[Julius Junghans]

hi,
is it true that theres a posibble higher Bmax value (not Bsat) if the core is gapped resultet by a lower remanence?

regards
julius

 

[Tim Wescott]

hi,
is it true that theres a posibble higher Bmax value (not Bsat) if the core is gapped resultet by a lower remanence?

regards
julius

Gapping the core does two things: When the core is gapped you end up
storing much more energy in the magnetic field associated with the gap
than in the core. It also reduces the sensitivity of the coil
inductance to changes in instantaneous permeability of the core as it
saturates.

The first means that even after you've added turns to bring your
inductance up you can run more current for a given core B, and you can
run the core up to a higher B without seeing as much effect at the coil
terminals (although you'll experience more core heating, which may or
may not be a big issue).

 

[Genome]

| hi,
| is it true that theres a posibble higher Bmax value (not Bsat) if the
core is gapped resultet by a lower remanence?
|
| regards
| julius

No.

There are two currents that flow in a forward converter transfomer
primary. One is the reflected output current, one is the current in the
magnetising inductance.

If you gap the core then magnetising inductance drops and the ripple
current due to it increases. The peak and peak to peak flux remain the
same.

Note that the equation for calculating turns is

N = Vin.Ton/B.Ae

Makes no reference to the magnetic path length, Le, or associated gap,
Lg. Flux is independant of these terms.

Unless you are trying to develop power, say for auxilliary supplies,
during flyback on transformer reset there are few reasons to gap the
transformer.

DNA

 

[Julius Junghans]

| hi,
| is it true that theres a posibble higher Bmax value (not Bsat) if the
core is gapped resultet by a lower remanence?
|
| regards
| julius

No.

There are two currents that flow in a forward converter transfomer
primary. One is the reflected output current, one is the current in the
magnetising inductance.

If you gap the core then magnetising inductance drops and the ripple
current due to it increases. The peak and peak to peak flux remain the
same.

Note that the equation for calculating turns is

N = Vin.Ton/B.Ae

Makes no reference to the magnetic path length, Le, or associated gap,
Lg. Flux is independant of these terms.

Unless you are trying to develop power, say for auxilliary supplies,
during flyback on transformer reset there are few reasons to gap the
transformer.

DNA

thx for your awnser, last question: if the remanence decreases with a gapped core it would be possible to start at 0H but with a lower B value right?
found a picture in my smps book from A. pressman that arguments with that explanation, the Bmax increase is about 800G with some ferruxcube material.
Bmax should be 1000G without an gap and 1800G with a gap of about 0.2 to 0.4 mils.

 

[John Popelish]

Perhaps you are confusing the Bfield with the Hfield.

Bsat for ferrites, irrespective of the gap, is in the order of 300mT to
400mT, 3000G to 4000G (at 100C). If you operate the core at higher
frequencies you have to reduce the value, Bmax, because core losses go
up.

Plots of the hysteresis loop are of B versus H where H is NI/Le ampere
turns per meter.... but you may be using Oersteds.

With a gap the required H field to drive the core to any given flux
density is increased.

However, in a transformer with a gap the magnetising inductance is
reduced so ripple current increases in proportion.

Although the core can support an higher H field that increase is negated
by the higher ripple current in the windings. The equation for
calculating required turns for a given Bmax doesn't make reference to
the H field.

(The following is a not so well thought out thought. Criticism
welcomed.)

There may be a small beneficial effect of adding a bit of air gap to a
forward converter core that makes the extra magnetizing current
worthwhile. It has to do with the shape of the BH loop.

To exaggerate the effect, assume a perfectly closed core that also has
a perfectly square BH loop. There will be an output the first time
the core is magnetized, but the squareness of the loop means that the
core has no ability to demagnetize itself. So the second (or third)
cycle of the forward converter will saturate the core. Adding a thin
gap linearizes the net BH loop, providing assistance in damaging the
core when the applied voltage is halted. The optimum gap is a trade
off between this benefit (and the net increase in flux swing possible
against the increased leakage inductance between windings and the
increased magnetization current. If the core material had a
completely flattened BH loop (no remnance) then the optimum gap would
be zero.

