Transformer hum...?

Been reading the S300 thread with interest, as the issue of transformer hum has cropped up.

[For those who don't suffer from it, and don't want to start hearing it turn away right now ]

It would be useful to understand reasons/cures for hum a little more, as most dealers and manufacturers just end up saying 'learn to live with it'

From my v. limited knowledge, there is hum:

- caused by ground loops, due to earthing, heard thru the speakers;

- and hum which is mechanical from the transformer itself. Why should transformers hum for this reason. Isn't it just poor design? Esp as some amps hum and others don't.
Is the hum due the windings/laminations rattling. It seems toroidal transf. are more sensitive to hum, esp if there is dc leakage into the transformer (not that I understand why). Or is it something also to do with the frequency of the ac current, and if this varies too much, the transformer may hum. Intermitent hum seems weird too. Is this just random, or induced by variations in the ac?

Anyhow, would be better to hear from more knowledgeable folks

cheers

 

I was concerned about transformer hum when I built my power amp but the huge 800VA toroidal I used is silent as the night. No audible hum even if I put my ear right up to it when it's working. I believe that transformer hum is caused if there's any DC component on the mains. The PSU I used from LC Audio has a DC filter before the transformer which they claime eliminates hum. I don't know if I would have had hum without it but I certainly don't have any with it

Michael.

 

hmm I am by no means an expert with transformers, much the opposite, I am only just beginning to understand them.

Having tinkered with valve amps. tho' with cheap tx's I can now say a couple of things.

The ones I had were designed by a clever guy, but built on the cheap, too a price.

Generally, tx makers try to get away with the cheapest they can, as with all things, usually cheaper means more pain in the long run.

Toroids have small cores, but less radiation, and they are more susceptible to dc, which is zero hz, which is the enemy of them.

The cores are magnets, and have a finite amount of magnetism, the bigger, the better, as they can magentise more, which means lower freq, or room to spare, visualize the size as car engine size, bigger=better.

it also depends on material, usually, cheapish iron is used, more expensive ones use silicon steel, with additives such as aluminium, nickel, cobalt, incidentally, the same as speaker magnets, alnico. there are only so many magnetic metals. also manganese.
These cost more.

The laminations have to be thin, and insulated from each other.

Then there is the coil, or copper, the thicker, the more current. higher volts demand better insulation.

If the current is too high, the magnet saturates, causing buzz, also DC, causes saturation, and toroids with their small cores don't like dc. Cores have to have magnetic resistance, a piece of paper, to limit the saturation.

Cheapness? cheap steel, insulation, operating the tx underated, ie small core, wire only just sufficietn, cheap steel.

you get what you pay for here.

If it buzzes, only cure is a new tx.

You can get copper bands to limit the radiation in el cores(hum induced by transformer, not by core buzz, and electrstatic shields, tx design is hugely complex, simpler for mains, where the freq is just 50 hz.

Its all about core size, no of turns of copper, size of copper, metal, frequency, and all that.

for induced hum you can't beat a toroid, even better, potted. Toroids cores don't seem to buzz as much, perhaps for lots of current.

summary

core buzz: too high current for too small a core(ie cheap skates), dc on the mains toroids hate, loose laminations, poor winding, poor insulation/glues.

only just found this out, if you have a yank tx designed for 60 hz, and our mains is 50 hz, it needs more iron core for the extra low freq., so will not be 100% happy with it, could buzz. lower freq=bigger core.

induced hum: toroids better

conclusion, spend a lot, get a good brand, or someone who understands them, not to a price. mount well away from small ac signals.

Hum in circuits, don't have a bloody clue, wiring, all sorts of things ,got some myself, devils job finding and curing

 

There's two main effects for txfmr in a piece of hifi kit.

One is loose windings/laminations. They aren't held together tightly enough and can vibrate. Potting will help fix this, as will tighter, more even windings.

The other is "magnetostriction". This is where the core of the transformer itself is subjected to such intense magnetic fields that it starts changing shape in the varying field. This will only occur when the transformer is heavily loaded, and can be avoided by using a much heftier spec transformer than necessary.

I believe the first effect is that which causes the hum in the S300, as it occurs at idle, when the txfmr isn't heavily loaded.

 

Interesting stuff LCD.
Probably a lot of your clear explanation has passed over my head

In the case of a tx humming when it was idling, it wouldn't be cos of a high ac current?
And does dc leakage occur when idling?
I get intermitent hum in this situation

As u can tell, I still know bugger all about this

Edit, just read Isaac's post!
It would still be useful to know if dc leakage is the culprit when the amp is idling, as I've come across some mains filters which claim to trap dc

 

DC can be removed by a common-mode choke (wrap the power leads around a toroidal ferrite core), or with an isolation transformer (1:1 primary-secondary transformer). DC can cause humming in toroidal transformers, and this would be the likely cause of a variable hum (ie if it changed late at night or something). However, I've not personally observed any variability in the S300 hum.

 

Ta Isaac for the info. The ferrite ring trick would be relatively simple & cheap I would imagine. Any ideas where these are available?

BTW, thought it might interest folks on the following explanation given by unison research on their transformer. It's refering to a usa user who I've been in contact with. Though, in the uk, this shouldn't be a problem as we use 50hz too. So, their answer doesn't still make sense to me Unless they are refering to dc leakage and not saying it. Or it's just their 'flawed' design. (Emailed them, but not got a reply)

..."Regarding your problem, noting what you explained, I think that the problem is due to the transformer's vibration. In our current production we are using a more performing filter capacitors, which require a bigger current during their charge in turn-on phase. During such a phase, or however for very few seconds, the power transformers must give high energy. The humming may be more evident also because of the higher mains frequency you have in your area (here in Italy it is 50Hz), so the charge current increases of about 20%. Moreover, in case the amp is used in areas in which the electrical situation is not good and has disturbed mains, the noise can increase more. Therefore, noting the above, I think the problem can be considered as normal..."

 

My Marantz CD6000OSE LE has hummed slightly since it was brand new, and my sisters Cambridge Audio CD32 hums quite badly too. I did buy a Fusion CD player for my sister from Richers before the Cambridge, the hum was so bad, it was like a substation so I took it back.

Do you think my slight hum from my Marantz is normal if not how I can easily fix it without replacing it?