Coupled inductor sepic and RMS current in primary.

[treez]

Hello,
The attached LTspice simulation is of an open loop sepic with coupled inductor. (schematic in pdf also attached)
The resonance frequency between the c(sepic) and the L(leakage) is equal to the switching frequency. (104khz)

I am amazed how very slight changes to the primary and secondary inductor's esr has such a big effect on the magnitude of the RMS current seen in each of these inductors.

With 1 milliohm of esr in each inductor, the rms current in the primary inductor is just 6.12 Amps.
With 30 milliohms of esr in each inductor, the rms current in the primary inductor is 1.93 Amps.

....Why such a big difference in RMS current?..after all, 30 milliohms, even though 30 times more than 1 milliohm, its still an incredibly low amount of resistance.

The sepic spec is
vin=5v
vout=15v
f(sw)=104khz
pout=7.5w
CCM

 

[FvM]

With 1 milliohm of esr in each inductor, the rms current in the primary inductor is just 1.93 Amps.
With 30 milliohms of esr in each inductor, the rms current in the primary inductor is 6.12 Amps.

I think it's the other way around. High current with low ESR.

What happened? You design a high Q series resonator and tune it carefully to the switching frequency. You shouldn't be surprized to get high currents,

 

[treez]

yes sorry , I got it the wrong way round.....(I've corrected it now in the top post) ..but 30mR is still low resistance, you wouldn't think the extra 29mR would damp it down that much , would you?...basically, I am trying to assess what level of esr is needed in order to ensure that these high rms currents don't happen in sepics.

 

[FvM]

I think, the leakage inductance in your simulation is unrealistically low. With real SEPIC inductors, you rarely get a series resonance at the switching frequency.

 

[treez]

The MSD1583 coupled inductor boasts coupling coefficient >=0.98......so with bifilar wound coils, you could be at k=0.9974, and by L(lk) = L(1-K^2), we see leakage of 140nH , as in the schamtic of the top post.
So are you saying that k=0.9974 is not possible, even with bifilar winding?

MSD1583 coupled inductor
https://www.coilcraft.com/pdfs/msd1583.pdf