 

[Genome]

| On Sun, 04 Jul 2004 21:50:05 GMT
|
| >
| > | > | hi,
| > | is it true that theres a posibble higher Bmax value (not Bsat) if
the
| > core is gapped resultet by a lower remanence?
| > |
| > | regards
| > | julius
| >
| > No.
| >
| > There are two currents that flow in a forward converter transfomer
| > primary. One is the reflected output current, one is the current in
the
| > magnetising inductance.
| >
| > If you gap the core then magnetising inductance drops and the ripple
| > current due to it increases. The peak and peak to peak flux remain
the
| > same.
| >
| > Note that the equation for calculating turns is
| >
| > N = Vin.Ton/B.Ae
| >
| > Makes no reference to the magnetic path length, Le, or associated
gap,
| > Lg. Flux is independant of these terms.
| >
| > Unless you are trying to develop power, say for auxilliary supplies,
| > during flyback on transformer reset there are few reasons to gap the
| > transformer.
| >
| > DNA
| >
| thx for your awnser, last question: if the remanence decreases with a
gapped core it would be possible to start at 0H but with a lower B value
right?
| found a picture in my smps book from A. pressman that arguments with
that explanation, the Bmax increase is about 800G with some ferruxcube
material.
| Bmax should be 1000G without an gap and 1800G with a gap of about 0.2
to 0.4 mils.

Perhaps you are confusing the Bfield with the Hfield.

Bsat for ferrites, irrespective of the gap, is in the order of 300mT to
400mT, 3000G to 4000G (at 100C). If you operate the core at higher
frequencies you have to reduce the value, Bmax, because core losses go
up.

Plots of the hysteresis loop are of B versus H where H is NI/Le ampere
turns per meter.... but you may be using Oersteds.

With a gap the required H field to drive the core to any given flux
density is increased.

However, in a transformer with a gap the magnetising inductance is
reduced so ripple current increases in proportion.

Although the core can support an higher H field that increase is negated
by the higher ripple current in the windings. The equation for
calculating required turns for a given Bmax doesn't make reference to
the H field.

DNA

 

[Julius Junghans]

| Genome wrote:
|
| > Perhaps you are confusing the Bfield with the Hfield.
| >
| > Bsat for ferrites, irrespective of the gap, is in the order of 300mT
to
| > 400mT, 3000G to 4000G (at 100C). If you operate the core at higher
| > frequencies you have to reduce the value, Bmax, because core losses
go
| > up.
| >
| > Plots of the hysteresis loop are of B versus H where H is NI/Le
ampere
| > turns per meter.... but you may be using Oersteds.
| >
| > With a gap the required H field to drive the core to any given flux
| > density is increased.
| >
| > However, in a transformer with a gap the magnetising inductance is
| > reduced so ripple current increases in proportion.
| >
| > Although the core can support an higher H field that increase is
negated
| > by the higher ripple current in the windings. The equation for
| > calculating required turns for a given Bmax doesn't make reference
to
| > the H field.
|
| (The following is a not so well thought out thought. Criticism
| welcomed.)
|
| There may be a small beneficial effect of adding a bit of air gap to a
| forward converter core that makes the extra magnetizing current
| worthwhile. It has to do with the shape of the BH loop.
|
| To exaggerate the effect, assume a perfectly closed core that also has
| a perfectly square BH loop. There will be an output the first time
| the core is magnetized, but the squareness of the loop means that the
| core has no ability to demagnetize itself. So the second (or third)
| cycle of the forward converter will saturate the core. Adding a thin
| gap linearizes the net BH loop, providing assistance in damaging the
| core when the applied voltage is halted. The optimum gap is a trade
| off between this benefit (and the net increase in flux swing possible
| against the increased leakage inductance between windings and the
| increased magnetization current. If the core material had a
| completely flattened BH loop (no remnance) then the optimum gap would
| be zero.
|
| --
| John Popelish

I think my brain just imploded, mind you I'm not too good at magnetics.

....Stargate..."Carter, my Head" - that was my first impression

I do like the idea of something providing assistance to damage the core.
That's wicked.

DNA

but you need a new bh curve made by the ferrite manufactor or by yourself to determince how the hysterese loop changed right?

