Forums

transformer core material

Started by Tom Del Rosso August 24, 2021
AIUI you use iron cores for low frequency and ferrite for high frequency 
because ferrite doesn't get magnetized, so why couldn't aluminum do the 
same?


-- 
 Defund the Thought Police 


 Tom Del Rosso wrote:
==================
> > AIUI you use iron cores for low frequency and ferrite for high frequency > because ferrite doesn't get magnetized,
** LOL - wrong. Most loudspeakers use FERRITE magnets !
> so why couldn't aluminum do the same?
** Or a piece of wood ? FYI the whole point of a core is that it DOES get magnetised. And as easily as possible and with the least energy losses when cycled in each direction. Iron / Silicon alloys are is the only materials that do that well at low frequencies. Ferrite has very low losses at at ALL frequencies but cannot compare with iron at low ones for sheer brute force. .... Phil
Tom Del Rosso wrote:
> AIUI you use iron cores for low frequency and ferrite for high frequency > because ferrite doesn't get magnetized, so why couldn't aluminum do the > same?
You *want* a transformer core to be easily magnetized! You don't want it to *stay* magnetized when the current goes to zero. Iron is good in low-frequency transformers because it has a high saturation field and high permeability, so you can get away with relatively few turns for the windings. Its disadvantage is that it is conductive, so there will be eddy current losses, which get rapidly worse with higher frequency. Those losses can be reduced by making the core out of thin insulated laminations, but this gets impractical quite fast. Ferrite has a lower permeability and lower saturation field, but it's an insulator, so it doesn't sustain eddy currents. That's the simple view. Magnetic materials are complicated and lots of effort has been spent on finding the best materials for specific applications. There are hundreds of different kinds of magnetic materials, maybe thousands. Jeroen Belleman
 Jeroen Belleman wrote:
===================
> > Iron is good in low-frequency transformers because it has a high > saturation field and high permeability, so you can get away with > relatively few turns for the windings. Its disadvantage is that it > is conductive, so there will be eddy current losses, which get > rapidly worse with higher frequency. Those losses can be reduced > by making the core out of thin insulated laminations, but this > gets impractical quite fast. >
** Typical iron core transformers intended for 50/60 Hz can be used to 20kHz and beyond with no such issue. As the operating frequency rises, core magnetisation falls cancelling any rise in losses. ....... Phil
Phil Allison wrote:
> Jeroen Belleman wrote: > =================== >> Iron is good in low-frequency transformers because it has a high >> saturation field and high permeability, so you can get away with >> relatively few turns for the windings. Its disadvantage is that it >> is conductive, so there will be eddy current losses, which get >> rapidly worse with higher frequency. Those losses can be reduced >> by making the core out of thin insulated laminations, but this >> gets impractical quite fast. >> > > ** Typical iron core transformers intended for 50/60 Hz can be used to 20kHz and beyond with no such issue. > As the operating frequency rises, core magnetisation falls cancelling any rise in losses. > > > > ....... Phil
Why isn't laminated iron good for RF transformer cores then? Jeroen Belleman
Jeroen Belleman wrote:
> Phil Allison wrote:
-----------------------------------------
> >> Iron is good in low-frequency transformers because it has a high > >> saturation field and high permeability, so you can get away with > >> relatively few turns for the windings. Its disadvantage is that it > >> is conductive, so there will be eddy current losses, which get > >> rapidly worse with higher frequency. Those losses can be reduced > >> by making the core out of thin insulated laminations, but this > >> gets impractical quite fast. > >> > > > > ** Typical iron core transformers intended for 50/60 Hz can be used to 20kHz and beyond with no such issue. > > As the operating frequency rises, core magnetisation falls cancelling any rise in losses. > > > > Why isn't laminated iron good for RF transformer cores then?
** You have misunderstood my post. ...... Phil
Phil Allison wrote:
> Jeroen Belleman wrote: >> Phil Allison wrote: > ----------------------------------------- >>>> Iron is good in low-frequency transformers because it has a high >>>> saturation field and high permeability, so you can get away with >>>> relatively few turns for the windings. Its disadvantage is that it >>>> is conductive, so there will be eddy current losses, which get >>>> rapidly worse with higher frequency. Those losses can be reduced >>>> by making the core out of thin insulated laminations, but this >>>> gets impractical quite fast. >>>> >>> ** Typical iron core transformers intended for 50/60 Hz can be used to 20kHz and beyond with no such issue. >>> As the operating frequency rises, core magnetisation falls cancelling any rise in losses. >> Why isn't laminated iron good for RF transformer cores then? > > ** You have misunderstood my post. > > ...... Phil
