Transformer is making a noise - how to cure?

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

I have a problem that I hope someone can help me with. I have a power amp that uses a toroidal transformer and is humming. Don't say it's because it doesn't know the words hence why the transformer hums :)

As soon as it is switched on, the noise appears if you put your ear to the transformer. Actually it is noticeable from a couple of feet away. However, if no signal is connected to the input, no noise appears in the speakers but as soon as you connect the inputs to the pre-amp, the noise appears in the speakers. It goes away with an isolation transformer, but I don't want to spend much money on it as it is a spare power amp. The amps are Soundstream monoblocks. I suspect the noise is due to one of two problems:

1. There is a significant amount of DC in the AC power supply hence why it is humming.

2. The windings on the transformer may not be as fixed as it should be so it is causing minor movements that induces the hum.

I don't think there is a ground loop problem at all, although it is possible there is a small contribution. The main problem is the transformer hum. As the power amp is old, I would not want to spend a lot of money to rectify this problem and ideally want to DIY a solution. I suppose I can dismantle the toroidal transformer and re-epoxy the windings to decrease any movement or even superglue each winding individually, is this likely to result in an improvement?

Secondly, does anyone know where or how I can DIY myself a DC filter for the incoming AC power line before it connects to the transformer? That way I figured it would eliminate the DC current before it gets anywhere near the toroidal transformer. What components do I need for this and how do I do this? Thanks in advance for any answers.................
 
The most easy way is to return the transformer to the guy who built it and tell hem the problem. The problem can be fixed in a few minutes.

If somehow you have to fix it by yourself, find a segment of the core close the winding end that is not fixed well. The segment is inserted to form a closed core loop when the windings are almost done. If the segment is not glued well at both ends with the bigger core segment, the transformer will cause you headache when powered.

Hope this helps and good luck to you!
 
DC mains filter is easily made with a cap wired in series with the mains and two antiparallel diodes for protection. The cap should have high capacitance but needn’t be high voltage. 10000uF 25v would be plenty. Search the forum for “dc mains filter” and I’m sure you’ll find lots more on the subject.
You say you get hum in your speakers when you connect a preamp and no hum if it is disconnected. This is not caused by your transformer, but most likely a ground loop.
On a side note, I have experienced that cheap bridge rectifiers have contributed to an increased level of transformer hum. But that is less likely to be your problem.
 
Hi,

Thanks very much for all the answers, much appreciated! I can't return the transformer because it is in a commercial amp, not a DIY one so I didn't buy the transformer on it's own :cannotbe:

I think I'll try the mains filter. Just to clarify before I do anything, connect it in seires with only the live wire of the mains? What value of antiparallel diodes? Would two 1N5404 do? Would this filter be good enough for power amps which are rated at 150W continous power into 8ohm (300W into 4 ohm)?

Last bit - no, it's not a ground loop as it makes the noise without a pre-amp connected but only a RCA plugged into the input. If you don't plug anything into the RCA input, the transformer still makes the noise but it doesn't come thru the speakers whilst as soon as anything is plugges into the input the noise appears thru the speakers as well as the transformer.
 
1n5404 would do just fine. I’m using 1n5406 myself and that is on 230v mains. You can make one filter for the live or you can make two, one for live and one for neutral. A schematic for a complete filter is found here http://www.diyaudio.com/forums/showthread.php?s=&threadid=2080&perpage=10&highlight=&pagenumber=2
This uses chokes and some other stuff that you can skip if you only want dc filtering.
For heavy currents consider getting caps with high ripple current capabilities. In fact you should do that anyway.
To clarify what I said about ground loop. Your transformer hum is not caused by a ground loop, that is impossible. But if the transformer hum somehow managed to find its way to your speakers the hum would be there regardless if something was connected to the inputs or not. That’s way I think you are dealing with two problems. Noisy transformer and a ground loop.
 
Hi Kermit,

Thanks again for your helpfulness :) I will make the filter with the diodes and a 10,000microF 50V capacitor to start with. Any suggestions as to what rating of capacitor is best or is this suitable? Just worried that may not be able to handle the amps (around 4 amps, it is also running on 240V AC mains like yours)

Hmmm, will check out the ground loop thing too.......
 
Hi,

Just a quick update: have made the filter with the one cap and two diodes. Good news is the transformer noise/buzz is definitely less (2dB less as measured on my SPL meter) but still more than when I plug it into an isolation transformer (which stops the noise completely).

