My aged hifi has recently been introduced to my new Samsung smart TV and is now producing pleasing sound which is better quality than the smart television itself. There is just one problem, mains hum. My hifi which is comprised of a Marantz CD50, and a Denon TU-260 tuner feeding an Audiolab 8000a which is driving Mission 763 speakers has always had some 50hz mains hum, right from when it was new in the mid 1980s. However recently I have moved to a new area and suspect the hum is louder. I have tracked the source of this hum down to the amplifier, specifically the amps power supply’s toroidal transformer. Just to confirm this was done with the amp disconnected from all the other components and with its lid off. In this state you can clearly trace the hum to the transformer in the 8000a. I have tried loosening/tightening the toroids mounting bolt and rotating the coil to no avail. Reading around the web I have found that this type of hum can be caused by high mains voltage which saturates the cores magnetic field producing magnetostriction and mechanical vibration, i.e. hum. I have measured my mains supply and found it to be 252 volts, which just about conforms to UK specs. The label on the transformer reads Talema 7529P1S2. Talema have a impressive website and catalogue of transformers but I cannot find anything about this particular transformer. I read somewhere its a special model for audiolab.
Now for my big question, has anyone actually sorted this hum by winding some extra turns into the primary winding of the transformer. I have seen this suggested on a forum, 20 turn were suggested. I guess the theory goes this will reduce the magnetising current and noise. I have lots of questions? but if it will fix the hum problem and also make the amp run a little cooler I am interested, not that this one runs particularly hot.
Propose to break into the primary winding where the coil emerges from the toroid and will need to wind in the correct direction, its a 50:50 guess but should be able to identify correct direction. propose to add 5 turns then test then 5 more up to 20 turns total then evaluate.
Any suggestions for about what wire to use, I have some insulated 2mm 28 gauge .3mm stranded wire rated at 17.5A which I think is intended for automotive use so is probably completely inappropriate as I don’t know its breakdown voltage. I was thinking of doing a test run with this wire and the amp supplied through a light bulb current limiter. Final installation with varnished transformer wire (of suitable dimension) with extra insulation layer of kaptan tape above and below.
I am out of my depth in the precise details of the specifics of this sort of problem but about 5 years before I bought this hifi I trained in this sort of stuff and then worked as an engineer for 30yrs. I take the safety aspect of electricity very seriously, which is why I am doing a lot of reading, thinking and consulting those that know more than I.
Sorry for length of post. P
Now for my big question, has anyone actually sorted this hum by winding some extra turns into the primary winding of the transformer. I have seen this suggested on a forum, 20 turn were suggested. I guess the theory goes this will reduce the magnetising current and noise. I have lots of questions? but if it will fix the hum problem and also make the amp run a little cooler I am interested, not that this one runs particularly hot.
Propose to break into the primary winding where the coil emerges from the toroid and will need to wind in the correct direction, its a 50:50 guess but should be able to identify correct direction. propose to add 5 turns then test then 5 more up to 20 turns total then evaluate.
Any suggestions for about what wire to use, I have some insulated 2mm 28 gauge .3mm stranded wire rated at 17.5A which I think is intended for automotive use so is probably completely inappropriate as I don’t know its breakdown voltage. I was thinking of doing a test run with this wire and the amp supplied through a light bulb current limiter. Final installation with varnished transformer wire (of suitable dimension) with extra insulation layer of kaptan tape above and below.
I am out of my depth in the precise details of the specifics of this sort of problem but about 5 years before I bought this hifi I trained in this sort of stuff and then worked as an engineer for 30yrs. I take the safety aspect of electricity very seriously, which is why I am doing a lot of reading, thinking and consulting those that know more than I.
Sorry for length of post. P
I have most of old (from 90's) large toroids buzzing as they were built for 220v, in new house I have over 240 occasionally.
Instead of rewinding primary, which is quite impossible, much more elegant is if you can use one small transformer /1/5th of main trafo size) and wire it as external bucking transformer. Original amp will be intact, but you need extra chassis and safety wiring for this extra trafo.
