited:
- Status
- Not open for further replies.