Author Topic: Toroidal transformer magnetostriction noise management  (Read 559 times)

0 Members and 1 Guest are viewing this topic.

Offline Coordonnée_chromatique

  • Contributor
  • Posts: 23
  • Country: fr
Toroidal transformer magnetostriction noise management
« on: October 24, 2020, 05:16:41 pm »
Hello,
I'm reading this forum since a long time and i have a question that require some expertise in order to get rid of an annoying problem, i can't resist any more because i know that there is some people here that have the ability to answer the right questions.
I'm using a linear PSU and the toroidal transformer is relatively quiet since i've choosen a small one, there is only a tad of magnetostriction noise in general, but there are some days that this noise is getting worse and can become very annoying few times.
I'm planing to immerge the transformer in a silicone matrix located in a polycarbonate box in order to get it away and isolated from me, have you any better idea in order to cure or understand my problem please ?
« Last Edit: October 24, 2020, 05:22:41 pm by Coordonnée_chromatique »
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 3870
  • Country: ca
Re: Toroidal transformer magnetostriction noise management
« Reply #1 on: October 24, 2020, 07:31:29 pm »
Sometimes transformer noise is from poor mains power quality - the higher harmonics can be heard at a higher frequency "buzz". You will notice it may be more noisy during the workday/week and quiet at night/weekend when industrial power usage is less.

How many VA, what mounting hardware are you using?
If you have steel (bolts, chassis etc.) near the transformer, you can get magnetostriction and the steel will vibrate - it's not transformer noise.
I've never heard a noisy toroid unless the mounting hardware was done wrong and acted as a shorted-turn, or the rubber washers were not used.

You can buy toroid shield/can/cover from eBay/Ali and then pot that. Best if the potting compound is non-flammable. Silicone I find it hard to get rid of air bubbles.
 

Offline mzzj

  • Super Contributor
  • ***
  • Posts: 1011
  • Country: fi
Re: Toroidal transformer magnetostriction noise management
« Reply #2 on: October 24, 2020, 07:48:34 pm »
Any DC offset will also break havoc on toroidal transformer cores.
Some crappy power controls like series diode can cause load imbalance that puts some fraction of volt imbalance on AC. Enough to saturate toroid transformers and make them buzz 
https://sound-au.com/articles/xfmr-dc.htm

In my previous apartment wifeys hair dryer caused some toroidal transformers to buzz rather loud.
 

Offline Coordonnée_chromatique

  • Contributor
  • Posts: 23
  • Country: fr
Re: Toroidal transformer magnetostriction noise management
« Reply #3 on: October 25, 2020, 07:19:55 am »
The PSU is composed of a Breve Tufvassons TTS150/Z230, a KBU8K rectifier and one standard 85° samwha 10 000µF for each rail. The mounting skews are made of amagnetic inox steel and the case is made of iron except the sides that are made of alumnium, the noise is changing a little (the harmonics are going up in frequency) when i put a book on the case vents mainly because the internal volume and the small grill row holes acts like a resonnator and must be colsed.
I’m using a 150VA Toroid, as the article mentioning it the input resistance should be high enough and my small transformer should be not affected by the DC (i’ve tested it with no success) but the practical humming is perceptible at less than two meters (if you are concentrated enough with a relatively low noise floor under 40dB(Z or C))
My problem is not really quantification of a weighted meaningless single number because the low frequency humming is extremely disturbing to me due to a large narrow peak of energy in the frequency response, it is triggering a patological injury to my nerve cells at the most sensitive peak of my ears measured transfer function... is litterally a resonnant coupling.
I must wait for the moment that annoy me the most and connect my DSO with a differential probe in order to see how the input waveforms should interact with the Xformer ?

Some crappy power controls like series diode can cause load imbalance that puts some fraction of volt imbalance on AC. Enough to saturate toroid transformers and make them buzz
In my previous apartment wifeys hair dryer caused some toroidal transformers to buzz rather loud.

Sometimes transformer noise is from poor mains power quality - the higher harmonics can be heard at a higher frequency "buzz". You will notice it may be more noisy during the workday/week and quiet at night/weekend when industrial power usage is less.

I have a DSO, a differential probe and an ampermetric probe, would you be able to help me to analyse the AC waveforms of my outlets please ? 
« Last Edit: October 25, 2020, 07:25:58 am by Coordonnée_chromatique »
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 8057
  • Country: de
Re: Toroidal transformer magnetostriction noise management
« Reply #4 on: October 25, 2020, 11:31:37 am »
A DC offset on the grid can cause trouble the toroidal transformers. It would be somewhat hard to see it would the voltage and directly probing mains can be dangerous.  One could look at the effect of DC, causing relatively large spikes in the magnetization current. These should be easy visible with a small current transformer as a scope probe.  For makeshift use one could use a toroidal common mode choke as a current transformer: some 50 Ohms on the secondary side (e.g. both coils in series, so they add up).

Normally toroidal transformers tend to use grain oriented material with relatively low magnetostriction - so there are usually quite low in noise. Only when driven very hard (high voltage or DC offset) there is a little hum. Having  a fuse before the rectifier can also help a little as the added resistance reduces the peak current to the filter capacitor.
 

Offline dietert1

  • Frequent Contributor
  • **
  • Posts: 650
  • Country: de
    • CADT Homepage
Re: Toroidal transformer magnetostriction noise management
« Reply #5 on: October 25, 2020, 02:41:18 pm »
Building a filter to get rid of a small DC voltage (e.g. 1 Volt) is fairly easy and there are schematics proposals on the web. Basically it's a 50 A diode bridge shorted to make a bidirectional diode with 1.4 V threshold and a large capacitor parallel to this, like 10 000 uF/4 V. Yes, you can use a hair drier to test, since it includes a diode for half power mode, so it will take some A of DC current in that mode, resulting in local mains imbalance. Anyway, when the hair drier is on, you may not hear the transformer hum...

