transformer inductance

[Abdelwakil]

Hi

I tried to simulate the smps fly back conv. in "Psim" app but i didn't understand what's the difference between Lm (magnetizing) and Lp(leakage) , and for that , i got a very inefficient results .

 

[Julie.]

In any transformer, the applied current generates a magnetic flux, some of this is used to magnitise the core, (search for magnetic saturation for an explanation) this is generally understood using a b-h diagram, and means that some of the input is lost in magnitising the iron core - this is Lm

In addition, some of the flux generated, actually misses the core, - it finds its way from the north pole of the winding to the south pole of the winding without going through the core, or through the secondary winding - this magnetic loss is Lp.

In addition, not all the flux that goes through the core actually goes through the secondary winding, - this is the Ls

- If you reversed the primary and secondary windings, then you would be able to measure the Ls just as you might measure Lp when not reversed etc.

So a core made up of poor magnetic material would increase Lm, and a core which has a poor design such that the magnetic field from the primary doesn't all go through the core would increase Lp, similarly if the design is such that the secondary doesn't link well with the core - this would increase Ls

 

[Abdelwakil]

In any transformer, the applied current generates a magnetic flux, some of this is used to magnitise the core, (search for magnetic saturation for an explanation) this is generally understood using a b-h diagram, and means that some of the input is lost in magnitising the iron core - this is Lm

In addition, some of the flux generated, actually misses the core, - it finds its way from the north pole of the winding to the south pole of the winding without going through the core, or through the secondary winding - this magnetic loss is Lp.

In addition, not all the flux that goes through the core actually goes through the secondary winding, - this is the Ls

- If you reversed the primary and secondary windings, then you would be able to measure the Ls just as you might measure Lp when not reversed etc.

So a core made up of poor magnetic material would increase Lm, and a core which has a poor design such that the magnetic field from the primary doesn't all go through the core would increase Lp, similarly if the design is such that the secondary doesn't link well with the core - this would increase Ls

I remarked that the decrising of the Lp is the most necessar thing to improve the effeciency , and Lm just make the response more or less quick .
Please am interested about the smps one , can you tell me why 90% of psus has very high Lp (more than 1mH) and they are efficient at the same time .

 

[Julie.]

I remarked that the decrising of the Lp is the most necessar thing to improve the effeciency , and Lm just make the response more or less quick .
Please am interested about the smps one , can you tell me why 90% of psus has very high Lp (more than 1mH) and they are efficient at the same time .

Not sure, when you refer to Lp is this actually the leakage inductance, or the primary inductance?

In psim it looks for the leakage inductance, however 1mH looks more like the total primary inductance - which would need to be fairly high or else the mosfet (or whatever is driving it) would need to drive loads of current as it won't be damped by a low inductance, a higher level of inductance (not leakage) would mean that the current would increase over a longer period, so say the mosfet applied 100v, - the current would climb up following a exponential curve, reaching peak at say 0.1 secs, after this the current would be constant - and no transformer action; now half the inductance and it would grow quicker - say taking 0.05 secs to reach maximum current - and then constant again, so you would lose effeciency as part of the time you wouldn't be able to transfer energy across the transformer, your choice would be more inductance, or higher switching frequency.

I would guess, given the optimal frequency / inductance balance that these generate the figures you find.