How does snubber work?

[AGCB]

I understand how a flyback diode placed across a DC inductor protects switching circuit from back EMF.
But how does a snubber (capacitor and resistor placed across AC load) work to protect switching circuit when turning off?
Inquiring minds want to know!
Aaron

 

[MikeMl]

Assuming the AC circuit has an inductive load, the RC snubber across the switch performs two jobs.

When the switch opens, the inductor current flows into the resistor in series with capacitor, creating a damped oscillation that dies out due to the dissipation of energy in the resistor.

When the switch closes, the capacitor might have a charge on it, and the resistor limits the initial capacitor discharge current.

Look at this. Switch is open for 20ms, then it closes, and then opens again at 40ms.
The capacitor C1 charges to 100V, and it discharges through R1 (10A pulse) as the switch closes at 20ms. The resistor is chosen to keep that pulse within limits.

Just before 40ms, the inductor current is 3A. As the switch opens, the snubber limits the first voltage peak to about 900V. It would have spiked high enough to create an arc across the opening switch otherwise (>20kV). The duration of the ringing is determined by L1 R1 C1.

 

[dknguyen]

An RC snubber is ideally chosen so that it critically damps the parasitic inductances and capacitances in the circuit to minimize ringing and overshoot. That's how it's supposed to protect the circuit. In this way it reduces voltage spike to a limited degree but it's not in the same way that a diode just outright clamps the voltage to a fixed level.

The other form of protection that an RC snubber provides that a diode does not is that it slows down rise and fall time of edges so the high dV/dT doesn't mess with the switch. I've had circuits where clamp diodes did not help at all because it wasn't the overshoot that was the problem: it was the high dV/dT and that was causing the switch to behave unreliably when turning off. It needed an RC snubber to fix the problem and with the RC snubber it ran fine without the diodes.

In addition, RC snubbers are much faster than diodes since they do not need to turn on like a diode. They are always "on" so they are much much faster. They can be used to "buy time" until your clamp diodes turn on.

The simplest way I describe an RC snubber is like this: just think of a RC snubber as a plain capacitor that acts as a short-circuit for very high frequencies so it slows down the edges. It gives the current from the inductance a momentary path to flow when the switch turns off which means the voltage spike is not as big (since the voltage spike exists because it's trying to force the current to keep flowing when the switch opens). That's basically what it's doing.

The resistor is added later to provide dampening characteristics and to give that energy absorbed by the capacitor a way to be dissipated. It's sort of like a high pass filter. Another way to think of it (which is not quite as accurate) is that it's a bit like a clamp diode, except a clamp diode discriminates based on voltage and clamp the voltage, whereas an RC snubber discriminates based on frequency/rate of change/time and "clamps" the rate of change (but not really since it just slows it down rather than limits the speed to a set level).

 

[AGCB]

Thanks guys for your time!

 

[MaxHeadRoom78]

These may help
Max.

 

[simonbramble]

If you understand V/L = di/dt.... for a constant value of L (leakage inductance), the faster you switch off the switch (dt) and/or the bigger the current (di), the bigger the back EMF will be (V) as the di/dt will be higher. So if you have a slow switch (like a bipolar transistor) you will get a smaller back EMF compared to a fast switch like a FET (everything else being equal). A snubber is designed to provide a path for the inductance current to flow one the switch has opened, thus slowing down the change in current with time (di/dt), thus reducing the back EMF. This is why you get snubbers from the collector/drain of the switch to ground as well as from collector/drain to input supply. They are not necessarily clamping the back emf to the supply voltage, more like diverting the current and slowing down the di/dt... Just another way of looking at its operation