Here is complete description how to get them back where they are ok, not only not humming, but also keeps secondary voltage where it should be.
https://sound-au.com/articles/buck-xfmr.htm
Hope it helps, for me it works very good.
Instead of rewinding primary, which is quite impossible, much more elegant is if you can use one small transformer /1/5th of main trafo size) and wire it as external bucking transformer. Original amp will be intact, but you need extra chassis and safety wiring for this extra trafo.
Here is complete description how to get them back where they are ok, not only not humming, but also keeps secondary voltage where it should be.
https://sound-au.com/articles/buck-xfmr.htm
Hope it helps, for me it works very good.
Put a resistor or a capacitor in series, common in fans, we had to put 10E / 10W in series with 50W LED flood lights as the lights were fed 270 when the plant supply was unloaded.
We had supplied lights to an Aluminum plant, and too many failures under warranty led to a site check...they were running on diesel generator when the grid supply was down for weekly off or maintenance. Going to 310 in extreme cases.
Best is a step down transformer.
Or simply change the transformer to an E-I type, toroidals are more sensitive to mains noise I think. That should have a 250V primary, of course.
I would not try to repair an old transformer, the copper goes stiff, and too much risk of coil breakage, meaning expensive repair.
A power conditioner connected externally will work, they are called voltage stabilisers here, auto or normal transformers with relay or triac switching of tapping.
IGBT versions, changing output voltage dynamically, are the latest I have seen.
Ferro resonant type is also an option.
We had supplied lights to an Aluminum plant, and too many failures under warranty led to a site check...they were running on diesel generator when the grid supply was down for weekly off or maintenance. Going to 310 in extreme cases.
Best is a step down transformer.
Or simply change the transformer to an E-I type, toroidals are more sensitive to mains noise I think. That should have a 250V primary, of course.
I would not try to repair an old transformer, the copper goes stiff, and too much risk of coil breakage, meaning expensive repair.
A power conditioner connected externally will work, they are called voltage stabilisers here, auto or normal transformers with relay or triac switching of tapping.
IGBT versions, changing output voltage dynamically, are the latest I have seen.
Ferro resonant type is also an option.
Last edited:
Hi
The toroid likely needs a DC mains blocker in series with its primary. Back in the 1980s there were not so many SMPSs powering things, but now your house is filled with them - including CFL and LED lamps.
The noise from those switching supplies is asymmetric on the mains and represents a DC offset. Large PTs have a problem with this because they have low-z windings. Low-power PTs have higher resistance and are not effected by the "DC".
The DC blocker is easy to make. Begin with a 25A bridge.
Short the ACs together.
Short the DCs together.
Now you have a pair of anti-parallel 50A diodes.
parallel this with as high a value of electrolytic cap as is economical to find.
Parallel the first cap with an identical second cap in opposite direction.
The caps block DC; the diodes limit the voltage both ways.
This is a common feature in large amps built since the end of the '90s or so.
Regarding adding extra turns to a toroid: That is frankly a ridiculous suggestion unless you want more noise and it does not address the noise source. As a point of interest, each winding on a toroidal transformer must completely cover the core, so the beginning and end of the winding are right beside each other - share the same lead-out point. This means that each winding shields the care, as well. You cannot half partial loops as with an EI transformer.
The toroid likely needs a DC mains blocker in series with its primary. Back in the 1980s there were not so many SMPSs powering things, but now your house is filled with them - including CFL and LED lamps.
The noise from those switching supplies is asymmetric on the mains and represents a DC offset. Large PTs have a problem with this because they have low-z windings. Low-power PTs have higher resistance and are not effected by the "DC".
The DC blocker is easy to make. Begin with a 25A bridge.
Short the ACs together.
Short the DCs together.
Now you have a pair of anti-parallel 50A diodes.
parallel this with as high a value of electrolytic cap as is economical to find.