But there are bad toroids, too. I remember buying some from ebay that turned out to be completely useless due to hum. Those were huge, though, like 2000 VA. Hum was more or less constant all the time. I tried different measures without success.

Regards, Dieter
 

Offline Coordonnée_chromatique

  • Contributor
  • Posts: 23
  • Country: fr
Re: Toroidal transformer magnetostriction noise management
« Reply #6 on: October 26, 2020, 05:45:24 pm »
resistance reduces the peak current to the filter capacitor.

I've tried to put a power resistor on each rail with a small capacitor in parallel on another epoch and another circuit and it has worked on a 500VA noisy transformer (factory potted talema toroidal transformer)

hair dryer caused some toroidal transformers to buzz rather loud.

you can use a hair drier to test

Yes, i've been sufficently stupid to test if a hair dryer is triggering the transformers noise...  :-DD evidently the hair dryer is a lot noisier than the transformer, could you immagine a safe simple cicruit the can simulate the hair dryer ? 
« Last Edit: October 26, 2020, 06:01:02 pm by Coordonnée_chromatique »
 

Offline dietert1

  • Frequent Contributor
  • **
  • Posts: 650
  • Country: de
    • CADT Homepage
Re: Toroidal transformer magnetostriction noise management
« Reply #7 on: October 26, 2020, 07:41:11 pm »
If you make that simple filter circuit, you can safely assume there won't be any significant DC current on your toroid - at least not on the primary. The same also applies to the secondary: It must not have DC current. That means hum can be caused by certain rectifier circuits that impose DC current onto one of the secondaries. The example i remember were certain rather expensive Backes & Müller active speakers that came with this type of mishap.

Regards, Dieter
 

Offline mzzj

  • Super Contributor
  • ***
  • Posts: 1011
  • Country: fi
Re: Toroidal transformer magnetostriction noise management
« Reply #8 on: October 26, 2020, 07:54:27 pm »
:-DD evidently the hair dryer is a lot noisier than the transformer, could you immagine a safe simple cicruit the can simulate the hair dryer ?
Any large load (like water kettle) wired with beefy diode in series.
 

Offline Coordonnée_chromatique

  • Contributor
  • Posts: 23
  • Country: fr
Re: Toroidal transformer magnetostriction noise management
« Reply #9 on: October 27, 2020, 05:37:14 pm »
:-DD evidently the hair dryer is a lot noisier than the transformer, could you immagine a safe simple cicruit the can simulate the hair dryer ?
Any large load (like water kettle) wired with beefy diode in series.

I remember a colleague at work that have forgotten to put some water in a electric plastic kettle, the magic smoke cloud was very small but black and smelly and i still remember that the dummy load must be quiet after the hair dryer episode... therfore i decided to test a small convector, fail again :

raw recording, no compression https://sndup.net/2478/radiateur.flac
« Last Edit: October 27, 2020, 05:39:35 pm by Coordonnée_chromatique »
 

Offline Coordonnée_chromatique

  • Contributor
  • Posts: 23
  • Country: fr
Re: Toroidal transformer magnetostriction noise management
« Reply #10 on: October 28, 2020, 05:59:52 pm »
For those who are eventually interested by this thread... this home appliance is silent as a dead horse  :horse:

 

Offline Coordonnée_chromatique

  • Contributor
  • Posts: 23
  • Country: fr
Re: Toroidal transformer magnetostriction noise management
« Reply #11 on: October 31, 2020, 10:16:22 am »
Hello gentlemens, you inellectual efforts in order to respond to this thread has been converted into data, and can pehaps be converted again into an inellectual sucess with some additional efforts  \$\Omega\$
Two 53W bulbs with a 1N5401 that gives 53W of power consumption.

Here is the mains AC voltage (yellow) and current (purple)


Here is the mains AC voltage (yellow) and current (purple) with the 1N5401.


There is no DC, could you tell me how many power consumption i need to trigger the DC please ?
 

Offline dietert1

  • Frequent Contributor
  • **
  • Posts: 650
  • Country: de
    • CADT Homepage
Re: Toroidal transformer magnetostriction noise management
« Reply #12 on: November 01, 2020, 02:40:24 pm »
I don't understand how you get 100 mV when measuring mains voltage. How much DC voltage did you expect? How much DC voltage do you need to see it on the scope?

Some years ago i made those experiments, too. I used a hairdrier in another room. Still i could not hear whether the toroid hum changed. Then i built the simple filter circuit i mentioned above. That filter gave me a way to measure the DC voltage (over the capacitor). As far as i remember i detected a fraction of a V, like 0,5 V and i concluded that my local mains circuit including the primary of my toroid had an effective DC impedance of about 0,1 Ohm. I assumed the hairdrier takes 5 A DC at 220 V in half power mode, since its nominal power is around 2 KW. My conclusion was: i did not need the DC filter. It's pure marketing that mains can deliver significant DC, except in special situations like running from a local generator or so.

In your case you installed about 100 W DC load, so your DC current may be about 0.5 A. Since your toroid is 150 VA, it may have a primary resistance of 3 or 5 Ohm and only a small fraction of that DC test current will be short-circuited through the primary of your toriod. When i look at your scope curves, your mains sine seems to have a flattened top i.e. harmonics. This may also generate audible noise. Cure it by testing the toriod without using a long mains cable with many other loads on the same end.

Regards, Dieter
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf