Snubber designing ckt for push pull

Hello i need to design a snubber circuit for push-pull dc -dc convertor input 385v output 135v i have calculated the leakage inductance and the parsatic capacitence i dunt know what will be the next step in the design

Are you trying to just snub voltage spikes on your drains? If that's the case, then you want either a simple RC snubber or a voltage clamping RCD snubber. RC snubbers generally damp the waveforms better but generate more losses while RCD clamps are less effective but more efficient (and somewhat easier to design).

For a RC snubber, you want the C to be about three times the parasitic Cs and R to be equal to sqrt(Ls/Cs).

why three times can we get a more logical solution

I'm not sure about 'logical' but to damp an LC resonance you would need a resistor across the circuit equal to the resonant impedance. The series capacitor is in place to block lower frequencies so the resistor only dissipates energy associated with the resonance.

OK, try it this way. This is me guessing so it will probably be wrong.

For critical damping you would expect the ringing to be damped in one half cycle [could be a full cycle]. Either way assume that the capacitor will charge to the applied voltage in that time. Assume Ls and Cs are your respective leakage and parasitic capacitances. Then the resonant frequency is,

Fr = 1/2.pi.SQRT[Ls.Cs]

A half cycle would be,

Thlf = pi.SQRT[Ls.Cs]

As Mtwieg says the resonant impedance is,

Xr = SQRT[Ls/Cs]

Time constant, to charge to 0.63 of the applied voltage, for an RC network is,

Tau = R.C

Substitute Xr,

Tau = Xr.C
Tau = SQRT[Ls/Cs].C

Tau will be your half cycle at resonance so,

pi.SQRT[Ls.Cs] = SQRT[Ls/Cs].C

Rearrange,

C = pi.SQRT[Ls.Cs]/SQRT[Ls/Cs]

Square both sides,

C^2 = pi^2.Ls.Cs.Cs/Ls
C^2 = pi^2.Cs^2

Take square roots of both sides,

C = pi.Cs

Which is, more or less...

C = 3.Cs

So the required series capacitor is three times the stray capacitance. Gosh!!! I'm surprised that worked.

Genome.

Yeah it just turns out that a factor of 3 gives a good damping factor. If you want to see all the gory detail, just search google. There are lots of white papers on this stuff.

[QUOTE=Genomerics; thanx man if u re intrested in smps design u can give me ur skype id or msn id if u can help me with my design

also, with damping resonances...i presume you are worried for emc(?)

You could just put a clamp there, so as to make sure that the drain voltage does not exceed its rated maximum.

....do you really want to damp resonances if you can get away with leaving them in there?

there will be a reflected voltage from sec to pri.....so use this to find max drain voltage, and just set your clamp above this voltage level.

will suppressors work my DC-DC input is 385volts and output is 145volts what the spike is of 800 volts approx

You need to know how much voltage is reflected to one of the primary coils when the "opposite" secondary coil is conducting.
...its the usual transformer equations...Ns/Np = Vs/Vp.

You want to catch the leakage spike, but you definetely dont want the de/magnetizing current going into your clamp....that will be too lossy....so get the clamp voltage high enough so you snip the spike but dont stop the de/magnetizing current from returning to your primary capacitor.

hasanferoz said:

Hello i need to design a snubber circuit for push-pull dc -dc convertor input 385v output 135v i have calculated the leakage inductance and the parsatic capacitence i dunt know what will be the next step in the design

Click to expand...

Just in case someone is following this :
How did you calculate the leakage inductance and parasitic capacitance ?
I think leakage inductance can be measured by shorting the secondary coils,but how can it be calculated ?