Selecting RC snubber parts for 400khz buck-boost inductor

[Psi]

Quick question about adding snubber RC on each side of an inductor on a 400khz buck-boost.  (Snubber is reducing ~230mhz ringing).

Excluding power dissipation and the R and C value, what are the considerations with regarding the physical part size (0603/0805 etc) and type of cap C0G/X7R or having a cap designed for RF.
Are there any specs that need to be considered for this application. Or can I throw a generic part in.

[moffy]

230MHz ringing sounds like a good use for some series ferrites.

[Psi]

The official design/app note just has a RC on each side of the inductor as satisfactory.
But i could do some tests with a ferrite

[moffy]

A suitable ferrite should provide good damping/absortion at that frequency, let us know how it goes.

[Psi]

Do you have any recommendation on how best to add it,
Currently there's a 2.2nF + 1R resistor (in series) to ground from each side of the inductor. (buck-boost)
The goal here is just to stop it causing any issues for the dcdc controller loop.   Less about in/out EMI.

[moffy]

I would assume placing it close and in series with the inductor, perhaps one on each side to mimic the symmetry.

[mtwieg]

Excluding power dissipation and the R and C value, what are the considerations with regarding the physical part size (0603/0805 etc) and type of cap C0G/X7R or having a cap designed for RF.
Are there any specs that need to be considered for this application. Or can I throw a generic part in.

The snubber circuit has to have a lower inductance than the circuit it is snubbing (not the inductance of your inductor, but the parasitic inductance in the switching components), which is why component size has to be kept small while still meeting power dissipation requirements. IME once the inductance being snubbed gets below a few nanohenries, designing effective RC snubbers becomes much more difficult. Splitting the snubber into multiple parallel networks can help somewhat.

You don't need a fancy capacitor type, class II dielectrics can work fine. Don't need super precise capacitance and some extra ESR is fine. Just keep in mind the effect of bias voltage on the effective capacitance.

I would assume placing it close and in series with the inductor, perhaps one on each side to mimic the symmetry.

Unlikely to help, as usually the inductor/choke is not participating in the high frequency ringing to begin with.

[SCSKITS]

230MHz (about 4.3nS) seems a bit high for ringing on a 400KHz converter.
Something else is going on there, maybe some parasitic inductance or capacitance in the layout.

[moffy]

230MHz (about 4.3nS) seems a bit high for ringing on a 400KHz converter.
Something else is going on there, maybe some parasitic inductance or capacitance in the layout.

It's an interesting frequency, reminiscent of the VHF oscillation of MOSFETs that don't have either a ferrite or resistor on the gate lead.

[Terry Bites]

You may find these design notes helpful.
Moffy's right, it does sound like you need a ferrite on the gate.