It's true I assumed constant Bmax. I should have said so. Constant voltage, as you assumed, is indeed more natural. I measured the frequency response of a few iron core transformers: A 75VA rectangular-core worked well up to 40kHz, while a similar sized toroid went up to only about 10kHz (-3dB), measured between the two independent 15V windings of each. This was with a tiny excitation voltage, which may mess up the results. Jeroen Belleman
Jeroen Belleman wrote:
==================
>>> > >>> ** Typical iron core transformers intended for 50/60 Hz can be used to 20kHz and beyond with no such issue. > >>> As the operating frequency rises, core magnetisation falls cancelling any rise in losses. > >> Why isn't laminated iron good for RF transformer cores then? > > > > ** You have misunderstood my post. > > > > It's true I assumed constant Bmax. I should have said so. > Constant voltage, as you assumed, is indeed more natural. >
** Well, I deal lot with audio transformers - from mic input to hundreds of watts.
> I measured the frequency response of a few iron core > transformers: A 75VA rectangular-core worked well up > to 40kHz,
** Yep. Audio output types go to about 60kHz or more.
>while a similar sized toroid went up to only > about 10kHz (-3dB), measured between the two independent > 15V windings of each.
** That is odd, toroidals are usually the best with -3dB responses to 100kHz. Just the fact the secondary is wound all over the primary does the trick. What cannot be done is having a tiny laminated iron core running at 100kHz and high power. As well as ferrite there are "powdered iron" cores and toroids that will. Another material is "amorphous steel" which as very low losses. ...... Phil
On 2021-08-25 23:21, Phil Allison wrote:
> Jeroen Belleman wrote: > ================== >>>> >>>>> ** Typical iron core transformers intended for 50/60 Hz can be used to 20kHz and beyond with no such issue. >>>>> As the operating frequency rises, core magnetisation falls cancelling any rise in losses. >>>> Why isn't laminated iron good for RF transformer cores then? >>> >>> ** You have misunderstood my post. >>> >> >> It's true I assumed constant Bmax. I should have said so. >> Constant voltage, as you assumed, is indeed more natural. >> > ** Well, I deal lot with audio transformers - from mic input to hundreds of watts. > > >> I measured the frequency response of a few iron core >> transformers: A 75VA rectangular-core worked well up >> to 40kHz, > > ** Yep. Audio output types go to about 60kHz or more. > > >> while a similar sized toroid went up to only >> about 10kHz (-3dB), measured between the two independent >> 15V windings of each. > > ** That is odd, toroidals are usually the best with -3dB responses to 100kHz. > Just the fact the secondary is wound all over the primary does the trick. > > What cannot be done is having a tiny laminated iron core running at 100kHz and high power. > > As well as ferrite there are "powdered iron" cores and toroids that will. > Another material is "amorphous steel" which as very low losses. > > > ...... Phil > > >
On 2021-08-25 23:21, Phil Allison wrote:
> Jeroen Belleman wrote: > ================== >>>> >>>>> ** Typical iron core transformers intended for 50/60 Hz can be used to 20kHz and beyond with no such issue. >>>>> As the operating frequency rises, core magnetisation falls cancelling any rise in losses. >>>> Why isn't laminated iron good for RF transformer cores then? >>> >>> ** You have misunderstood my post. >>> >> >> It's true I assumed constant Bmax. I should have said so. >> Constant voltage, as you assumed, is indeed more natural. >> > ** Well, I deal lot with audio transformers - from mic input to hundreds of watts. > > >> I measured the frequency response of a few iron core >> transformers: A 75VA rectangular-core worked well up >> to 40kHz, > > ** Yep. Audio output types go to about 60kHz or more. > > >> while a similar sized toroid went up to only >> about 10kHz (-3dB), measured between the two independent >> 15V windings of each. > > ** That is odd, toroidals are usually the best with -3dB responses to 100kHz. > Just the fact the secondary is wound all over the primary does the trick.
It surprised me too. I also measured an inter-winding capacitance of 2nF, which strikes me as high. The drop-off was a resonance dip.
> > What cannot be done is having a tiny laminated iron core running at 100kHz and high power. > > As well as ferrite there are "powdered iron" cores and toroids that will. > Another material is "amorphous steel" which as very low losses.
I use transformers for RF. I've used ferrite, of course, but also metglas and similar materials. The cores are mostly there for the low end of the frequency range. Beyond a few MHz, you really want to keep the flux out of the core, which is done by using transmission lines for the windings. Some of my transformers are good to 9GHz, but those bear little resemblance to a traditional wound transformer. Jeroen Belleman