So any ideas as to what additional things I can try to reduce the noise further? Would adding another filter to the neutral line help? I've only got it on the live line at the moment. Or do I need to change to a higher capacity cap in view of the incoming 240V AC instead of 110V AC?

Last question - does the quality of diodes matter as the 1N5404 I am using are really cheap stuff.......
 
audionutter said:

Hmmm, will check out the ground loop thing too.......


If transformer noise gets to the speaker when the pre-amp is connected to input but not when pre-amp is connected, it doesn't sound like amp power supply circuit problems. If it was a bad filter cap or rectifier this noise would probably occur all the time. The problem sounds like it is related to an input. Try to connect just a plain old resistor accross the input, value that matches input impeadence, and see if the noise occurs. Then add small caps & or coils and see if you can get the noise to occur. (You may be getting RF noise related to the input impeadence and this could affect bias.) If it does not, the problem is probably with the pre-amp. If it does, then the amp circuit has problems.

Try this......measure the current bias that the collector of the output transistors take. Break the circuit from collector of one output and the voltage rail and insert a DC ammeter. Compare the current bias(with no signal) without the pre-amp connected and no noise, to the bias current when the pre-amp is connected and noise occurs. If this bias increases, the amp circuit may have RF noise bouncing around in there (which can affect the DC bias on some amps) or bias problems. PSU ripple voltage is related to the amount of current drawn.

Have you checked this problem whith a different input source? Always try to eleminate the obvious first. ;)

BTW ground loops are a pain in the keister to track down.
 
Hi,

Hmmm, I wonder if I am confusing people with my descriptions? Let's try again:

The main problem is a noise/buzz/hum coming from the toroidal transformer. This noise is present in the transformer from the moment it is switched on and the same noise is reproduced thru the speakers when the RCA input has something connected to it (doesn't matter if it is just an open ended RCA cord or a pre-amp or CD or anything else).

The same noise goes away completely IF the amp/toroidal is connected to an isolation transformer. Even the noise from the speakers. So that is why I do not think it is a ground loop problem, but rather that of DC current leakage into the transformer. The noise is reduced by 2dB with the DIY filter that I have just built as suggested by Kermit (thanks, man!) but any ideas as to what additional things I can try to reduce the noise further? I don't want to use the isolation transformer as it is a cheap, spare, power amp. Would adding another DIY filter to the neutral line help? I've only got it on the live line at the moment. Or do I need to change to a higher capacity cap in view of the incoming 240V AC instead of 110V AC?

Last question - does the quality of diodes matter as the 1N5404 I am using are really cheap stuff.......
 
audionutter said:
Hi,

Hmmm, I wonder if I am confusing people with my descriptions? Let's try again:

The main problem is a noise/buzz/hum coming from the toroidal transformer. This noise is present in the transformer from the moment it is switched on and the same noise is reproduced thru the speakers when the RCA input has something connected to it (doesn't matter if it is just an open ended RCA cord or a pre-amp or CD or anything else).

The same noise goes away completely IF the amp/toroidal is connected to an isolation transformer. Even the noise from the speakers. So that is why I do not think it is a ground loop problem, but rather that of DC current leakage into the transformer. The noise is reduced by 2dB with the DIY filter that I have just built as suggested by Kermit (thanks, man!) but any ideas as to what additional things I can try to reduce the noise further? I don't want to use the isolation transformer as it is a cheap, spare, power amp. Would adding another DIY filter to the neutral line help? I've only got it on the live line at the moment. Or do I need to change to a higher capacity cap in view of the incoming 240V AC instead of 110V AC?

Last question - does the quality of diodes matter as the 1N5404 I am using are really cheap stuff.......


So the noise is present when there is no RCA connection? Or is only when a wire(RCA) is connected. Is there any condition that the noise is not present when the power transformer is connected to a line not through an isolation transformer?

If there is no damage to the transformer; change it with another one and see if the same results occur. Also try inserting an AC ammeter in series with the primary coil and see how much current is flowing with the secondary completely disconnected from the circuit. There should be very little AC current pulled from the line when the secondary is open circuited. Also measure with ohmmeter to make sure there is no leakage from primary to secondary.
It seems that if the transformer is buzzing and is not damaged then there is probably current taken from it. What kind of enclosure is the amp in? sometimes mounting a transformer on metal (steel or any other metal with iron) the magnetic field around the transformer can cause vibrations and rubber gromets are needed in mounting to absorb the movement of the metal vibrations. A metal case is however much better because the metal enclosure will help to shield RF noise out.

