SEPIC converter: RC snubber sizing?

[treez]

Hello,
The attached are schematics of two like-for-like SEPIC converters, however, one has an RC snubber across the sepic capacitor and the other doesn't.
The one without the snubber suffers terrible ringing of the sepic inductors and the sepic capacitor.

The (unwanted) L,L,C ringing frequency is 6850Hz.

The snubber 3dB frequency is 1590Hz.

The snubber obviously dissipates virtually nothing when the sepic is in steady state.

Why is this snubber frequency of 1590Hz so effective in totally snubbing out the unwanted resonant ringing of the SEPIC?

The SEPIC spec's are as follows:
F(sw) = 104KHz
Vin = 5V
Vout = 15V
Pout = 7.5W
Continuous Current mode (with slope compensation)

 

[FvM]

A usual rule-of-thumb sets the snubber R equal to the characteristic impedance of the LC resonator which should result in lowest Q. The rule would give a slightly lower R value, e.g. 2 or 3 ohm. But your value is still in a effective range.

 

[treez]

Thanks FvM...do you find it strange that in steady state , this snubber dissipates virtually nothing?.....its unbelievable that this awful L,L,C ring that is the scourge of the uncoupled inductor sepic is so easily gotten rid of...it seems too easy to be true?

 

[schmitt trigger]

How do you know that the snubber is dissipating nothing? Is that from the simulation, or have you actually built the circuit?

 

[mtwieg]

Thanks FvM...do you find it strange that in steady state , this snubber dissipates virtually nothing?.....its unbelievable that this awful L,L,C ring that is the scourge of the uncoupled inductor sepic is so easily gotten rid of...it seems too easy to be true?

The snubber doesn't damp the ringing with the leakage inductance, but rather the resonance formed with the magnetizing inductance. The reason that it doesn't dissipate much power is that it does not significantly affect the Q of the capacitor at the switching frequency (since that damped resonance is far below the switching frequency). In the case of a resonance that forms near or above the switching frequency (like that caused by using very small leakage inductance) you wouldn't be able to damp that without significant losses.

 

[treez]

How do you know that the snubber is dissipating nothing? Is that from the simulation, or have you actually built the circuit?

...simulation...my actual circuit that I have built uses a coupled inductor.

In the case of a resonance that forms near or above the switching frequency (like that caused by using very small leakage inductance) you wouldn't be able to damp that without significant losses.

...this then seems like a big disadvantage of the coupled inductor sepic converter?

-in smps work, as you know, most snubbers damp ringing which is well above the switching frequency...in such cases, if the ringing is miles above the switching frequency, then is it possible to snub it without significant losses?

Also, supposing I am doing a coupled inductor sepic, and the leakage is 460nH, and the sepic cap is 22uF, then the LC resonance is at 50KHz......what if my switching frequency is also at 50KHz?....does that mean it will be very difficult to effectively snub this resonance without incurring huge losses?