Parallel the first cap with an identical second cap in opposite direction.
The caps block DC; the diodes limit the voltage both ways.
This is a common feature in large amps built since the end of the '90s or so.
Regarding adding extra turns to a toroid: That is frankly a ridiculous suggestion unless you want more noise and it does not address the noise source. As a point of interest, each winding on a toroidal transformer must completely cover the core, so the beginning and end of the winding are right beside each other - share the same lead-out point. This means that each winding shields the care, as well. You cannot half partial loops as with an EI transformer.
Last edited:
Hmmm... If you think you are out of your depth, check the transformer's secondary voltage when the amp is idling (the total draw for that amp should be around 500mA max on the DC side), and then get another one made with primary @ 250V AC mains voltage.
... good. You shouldn't struggle at all, then.
Adding windings is a lot of hassle, it is unsafe, and will cause the windings to become loose (more ringing).
Hi Drbulj,
Thanks for your suggestions.
I am not convinced the addition of turns to the primay is not possible but I can't prove it so I don't have a leg to stand on at the moment.
The bucking transformer option looks promising. I need to do some reading and then hopefuly put it into practice
Thanks again P
I have worked on a number of Audiolab 8000s and they all hum. The design of the chassis with virtually no ventilation means internally they run hot. Evidence is the burn marks on the PC board and not good for components. Best advice has already been given, replace the mains transformer. Simplest and least risk option.
Hi Nauta, thanks for your reply.
From what I can see the hum is 50hz mains caused by the supply voltage being relatively high at 252volts, causing noise through magnetostriction, mechanical movement in the core and windings
I agree that switch mode power supplies and CFL and LED lamps are a modern source of noise but I am not convinced this is the cause in this instance, as the hum is very smooth, not dirty as I would expect of some of these other noise sources.
On the subject of winding turns on toroidal transformers it would be fair to say I am fairly clueless and I am looking for information on the subject.
Overall I am not an expert in this subject and may well be barking up the wrong tree but I do know that high supply voltage is a factor and I am working to correct that.
again many thanks P
Hi ,this seems like sounds advice, any suggestions as to where to go to to get a good transformer made. Adding windings was a suggestion I came across on the web, made by someone who appeared to be speaking from a position of knowledge and experience
thanks P
Hi Totally Analogue,
I have to agree with your coments about the tendency of these amps to run hot but in its defence my amp while showing some signs of running hot it doesn't display the excesive heat and damage I have seen on other 8000as on the web. If I get the hum fixed I will replace the 1/4 w resistors that get hot with 1/2 or 1 w resistors mounted clear of the board for better circulation.
Any suggestions on whare to get a replacement transformer that wont hum?
Thanks P
I've only recently returned to the UK so not up to speed with suppliers so I would probably look at Farnell. Make sure you check the height that is a small chassis. As mentioned above the uneven waveform due to the modern fad for switch mode psu can be a problem. If you have a socket close to your incoming supply try plugging in there and see if it improves.
Same issue, toroids are less forgiving of mains variation and noise.
Putting another toroid is not a good idea.
And in India, the quality of one-off toroids is doubtful, the factory made units are better, but not easily available everywhere, tend to be expensive.
The winding machines, wire quality, and proper winding on the cores, then impregnation are in need of better quality than with E-I, and the toroid core has to a continuous strip of 0.3 mm or so electrical steel, several meters long, and absolutely flat.
Better put a 10E/20W wire wound resistor in series with mains, see what happens...it is only about $2.
Then think of other solutions.
Putting another toroid is not a good idea.
And in India, the quality of one-off toroids is doubtful, the factory made units are better, but not easily available everywhere, tend to be expensive.
The winding machines, wire quality, and proper winding on the cores, then impregnation are in need of better quality than with E-I, and the toroid core has to a continuous strip of 0.3 mm or so electrical steel, several meters long, and absolutely flat.
Better put a 10E/20W wire wound resistor in series with mains, see what happens...it is only about $2.
Then think of other solutions.