I was thinking that if the preamplifier case was referenced to neutral and the amp case was plugged in backwards and amp case was referenced to hot, this could gererate noise. But if the noise occurs, when just a wire is connected to the input, this is an unlikely scenario.

My previous thought is that if there is no wire connected to the input and no noise was generated, and when a wire is connected to the input, noise was generated, then the wire could be acting like an antenea and introducing RF noise into the amp circuit and cause it to occilate at radio frequencies. This radio frequency noise can cause the DC bias to be unstable or just amplify the RF and cause the amplifier to draw current. This would cause a larger ripple voltage on the PSU rails, and cause 120Hz noise (100Hz for 50Hz line) in the speaker and cause the transformer to vibrate because current is being taken. You wouldn't hear RF noise and would need a scope to determain this.
BTW... do you have a scope?

Quick question...When the amp is on and makes the noise, does the output devices get any warmer then when there is no noise?
It sounds like there is a current drain someware, that is of course if the transformer isn't damaged. 240V or 110V doesn't matter if the transformer is rated 240V for the primary and has 240V connected to it.

As for those diodes, I'm sure they wouldn't cause any problems if and would be just fine if they aren't damaged. Bad rectifier diodes or filter caps would cause noise all the time.
 
Well, some good news. I've got some positive results and managed to stop the transformer noise. Well, not quite stop completely but reduced to a very, very low level of less than 40dB (below the measureable limits of my SPL meter). Achieved this by wiring two caps in series, each with an opposing diode paralleled up. That way DC is completely blocked, better than the other way of one cap but two opposing direction diodes in parallel.

Now that the problem is overcome, I have another question. How can you be sre the capacitors will be enough for the current the power amp uses? The AC voltage is 240V and power amp fuse is rated at 4amps. So what value of capacitors do I need? I am using two capacitors in series at the moment - only 10,000 microFarads and 50V, will try to get higher value ones when they are in stock. Any suggestions?
 
You won't be needing verry large caps because the caps only pass smaller AC currents below the 0.7 Volts or so of treshhold voltage of the diodes.
Once the voltage is > 0.7 Volts, all current will flow thru the diodes.
If there is a DC component present in the AC, these are mostly smaller than 0.7 Volts, but can be enough to cause the transformer core to saturate short term (and cause the humming) it will be blocked by the caps.
 
If I understand you correctly - you've did something like this:
dcblocker.gif

(I've found picture on diyparadise)
 
Hi
the AC goes through the back to back caps. measure the volts drop using your dmm. At switch on the high current pulse will generate a larger volts drop & the diodes short this out to prevent damage to the caps.
You need parallel diodes to ensure that current can flow both ways if the voltage rises too high. You need back to back caps or bipolar caps because the AC has a plus & minus component.
I have seen some circuits using two diodes in series then paralleled to give 1.4 volts before shorting out this would better suit 240vac supplies. because the caps only see a maximum of the diode drop across them they only need to be low voltage types plus a generous factor of safety.
I have also seen the caps in the neutral line only. I believe this is to avoid the high voltages being exposed to prying fingers, there is a lot of circuit and components operating at line level if the circuit is in the live lead.
 
I don't recommend very high value capacitors since mains line suffers DC fluctuations, DC component is not constant. Too big capacitors will produce also saturation when DC value changes.

Use something with low enough impedance at 100Hz to prevent the diodes from conducting due to the own ripple of the capacitors [dependent on actual peak current consumption]
 
Diagram of Audionutter's solution.

Hi,
I am following this discussion with interest as I am having similar troubles with a noisy transformer. Audionutter, as you seem to have cracked it now, can you please tell us which, if either, of the two diagrams above is the one which describes best what you have done to cure the problem.
If neither is correct would you please post the diagram of your solution.
Thanks, hopefully........................
 
I wouldn't try the circuit found/posted by Stabist....

You will probably be sitting in the dark after a loud bang!:hot:

If you follow the flow of current you will see that it flows through one of the capacitors during the whole duration of each half cycle.
The diodes do not protect/shunt the capacitors correctly.

I didn't test myself the circuit posted by me, but if you search for the original thread you will read experiences of people who did it with the correct result.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.