 

[Genome]

| Genome wrote:
|
| > Perhaps you are confusing the Bfield with the Hfield.
| >
| > Bsat for ferrites, irrespective of the gap, is in the order of 300mT
to
| > 400mT, 3000G to 4000G (at 100C). If you operate the core at higher
| > frequencies you have to reduce the value, Bmax, because core losses
go
| > up.
| >
| > Plots of the hysteresis loop are of B versus H where H is NI/Le
ampere
| > turns per meter.... but you may be using Oersteds.
| >
| > With a gap the required H field to drive the core to any given flux
| > density is increased.
| >
| > However, in a transformer with a gap the magnetising inductance is
| > reduced so ripple current increases in proportion.
| >
| > Although the core can support an higher H field that increase is
negated
| > by the higher ripple current in the windings. The equation for
| > calculating required turns for a given Bmax doesn't make reference
to
| > the H field.
|
| (The following is a not so well thought out thought. Criticism
| welcomed.)
|
| There may be a small beneficial effect of adding a bit of air gap to a
| forward converter core that makes the extra magnetizing current
| worthwhile. It has to do with the shape of the BH loop.
|
| To exaggerate the effect, assume a perfectly closed core that also has
| a perfectly square BH loop. There will be an output the first time
| the core is magnetized, but the squareness of the loop means that the
| core has no ability to demagnetize itself. So the second (or third)
| cycle of the forward converter will saturate the core. Adding a thin
| gap linearizes the net BH loop, providing assistance in damaging the
| core when the applied voltage is halted. The optimum gap is a trade
| off between this benefit (and the net increase in flux swing possible
| against the increased leakage inductance between windings and the
| increased magnetization current. If the core material had a
| completely flattened BH loop (no remnance) then the optimum gap would
| be zero.
|
| --
| John Popelish

I think my brain just imploded, mind you I'm not too good at magnetics.

I do like the idea of something providing assistance to damage the core.
That's wicked.

DNA

 

[John Popelish]

| Genome wrote:
|
| > Perhaps you are confusing the Bfield with the Hfield.
| >
| > Bsat for ferrites, irrespective of the gap, is in the order of 300mT
to
| > 400mT, 3000G to 4000G (at 100C). If you operate the core at higher
| > frequencies you have to reduce the value, Bmax, because core losses
go
| > up.
| >
| > Plots of the hysteresis loop are of B versus H where H is NI/Le
ampere
| > turns per meter.... but you may be using Oersteds.
| >
| > With a gap the required H field to drive the core to any given flux
| > density is increased.
| >
| > However, in a transformer with a gap the magnetising inductance is
| > reduced so ripple current increases in proportion.
| >
| > Although the core can support an higher H field that increase is
negated
| > by the higher ripple current in the windings. The equation for
| > calculating required turns for a given Bmax doesn't make reference
to
| > the H field.
|
| (The following is a not so well thought out thought. Criticism
| welcomed.)
|
| There may be a small beneficial effect of adding a bit of air gap to a
| forward converter core that makes the extra magnetizing current
| worthwhile. It has to do with the shape of the BH loop.
|
| To exaggerate the effect, assume a perfectly closed core that also has
| a perfectly square BH loop. There will be an output the first time
| the core is magnetized, but the squareness of the loop means that the
| core has no ability to demagnetize itself. So the second (or third)
| cycle of the forward converter will saturate the core. Adding a thin
| gap linearizes the net BH loop, providing assistance in damaging the
| core when the applied voltage is halted. The optimum gap is a trade
| off between this benefit (and the net increase in flux swing possible
| against the increased leakage inductance between windings and the
| increased magnetization current. If the core material had a
| completely flattened BH loop (no remnance) then the optimum gap would
| be zero.
|
| --
| John Popelish

I think my brain just imploded, mind you I'm not too good at magnetics.

I do like the idea of something providing assistance to damage the core.
That's wicked.

DNA

Damn spell checker! That should have read, "demaging the core" as in
demagnetizing the core".

 

[John Popelish]

but you need a new bh curve made by the ferrite manufactor or by yourself to determince how the hysterese loop changed right?

