Rate of change of core saturation

Main Page | Report this Page

Science Forum Index » Physics - Electromagnetic Forum » Rate of change of core saturation

Page 1 of 1

Author

Message

Theo Markettos

Posted: Tue Mar 18, 2008 7:16 am

Guest

Imagine I have an air-cored transformer. This has some frequency response
up to (say) 100MHz where it tails off.

Then I insert a ferrite core. This increases the mutual inductance so I
have a greater frequency response at low frequencies. But at high
frequencies the response is (probably) worse than the air-cored case.

What effect is causing this? Is it just as a result of adding the core that
I increase the inductance of the transformer so I can't get much current to
flow in the primary? Is it that hysteresis in the core is converting much
more energy into heat - but there's still primary current flowing? Is it
eddy currents (but air and ferrite are insulators)?

What I'm trying to work out is what's going on in the core. I have a core
with a given susceptibility. I apply a strong AC field (so we can neglect
inductance). What's the quantity that controls how much magnetisation there
is in the core at a given time? In other words, the magnetic domains inside
the material need time to align. In air this is minimal (there's nothing to
align, unless you count diamagnetism), while there quite a lot of alignment
to be done in a ferr[o|i]magnetic material. So what sets this time constant
and what is it called?

Thanks
Theo

Wimpie

Posted: Tue Mar 18, 2008 11:29 am

Guest

On 18 mar, 18:16, Theo Markettos <theom+n...@chiark.greenend.org.uk>
wrote:

Quote:

Hello Theo,

Talking about traditional transformers (not transmission line types)
and wire length << 0.25 lambda, HF performance is dictated by
capacitance between turns of own windings and capacitance between the
two windings. Another one is the the leakage flux. It is the flux
generated by the primary that does not encircle the secondary.

The leakage inductance is the inductance that you measure when you
short circuit the secondary. The more air between the two windings,
the more the leakage (flux in between the two windings), the higher
the leakage inductance and the lower the high frequency roll-off. That
is the reason that in many HF transformers secondary and primary
winding are made simultaneously (so primary wire close to secondary
wire).

Adding a core generally does not greatly affect the leakage flux (so
the leakage inductance), but it does affect the magnetization
inductance. This is the inductance that you measure with open
secondary.

When driving a transformer with no ohmic losses with a constant
voltage amplitude and frequency, the total flux is not depended on
whether a core is present or not as flux is dictated by number of
turns and Vs product. Only the resultant current (at given drive
situation) to establish a certain flux is far lower in a coil/
transformer with appropriate core material.

What exactly happens in the core, is a different story. When you use
a low frequency MnZn power ferrite at (for example) 250 MHz. The
material will behave like a strong lossy capacitive medium. It can be
such that due to EM wave propagation there will be less flux in the
core (with respect to the situation withoug core). The core more or
less behaves as metal, and RF field will not penetrate to metal. So
using the wrong type of magnetic material may reduce the magnetization
inductance.

In a macroscopic DC view, Flux density (B, Vs/m^2) = Permeability (H/
m) * Magnetic field (A/m). In a macroscopic HF situation, it depends
on the properties of the magnetic material (that affects EM wave
propagation speed) and the size of the core with respect to wavelength
within the core material. For microscopic properties, you might check
various core producers like Epcos, ferroxcube/yageo, fairrite, etc.

Hope this helps a bit.

Best regards,

Wim
PA3DJS
www.tetech.nl
remove abc from the mail address.

Theo Markettos

Posted: Wed Mar 19, 2008 9:31 am

Guest

Wimpie <wimabctel@tetech.nl> wrote:
[big snip]

Quote:

Hope this helps a bit.

Thanks, it does a lot.

A bit of explanation about the other half of my question. I'm thinking of a
fluxgate-type device here, where it's possible to measure the background
field by seeing how much flux you have to apply before the core goes into
saturation. They're pulsed at a fairly low frequency (say 10KHz) and I was
wondering what the upper limit was set by - is that because of
inductance/capacitance of the device (let's assume linearity isn't
important)? Ignoring that (assume the core is in some strong
preexisting AC field), what controls how quickly a core can go in and out of
saturation? Or is it not possible to get to this situation and other EM
effects take over?

Theo

Wimpie

Posted: Thu Mar 20, 2008 10:34 am

Guest

On 19 mar, 20:31, Theo Markettos <theom+n...@chiark.greenend.org.uk>
wrote:

Quote:

Wimpie <wimabc...@tetech.nl> wrote:

[big snip]

Hope this helps a bit.

Thanks, it does a lot.

A bit of explanation about the other half of my question. I'm thinking of a
fluxgate-type device here, where it's possible to measure the background
field by seeing how much flux you have to apply before the core goes into
saturation. They're pulsed at a fairly low frequency (say 10KHz) and I was
wondering what the upper limit was set by - is that because of
inductance/capacitance of the device (let's assume linearity isn't
important)? Ignoring that (assume the core is in some strong
preexisting AC field), what controls how quickly a core can go in and out of
saturation? Or is it not possible to get to this situation and other EM
effects take over?

Theo

Hello Theo,

When talking about low frequency varying current and fields,
Saturation depends on many factors. For closed magnetic circuits
without air gap (for example a ferrite or wound ring core), the
magnetic path length, Bsat and permeability determines the amount of
ampere*turns to get saturation. So for closed circuits, magnetic
properties of the material and path length dominate.

When an air gap is present, or a completely open circuit (like a
ferrite bar as used in AM/LW radio receivers), the larger the length/D
ratio, the less Ampere turns are required. Search for effective
permeability of ferrite bars (check Ferroxcube, Epcos, Amidon, etc).
For open circuits, field path through air can play a dominant role and
will always lead to higher Ampere*turns to get core saturation.

The AC response depends on the cross section of the magnetic material
and the material itself. Real metals, except the very thin ribbons,
have poor frequency performance. This is mostly because of eddy
current phenomena. The thinner the ribbon (in case of wound cores or
half cores), the better the high frequency performance (up to above
100 kHz). Thin (amorphous) metal strips are used as tags in Electronic
Article Surveillance.

Ferrite comes in a many varieties. The choice in ferrites with steep
rectangular BH curve, is limited. The low permeability materials have
good frequency response (above MHz frequencies), but BH curve isn't
square shaped.

In generally, you can calculate the flux inside a core from the
inductance, magnetic cross section and number of turns.
B = I*inductance/(turns*crosssection).

For your application, I do not expect capacitance to be a problem
(unless you are going to used many turns in combination with very low
drive current).

Best regards,

Wim
PA3DJS
www.tetech.nl
remove abc from the mail address.

Theo Markettos

Posted: Mon Mar 31, 2008 9:08 am

Guest

Wimpie <wimabctel@tetech.nl> wrote:

Quote:

For your application, I do not expect capacitance to be a problem
(unless you are going to used many turns in combination with very low
drive current).

Thanks for that. It's not really an 'application', it was just a
hypothetical example trying to understand the second-order effects. I think
I understand a little better now :)

Theo

Page 1 of 1 All times are GMT - 5 Hours The time now is Tue May 13, 2008 5:47 pm Science Forum Index