Loop breakers

[felixx]

Two amplifiers are connected to the PSU by +ve, -ve and power ground.
I have two screens from the signal ground to the RCAs. Both RCAs are isolated from the chassis.
When I connect one input plug, still silent.
Now connect the other input plug and look at what I have -A BIG LOOP.
Start at power ground to PCB power ground through PCB to signal ground to input screen to RCA to interconnect to source component across the source component to the other interconnect to the other RCA to the other signal screen to signal ground to PCB power ground to power ground= closed loop= hum.

I must break this loop.How?

[roender]

Hi Felix,
Make two separated PSU, one fore each amplifier, like in a true monoblock but with a single case. Then, you will have only one common ground point, at input of the amplifiers.

Mihai

[richie00boy]

That is one way, but the most complex/expensive one. It is perfectly possible to have no hum with a shared transformer - look how many commercial amps do it.

Are the amp boards made by yourself? Does the input lead screen just go back to the amp board which in turn has a ground back to the star at the PSU?

[AndrewT]

Hi,
a dual rectifier PSU for each power amp will separate the power grounds between the channels. This won't eliminate the hum but it will reduce it significantly. The power grounds must be taken separately through two disconnecting networks to the safety earth.
Better still is a four secondary transformer that completely isolates the two channels. Again two disconnecting networks are required.

Taking the two RCA grounds back to their main audio grounds so that the signal grounds go via the input plug ground may be quieter than running the PCB signal ground to the PCB power ground.

[ricobasso]

A loop is only a problem if the area is large and there is an interfering magnetic field crossing the area. For example, all twin phono leads complete a loop via their screens when they are plugged in at both ends, but, that does not produce hum because the area is small.

Try re-routing your input wires inside the amplifier to reduce the loop area and to get them away from the mains transformer.

[roender]

You may also try to split the power gnd from input gnd trough a 10ohm resistor, like in the schematic below

[felixx]

Mihai....all that is done.
10 ohms parallel with 0,1uf.

And the problem is ONLY when I attached the other channel interconnect to the cd player.
When I run amplifiers separate all is ok.


I replaced the toroid tranny with R-core and the problem disapear....too bad because the R-core was borrowed from a friend.


But now I want to try to modify the star ground to make that 1kw toroid to do the job right....I hope.
I will try to disconnect the case of the amp from the 0V buss rail and
connect a 5 watt 27 ohm R between case and 0V,close to RCA inputs,like in the picture.
The 27 ohms is a much higher resistance than the shielded wire to the cd player,so any ground loop voltage should not appear across the lengths of shielded wires used anywhere.


I suspect the magnetic field of the power tranny is creating some small hum voltage in the earth wiring by means of stray magnetic coupling.

On R-core that is not noticeble.
Anyway an R-core tranny will be GOLD on audio amps.

[janneman]

Is it hum or buzz? Buzz is rectified ground current.

Simple solution: run each screen from it's RCA to the point on the board where you put the signal. For instance, you probably have a resistor from the input to ground on the board. Put the screen on the same ground point as the resistor - or connect directly to the resistor terminal. No connection from screen to any other point.

Then, the speaker return goes to the supply star point for that channel, not to any other point

Jan Didden

[AndrewT]

Hi,
toroid radiated field is very low but often there is a peak at a few locations around the perimeter.
This can oriented to reduce it's affect on the circuit.
Try rotating the toroid around it's mounting bolt and measure the hum change at the output terminals. Flexible leadouts are needed for this and preferably quite long ones.
The mains input leadouts should be twisted as a pair or quad, if dual primary. The secondaries should also be a twisted triplet right up to the rectifiers, then the centre tap goes directly to the power ground.

[felixx]

janneman...all these was done.
AndrewT-All the a.c. wires is twisted.

The only thing I CAN'T do is to move the toroid from the place....The position is verticaly-mounted.I don't have too much place for that but tomorow I will try to place the tranny from some distance and see what's happend.
With R core the amp is dead.With toroid the problems apears.
Stray effect is the cause...I'm 90% sure.