Maker Pro
Maker Pro

Transformer noise - what can be done about it?

Eadventure

Aug 1, 2012
5
Joined
Aug 1, 2012
Messages
5
I made a power supply that uses a small audio transformer and a regulator, in addition to a few discretes. I'm getting what I want out of it but, the transformer gives out a high pitch noise. From what I've read this is due to the construction of the core, constructed of plates that vibrate at certain frequencies. The problem for me is I think the core is solid ferrite. I don't want to destroy it to find out for sure. And I'm not positive that I'm right about the plate vibrations anyway.

Is there anything I can do to silence the hum?
 

Laplace

Apr 4, 2010
1,252
Joined
Apr 4, 2010
Messages
1,252
I would try two things. First, use rubber stand-offs to mount the transformer so that it does not contact any hard surface that could amplify the sound. If that does not work then try potting the transformer in slow-curing epoxy. Use a vacuum pump to exhaust the air from the crevices of the transformer. The sound may be coming from loose contact between the wire coil and the core; epoxy should fix that. If cooling is a problem, let most of the epoxy drip off evenly.
 

davenn

Moderator
Sep 5, 2009
14,011
Joined
Sep 5, 2009
Messages
14,011
Hi Eadventure
Welcome to the forums :)

you are using an AUDIO transformer as a POWER transformer ?

doesnt sound like a wise thing to do
best you explain and clarify what you are doing, may post a schematic of the circuit wouldnt go amiss either :)

Is there anything I can do to silence the hum?
yeah, use a transformer designed for the job :)

cheers
Dave
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
A transformer with a laminated core will show as metal when scratched. Ferrite will stay dark.

Ferrite cored transformers are used for high frequencies, they are not suitable for audio or mains.

A transformer needs to have sufficient inductance for the job in hand. Your audio tranformer may not have enough, perhaps because the volyage is too high or the frequency is too low. If there is insufficient inductance, the core will go into saturation and over-heating will occur.
 

john monks

Mar 9, 2012
693
Joined
Mar 9, 2012
Messages
693
Laplace nailed it.
When I hear a buzzing coming from a power transformer sometimes I squeeze Silicon rubber, or poor epoxy, or super glue between the laminations.
 

Eadventure

Aug 1, 2012
5
Joined
Aug 1, 2012
Messages
5
Thanks for the welcome and the replies.

I'm making a radiation detector. The Geiger-Meuller tube requires 400VDC and only 25 microamps. There are a number of examples out there using audio xfmrs for detectors (how successful I don't know), but I went with that. Trying to keep it small.

The circuit is a typical boost design for the regulator I'm using. I don't care to advertise for free but the device has 34063 in its designation. Any large distributor will have the datasheet (with example circuits). The part can switch at 100KHz according to spec. Would that be suitable for use with a ferrite core? I'm considering winding my own if it will be worth the effort.

I had read in the past about potting a xfmr. Since I'm short on vacuum chambers I think I'll try the super glue approach.

I'm no engineer, actually a pcb designer. Been ramping up pretty well with online tutorials since last october. Wish I'd started about 10 years ago. Well ok, 20...
 

Eadventure

Aug 1, 2012
5
Joined
Aug 1, 2012
Messages
5
A transformer needs to have sufficient inductance for the job in hand. Your audio tranformer may not have enough, perhaps because the volyage is too high or the frequency is too low. If there is insufficient inductance, the core will go into saturation and over-heating will occur.

Yup, I've been aware that these things are occuring in my circuit, especially the inductance aspects. I've worked hard to try and keep from saturating and have been fairly succesful. Its a fun, fascinating challenge for me.
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
The MC34063A should be fine. I might get one to produce a high voltage supply for the primary of an electric fence controller.
You will need a ferrite cored transformer and a pot core would be the one to use. You will not need to run it very fast with your minimal load. If you run at ultrasonic frequencies, you will not hear anything and there will not be any laminations to vibrate.
The chip can manage 40V so to get 400V with a bit of spare, you will need a turns ratio of about 15.

A G-M tube measures pulses so you might need a very smooth supply.
 

davenn

Moderator
Sep 5, 2009
14,011
Joined
Sep 5, 2009
Messages
14,011
Thanks for the welcome and the replies.

I'm making a radiation detector. The Geiger-Meuller tube requires 400VDC and only 25 microamps. ...

ahhh now we find out the fun use of the cct :) I built up a kits many years back from an Australian supplier. Oatley Electronics here in Sydney

attachment.php


I didnt get that black meter with my kit and my G-M tube was brass not glass
I have a reasonable number of rock and mineral samples in my collection that make the unit click crazily :)

Dave
 

Attachments

  • K305P.jpg
    K305P.jpg
    38.5 KB · Views: 346

Eadventure

Aug 1, 2012
5
Joined
Aug 1, 2012
Messages
5
Wow very cool! Thats the same tube I have. Looks like a different power supply from mine. I've made a few efforts with BJT and FETs. Even did some darlington configurations. Also tried two successive inductors. They've all worked reasonably well but I want to make this thing with the chip I picked out.
 

davenn

Moderator
Sep 5, 2009
14,011
Joined
Sep 5, 2009
Messages
14,011
do you have anything to use as a particle source ?
of nothing obvious, use one of the smoke alarms that have a small sample of Amercurium in it, they are not a strong emitter but it will make the unit tick happily if held close, within ~ 10 cm

Dave
 

Eadventure

Aug 1, 2012
5
Joined
Aug 1, 2012
Messages
5
Nope, I don't have anything like that yet. I did have some breadboards set up and got the thing to sense ambient radiation, outputting a signal to a tiny speaker and LED. They say you sense 1-20 blips per minute. that's about what I was getting.
 
Top