how to get rid of transformer hum

[case-sensitive]

first I used a kit. It did buzz. Then I got the schematic of that kit and made my own layout. Buzzes as well. It did buzz at the old flat in Europe and still buzzes here in Australia.
You reckon, the speaker cables feed back, not enough dampening?

 

[spec]

first I used a kit. It did buzz. Then I got the schematic of that kit and made my own layout. Buzzes as well. It did buzz at the old flat in Europe and still buzzes here in Australia.
You reckon, the speaker cables feed back, not enough dampening?

First may I say that it would not be damping.

I am always banging on about instability, but with audio amplifiers that is often the root cause.

Now that you have explained that the amps are specials, I am not at all surprised that you have hum etc, especially with the chips you have chosen, which are heap powerful.

In passing, just to mention that I have been designing and building audio amps of all kinds for years and in nearly every case there has been similar problems to yours.

If you look at the schematic of a commercial audio amp, you will often see all sorts of weird components scattered about, mainly capacitors. They are there to cure the many little annoyances that arise.

With any audio amplifier, let alone the high power version you have, the layout is super critical as are many of the component types.

Could you post a schematic of your amplifier and also a picture of the physical layout.

spec

 

[case-sensitive]

found an 8.2 ohm resistor, the buzz gets a bit less but its still there. Power consumption in both cases is identical. Even when the speakers are not connected.
The easiest way out will be a switch mode power supply. As its quiet when powered by the lab supply. When on the lab supply, the pure 50Hz tone, I have to be 5cm away from the driver to just so hear it.
The amps are at 30dB amplification, I will wind dit down to the minimum they do, about 22. When switching them on for split seconds there are 2 V in the bass drivers (2 in bridge), that nearly pushes the membrane out of the driver. I reckon I should wind down the supply voltage. Its not about being loud, its about producing a clear sound.

 

[spec]

found an 8.2 ohm resistor, the buzz gets a bit less but its still there. Power consumption in both cases is identical. Even when the speakers are not connected.
The easiest way out will be a switch mode power supply. As its quiet when powered by the lab supply. When on the lab supply, the pure 50Hz tone, I have to be 5cm away from the driver to just so hear it.
The amps are at 30dB amplification, I will wind dit down to the minimum they do, about 22. When switching them on for split seconds there are 2 V in the bass drivers (2 in bridge), that nearly pushes the membrane out of the driver. I reckon I should wind down the supply voltage. Its not about being loud, its about producing a clear sound.

I would not advise going for a switch mode power supply.

You should be able to fix the problem with your original power supply using the toroidal transformer.

Can you post the exact schematic of your power supply including all component values and types?

By the way, these are the main ways to get hum in the audio chain:

(1) Capacities coupling
(2) Inductive coupling
(3) Ground current
(4) Ripple voltage

spec

 

[case-sensitive]

top left supply for the relay that switches mute on/off

top right supply for crossover

below transformer 3pass crossover

underneath crossover the power supply

 

[spec]

top left supply for the relay that switches mute on/off

top right supply for crossover

below transformer 3pass crossover

underneath crossover the power supply

No reservoir capacitors or are they underneath?

I hope you don't mind me saying this, but I suspect layout problems.

The two auxiliary power supply transformers will also be chucking out qute a magnetic field

Like the toroid and the case.

spec

 

[case-sensitive]

the accessories are on the line of 100mA or less. The relais draw something like 10mA at 5 Volt, the crossover no idea, worst case 100mA at +-15V. I did not want to run them off the main transformer.
Here power supply. The Cs across the diodes are film, have to replace them with ceramic.
I disconnected the 2 small transformers, does not make a change.

 

[spec]

PLEASE OBSERVE SAFETY PRECAUTIONS- NEVER TOUCH AN EXPOSED CIRCUIT THAT IS POWERED FROM THE MAINS SUPPLY

the accessories are on the line of 100mA or less. The relais draw something like 10mA at 5 Volt, the crossover no idea, worst case 100mA at +-15V. I did not want to run them off the main transformer.
Here power supply. The Cs across the diodes are film, have to replace them with ceramic.
I disconnected the 2 small transformers, does not make a change.

Wow! nice power supply.

I assume that the torodial transformer has two center-tapped secondary windings, making a total of six wires.

Would you like to try this:
(1) Completely remove all mains supplies from the amplifier.
(2) Disconnect the toroidal transformer from the amplifier (6 wires + mains)
(3) Physically remove the toroidal transformer from the amplifier
(4) Connect six substantial wires, about two feet long, to the secondaries of the torodial transformer.
(5) Twist the two non-center tap wires from each of the two secondary windings together.
(6) Connect the six wires to the power supply board
(7) Place the amplifier and toroidal transformer on a non magnetic non-conductive surface (wood for example)
(8) Place the torodial transformer as far as possible from the amplifier and any input or output leads.
(9) Connect the toroidal transformer to the mains supply.
(10) Switch the mains supply on.

Check if the mains hum has gone (you can try changing the orientation of the toroidal transformer but do not touch anything while the mains is switched on).

spec

 

[case-sensitive]

I messed up the supply, transformer has 2 centre and 1 each end, so I had to do some magic. Anyway, that was my next plan. 22:30 in here now, see if I get it done

 

[spec]

I messed up the supply, transformer has 2 centre and 1 each end, so I had to do some magic. Anyway, that was my next plan. 22:30 in here now, see if I get it done

OK no rush- I have to do some building work pretty soon anyway.

Can you post a schematic showing the transformer primaries and secondaries?

spec

 

[case-sensitive]

building work, got me sidetracked from this project .-)
here the secondary conns. Had to wire that op on the fly as the transformer has 4 secondary conns. Will do the wiring tomorrow. Primary goes from power at the back in to a switch on the front panel then back to the transformer, the other to the power in ar the back.

 

[case-sensitive]

ok, did that, transformer about 1m/3 feet away. any other transformer inside the housing disconnected. Power for mute in/out relais from lab supply.
Hum is still there. But, when I switch the mains transformer off, mute is off (as in speakers running of the power supplys caps while they are discharging) the hum is still there. Something must be oscillating. Now remember, the crossover sits over /on top of the power supply. When I disconnect the cable from the xover to monoblock in, there is small white noise, cable back on, buzz is back. So next step is replace foil caps parallel to rectifier diodes in power supply with ceramics, as that seems to add some riesistance and it acts as a snubber. But still, without transformer, why is it still oscillating? Crossover is not powered.
With monoblocks in mute, it draws 1.6 or so W, monoblocks on 12W. The transformer does not get louder wirh more load.
diodes are these
**broken link removed**
thread on bypass resistors on diodes
https://www.diyaudio.com/forums/power-supplies/133916-bypass-capacitors-diodes.html

 

[spec]

Hi CS,

ok, did that, transformer about 1m/3 feet away. any other transformer inside the housing disconnected. Power for mute in/out relays from lab supply.
Hum is still there. But, when I switch the mains transformer off, mute is off (as in speakers running of the power supplys caps while they are discharging) the hum is still there.

Hmm, interesting, but not unusual.
Can you confirm this:
Configuration
(1) There is no mains connected to the amplifier whatsoever.
(2) The amplifier is in no magnetic fields: near to cables carrying mains, TV, fan etc
(3) The amplifier is on a non magnetic electrically insulating material.
(4) The mains torodial transformer is outside the amplifier case 1m away and is powered by the mains (230V 50Hz)
Symptoms
(1) With the above configuration the hum can be heard from both left and right speakers
(2) When the mains supply is completely removed from the transformer (live neutral and earth) the hum is still present from both speakers while the reservoir capacitors still have a voltage across them.

Now remember, the crossover sits over /on top of the power supply. When I disconnect the cable from the xover to monoblock in, there is small white noise, cable back on, buzz is back.

Can you also confirm this. With the configuration above, if you disconnect the two cables connecting the xover to the monoblocks, the hum stops and all you can hear is some background noise. Is that background noise normal and acceptable to you?

So next step is replace foil caps parallel to rectifier diodes in power supply with ceramics, as that seems to add some resistance and it acts as a snubber. But still, without transformer, why is it still oscillating? Crossover is not powered.
With monoblocks in mute, it draws 1.6 or so W, monoblocks on 12W. The transformer does not get louder wirh more load.
diodes are these
**broken link removed**
thread on bypass resistors on diodes
https://www.diyaudio.com/forums/power-supplies/133916-bypass-capacitors-diodes.html

I hope you do not mind me saying this, but it would be best to avoid considering any solutions at this stage until the root cause of the problem is established.

But just to say that from the symptoms, the fault is unlikely to be connected with the diodes or suppressor capacitors (to suppress diode commutation noise) across the diodes .

It is unwise to use ceramic capacitors in an audio amplifier, except possibly as suppressors across the rectifier diodes, but even so, polypropylene film capacitors would be a better choice. Ceramic capacitors not only generate distortion but their capacitance varies with voltage, temperature, and age. Ceramic capacitors can also act as electrical/sound transducers. These shortcomings are outweighed by positive characteristics for other applications.

Looks like no building work for me today- it's mild but pouring with rain.

spec

PS: I see no expense has been spared on components- those film capacitors must have cost a buck or two.

 

[spec]

Hi again CS,

Could you post a schematic of the xover board giving exact component types?

I see some ceramic capacitors, which look like decouplers, on the xover board.

Where do the supply rails for the xover come from. Can you supply exact details of the xover power rails?

Where did the xover board come from?

spec

 

[case-sensitive]

the earth is still connected to the chassis as I am a bit scared to have it completely floating, voltage wise. Edid: earth (green yellow) disconnected, it becomes quieter.
the xover I did, ther ceramic caps are 100nF in the rails for the +-15 V to the opamps. The xover is not connected to anything, except that cable to the input ot the mono block. This amp is just 1 channel, no need to worry about other speakers.
There is a power cable in the wall about 600mm away, goes vertical. And the only thing it does is power my experiment, so not much magnetic field there.
its all MKPs and metal resistors, opamps are OPA134, then a bunch of Mundort tin caps.
Ok remover all the clutter around and listen to it again. But the buzz kept going with the lab supply for the mute relays unplugged, until its caps dischaged and the relais tripped.
The only thing that goes in is 2x22V DC from the transformer and 3V DC, the minimum to switch the mute relais on. The relais are galvanically separated from the monoblocks. Thats the last thing I can get rid off, feed 3V in from batteries.
For prices, does Reichelt supply to the UK, everything in there feels like half price.

 

[spec]

the earth is still connected to the chassis as I am a bit scared to have it completely floating, voltage wise
the xover I did, ther ceramic caps are 100nF in the rails for the +-15 V to the opamps. The xover is not connected to anything, excapt that cable to the input ot the mono block. This amp is just 1 channel, no need to worry about other speakers.
There is a power cable in the wall about 600mm away, goes vertical. And the only thing it does is power my experiment, so not much magnetic field there.
its all MKPs and metal resistors, opamps are OPA134, then a bunch of Mundort tin caps.
Ok remover all the clutter around and listen to it again. But the buzz kept going with the lab supply for the mute relays unplugged, until its caps dischaged and the relais tripped.
For prices, does Reichelt supply to the UK, everything in there feels like half price.

OK. Can you try this:
(1) Put the torodial transformer back in the amp (put the amp back to its full normal configuration).
(2) Check that the hum is still audible from both speakers
(3) Disconnect the cable from the Xover to the left monoblock. Is the hum absent from the left speaker but present on the right speaker?
(4) Connect the left cable back. Is the hum present from both speakers
(5) Disconnect the cable from the Xover to the right monoblock. Is the hum absent from the right speaker but present from the left speaker?
(6) Disconnect both wires to the monoblocks. Is the hum absent from both speakers.

spec

 

[case-sensitive]

I tried again, there was another earth cable to the scope, but the scope was not connected to anything. When I took that one off and the other earth, now the buzz is quieter without earth. I put those molex CAT 5 connetors in, as in the past I got rid of noise with CAT5 cables. This time it makes things only worse.
The transformer, there is 2 pairs of cables. Each pair has 22V across, but overall I dontget 44V. So No idea, if its 2 windings parallel to each other, then they have the same phase, fair enough then. Can only hope there is no offset.
At the rectifier diodes, there are 2 100nF parallel to each diode. Maybe get rid of one. There are divided opinions, foils or ceramics parallel to diodes. Multisim does not simulate it.
http://www.hagtech.com/pdf/snubber.pdf
maybe need some resistors in series to these caps

the inputs of the monoblocks are really sensitive, evan just touching the cables increases the buzz, or pulleing them off. They are shielded already. Maybe the inlet resistance is to low, but thats based on a schematic for a 80W monoblock, just did the routing so I can get spareparts.

 

[case-sensitive]

ok.
1. Amps with transformers in
-pull cable off on xover, so there is 20cm free cable as antenna to monoblock input. Buzz gets louder, more like sawtooth. Pull cable off on monoblock. Buzz is gone. If I put a jumper over the inlet (shorten it to GND) its completely quiet.
2. Amps without any transformer in, connected to lab supply
-pull cable off on xover, leave on monoblock. Loud hum (just low frequency, less overtones), louder than 1. But only when I touch the cable. Otherwise its quiet, extra ground or not. Pull cable off on monoblock, quiet.

Inlet caps on the monoblock would not fix that, as its not DC.

When running of the transformer, the amp with transformer inside is louder than the one with transformer 1m away.

Double checked, the schematic I got is very close to the datasheet from ST

edit: have to redo the xover to suit 24dB, 12dB does not have enough overlap. Then could boost the signal, put bigger resistance into the monoblock, so it gets rid of the noise.