Peak flux density in flyback transformer core?

[treez]

Hello,
We are doing an offline flyback which is 60W average output power.
However, it has a "peaky" load and the output power can be 120W for 100ms intervals.
During these intervals, our EFD30 flyback transformer design would have a flux density of 350mT.
Is this too high for gapped (0.7mm) N87 ferrite material?...the datasheet doesn't say.

Datasheet of the 0.7mm gapped EFD30 ferrite core:
https://www.farnell.com/datasheets/80944.pdf

 

[FvM]

The N87 material data can be found in the first part of the ferrite data book. You can also use the Epcos Ferrite calculation tool to estimate core losses.

350 mT is below saturation but involves huge core losses. Question is about switching frequency and duty cycle of peak load.

If it's more than a few percent, the average magnetization must be reduced to compensate for the extra losses.

 

[Anna Conda]

350mT is acceptable for short overloads....

 

[treez]

I have downloaded the the EPCOS "ferrite magnetic design tool", and this tool does not appear to be at all useful in analysing flyback designs.
I have a CCM flyback design using an ETD44 (1mm gap) core (NP/NS=2), LP=177u, Vin=70VDC, vout=24V, IOUT = 5A (peak for 100ms intervals)
and the tool doesnt seem to help much.
It tells me what the RAC/RDC value is, but i cannot understand how it can possibly have worked this out when it did
not ask me for the peak and pedestal currents of the primary trapezoidal waveform?

Also, it asks me for a "B(mT)" value and then calcualtes what it says is a core loss figure in kW/m^3.
This is surely incorrect, since again the peak and pedestal flux values would have to be known, as well as the operating frequency,
in order to calculate the core loss in the flyback ferrite.

I read the manual aswell but it offers no enlightenment......i am wondering if this tool is simply for
sinusoidal waveforms only?

EPCOS ferrite magnetic design tool
https://en.tdk.eu/tdk-en/180490/design-support/design-tools/ferrites/ferrite-magnetic-design-tool

The databooks are equally unhelpful in working out core losses in flyback cores....they simply seem to
only address the situation of sinusoidal waveforms, which is totally unhelpful when you have unipolar
trapezoidal waveforms as in the flyback SMPS.

 

[Anna Conda]

triangular flux is not very different from sinusiodal (only 3% different), so if you know the p-p flux variation you can use half of this to estimate core losses. Of course the step applied volts in a switchmode mean that some materials suffer from induced dv/dt eddy current losses more than others (those with lower resistance) which is what sets apart the good core materials. As a rule of thumb, for 100kHz keep Bpk to 80mT (160mT pp) max, for reasonably good core materials...

 

[treez]

thanks, but we can't keep Bpk to 80mT.....the core would need to be too big. We let Bpk go up to 300mT in our flyback....oh, sorry, did you mean delta B = 160mT max?

 

[Anna Conda]

delta B = 160mT peak-peak ( Bpp) for steady state operation, Bpk can be 300mT or a bit higher but the flux swing must be constrained for reasonable core losses....