What is the relationship between leakage inductance and core gap size?

[treez]

I think it is better to equally distribute the total gap over the three legs

When I worked in the PSU dept of a multinational TV company, they had a 40W Flyback in the TeleVision’s (vin=390v,vout=25v,fsw=80khz or so, pout=40w, BCM)
We noticed that it had a very very small RCD clamp, which indeed was dissipating very little. We “unwound” the flyback transformer to see what special winding technique had been used…and noticed no gap at all. It must have been some special ferrite mix that “integrates” the gap in the body of the ferrite.
(As an aside, I once worked in a huge company where the engineers developing an offline, 25W, BCM flyback thought that quasi-resonant flybacks re-cycle the leakage energy...they hadn't used any RCD clamp at all..they had to go back to the drawing board)

- - - Updated - - -

Warpspeed I like your test rig of post#13, I will likely use that as the number of places ive worked where there is either

1...no LCR meter
1A...Have An LCR meter but its lost and they wont buy another till someone owns up to nicking it.
2...They have an LCR meter but its constantly hogged by some individual
3...The LCR meter doesn't measure leakage at the actual frequency that you want it to.

- - - Updated - - -

I have often wondered about the wisdom of gapping the centre limb of a transformer, and then winding right over the top of the gap.

I concur with this, and I don't know if it helps, but I have a personal theory that if you do have to wind over a gap, then it gives lower leakage if you can use either thick ECW, or multiple parallel ECW strands so that at least some of the conductor is not "over the gap".

Thanks FvM, great mag sim results, shows that the gapped ones have less Llk, but the Llk that exists is a greater percentage of the primary inductance.
The Air core one is interesting. Mind you, your "distributed gap having more leakage than the discrete gap" result throws my theory about the TV company flyback transformer and how that got so little leakage inductance.
These are all fantastic replies in this thread. Thankyou all.

 

[Warpspeed]

It must have been some special ferrite mix that “integrates” the gap in the body of the ferrite.

There are a fairly wide range of manganese zinc ferrite’s available, but most of what we use in switching supplies are the very common "power ferrite" grades.
But there are also a wide range of very low permeability RF grades of nickel zinc ferrite, particularly in toroids.

The ferrite mix would already exist, and the ferrite manufacturers would have suitable dies for the core shape. With a large enough order it might be possible to get just about any combination.