Air is pretty linear. I think it is possible to infer the air gap
plus core BH curve from the material BH curve.

 

[Tim Wescott]

Damn spell checker! That should have read, "demaging the core" as in
demagnetizing the core".

If spell checkers do nothing else for me, they do encourage me not to
use abbreviations any more.

 

[Tim Wescott]

| hi,
| is it true that theres a posibble higher Bmax value (not Bsat) if the
core is gapped resultet by a lower remanence?
|
| regards
| julius

No.

There are two currents that flow in a forward converter transfomer
primary. One is the reflected output current, one is the current in the
magnetising inductance.

If you gap the core then magnetising inductance drops and the ripple
current due to it increases. The peak and peak to peak flux remain the
same.

Note that the equation for calculating turns is

N = Vin.Ton/B.Ae

Makes no reference to the magnetic path length, Le, or associated gap,
Lg. Flux is independant of these terms.

Unless you are trying to develop power, say for auxilliary supplies,
during flyback on transformer reset there are few reasons to gap the
transformer.

DNA

But what about power inductors wound on a bobbin core. Isn't that more
or less a pot core with a whopping big gap?

 

[John Popelish]

But what about power inductors wound on a bobbin core. Isn't that more
or less a pot core with a whopping big gap?

More gap than core.

 

[John Popelish]

John Popelish wrote:
If spell checkers do nothing else for me, they do encourage me not to
use abbreviations any more.

A lesson I need to learn.

 

[Tim Wescott]

| But what about power inductors wound on a bobbin core. Isn't that
| more or less a pot core with a whopping big gap?
|
| --
|
| Tim Wescott
| Wescott Design Services
| http://www.wescottdesign.com

How is that relevant to the latest telephone bill?

DNA

Well, you were discounting the value of gapping a core, yet the
bobbin-core inductors seem to give most of the advantages of a gapped
core -- much higher resistance to core saturation, more current per
volume, better ability to talk to every other conductive loop on your
board, etc.

 

[John Popelish]

|
| Damn spell checker! That should have read, "demaging the core" as in
| demagnetizing the core".

That would be demagging }~~~

Thank you.

Sorry, like I said.... I'm not too good at magnetics. I can't see the
original concept.

I still love the original whoopsy

Chisel

and hammer

DNA

I gap to a magnetic fields something like a resistance is to a
voltage.
A resistor develops a voltage drop that is proportional to the current
that bucks the applied voltage. An air gap develops an H field that
is proportional to the flux density that bucks the H field creating
the flux. So whether the flux is caused by amp turns, or by remnance
in the core, the air gap produces an H field that bucks the flux
toward zero. In theory, no matter how big the gap, it cannot buck a
remnance produced field all the way to zero, but can push the total
remnant flux arbitrarily close to zero. If the core material has a
small remnance, a tiny air gap can significantly lower the remnant
flux in a self resetting process like a forward converter, allowing a
larger swing during the next magnetization. If you need a large air
gap to make it work, you are using the wrong core material, to begin
with. Square loop magnetic amplifier material just sucks (energy) in
a forward converter.

 

[Genome]

|
| Damn spell checker! That should have read, "demaging the core" as in
| demagnetizing the core".
| --
| John Popelish

That would be demagging }~~~

Sorry, like I said.... I'm not too good at magnetics. I can't see the
original concept.

I still love the original whoopsy

Chisel

and hammer

DNA

 

[Genome]

| If spell checkers do nothing else for me, they do encourage me not to
| use abbreviations any more.
|
| --
|
| Tim Wescott
| Wescott Design Services
| http://www.wescottdesign.com

Always check your spell checker.

DNA

 

[Julius Junghans]

Air is pretty linear. I think it is possible to infer the air gap
plus core BH curve from the material BH curve.

good idea, btw found this gap "tweak" in another smps book by billings, keith. its from 1989 but wins in the first view over the pressman one from 1997.

 

[Genome]

| But what about power inductors wound on a bobbin core. Isn't that
| more or less a pot core with a whopping big gap?
|
| --
|
| Tim Wescott
| Wescott Design Services
| http://www.wescottdesign.com

How is that relevant to the latest telephone bill?

DNA