Transformer HUM.

[zdr]

Hello folks,

Need some help in trying to eliminate transformer hum. It's not a buzz, it's a hum. It's audible from the transformer itself, and also on speakers, from a meter away. Transformer is 500VA 2x25V Nuvotem Talema:

0500P1-2-025 | 2 Output Toroidal Transformer, 500VA, 2 x 25V ac | RS Components

What I tried so far:

1. Steel shielding cover placed, no change
2. DC blocking circuit from here: Mains DC and Transformers, no joy.

The only thing left that comes to mind is to purchase more expensive transformer from RS which is specifically declared as very low induced noise:

0500P1-2-025K | 2 Output Toroidal Transformer, 500VA, 2 x 25V ac | RS Components

 

[zdr]

 

[JonSnell Electronic]

The transformer you have looks like a potted toroid type.
They are normally low noise.
Make sure there is no electrical connection to the fixing bolt in the centre of the transformer, if there is it will produce a noise as it is a shorted turn.
There is also a connector block that looks like it has been screwed to the transformer .... I hope not!

 

[zdr]

The transformer you have looks like a potted toroid type.
They are normally low noise.
Make sure there is no electrical connection to the fixing bolt in the centre of the transformer, if there is it will produce a noise as it is a shorted turn.
There is also a connector block that looks like it has been screwed to the transformer .... I hope not!

The transformer is the one from the first url, with added outer steel shell. PCB is attached to outer shell, not the transformer itself. So it's not potted, nor declared as "low hum/low noise", unlike the one from second url. There is no connection from center screw to outer shell. It hummed the same way before shell was added anyway.

 

[zjjwwa]

Which problem is the bigger one?
Physical mechanical hum of the transformer, Or the hum from the speakers?

 

[DF96]

Low induced noise does not mean low acoustic noise, so the alternative transformer may be no better.

Did you use the 'rubber' washers when mounting the transformer? Did you tighten the bolt too much?

 

[zdr]

Which problem is the bigger one?
Physical mechanical hum of the transformer, Or the hum from the speakers?

Hum from the speakers, although they sound very similar.

 

[zdr]

Low induced noise does not mean low acoustic noise, so the alternative transformer may be no better.

Did you use the 'rubber' washers when mounting the transformer? Did you tighten the bolt too much?

Yes, used rubber washers on both top and bottom. Did not tighten the bolt too much, as it would bend the bottom chassis plate otherwise.

 

[zjjwwa]

Did you tightly twist all relevant pairs of cables together?
Such as:
Twisted pairfrom transformer to rectifier bridge?
From rectifier to the input legs of filter cap?
From filter cap output legs to amp?
...
Did you treat each of the filter caps as a Four legged device?

 

[scottjoplin]

Are you hearing the hum from your speakers or have you measured it? That kind of low frequency noise can travel around in strange ways.

 

[zdr]

Are you hearing the hum from your speakers or have you measured it? That kind of low frequency noise can travel around in strange ways.

I can hear it. Did not try to measure it.

 

[zdr]

Did you tightly twist all relevant pairs of cables together?
Such as:
Twisted pairfrom transformer to rectifier bridge?
From rectifier to the input legs of filter cap?
From filter cap output legs to amp?
...
Did you treat each of the filter caps as a Four legged device?

1. yes
2. they are on the same board, very close.
3. too short to twist.
4. ?

 

[scottjoplin]

I would be tempted to measure to confirm

 

[zjjwwa]

4. ?

As in: no "T" junctions. The capacitor has physically 2 legs. But for the sake of lowest hum, I always treat them as 4 legged devices: 2 inputs and 2 outputs.

 

[zdr]

As in: no "T" junctions. The capacitor has physically 2 legs. But for the sake of lowest hum, I always treat them as 4 legged devices: 2 inputs and 2 outputs.

They are on PCB, layout part of awsome RMI-FC100 design, pretty much looks as you described.

 

[zjjwwa]

How are all the ground returns mapped out?
Speaker return?
Zobel return?
Input stage differential pair return?
Nfb decoupling cap return?
Rail decoupling cap returns?
How are these connected to mid point of main PSU filter caps? What sequence?
...
No offence. Just fishing for possible trouble spots.

 

[zjjwwa]

Do you have a single (as in not more than 1) connection between Protective Earth and Ground?
Is it a direct hard wired connection or via a low ohm resistor?
Example 2.2 ohms / 10W?

 

[zjjwwa]

Did you try foil 47nF, rated for a voltage at least 2.2 times higher than the DC Rail voltage, such capacitors connected in parallel to each of the rectifier diodes?

 

[zdr]

There is no connection between audio ground and chassis. Chassis is connected to mains earth.

 

[scottjoplin]

That could be a problem, usually best that audio ground is connected to the chassis