Transientreduction

[Wedwin]

I have by now read a lot about snubbers.

I think snubbers are a remedy for a symptom, not a solution. Well, induktive loads are ... inductive.

The spikes occur due to the transistors switching off very fast.

Would it be feasable to turn off the transistor slowly enough to reduce the spikes, yet fast enough not to release too much power inside the transistor?

Rather, is this simple to do or is it just as hard as dimensioning snubbers?

Sincerely, Wed

 

[Skogsgurra]

It is possible to turn the transistors off slowly, yes. But it is not something one wants to do since the power dissipated during the switch-off will be high. That is, in fact, why faster and faster switches are used in many power applications. Fast switching results in better efficiency.

There is another thing that should be said about snubbers; they are not always used only to absorb spikes when transistors are switched off. There is also the "carrier charge effect" in thyristors that has to be taken care of - and also the fact that external spikes from the grid can disturb a mains connected thyristor rectifier and those spikes are absorbed by the snubbers as well.

There are frequency inverters where the switching has been slowed down to avoid EMI (a Norwegian system) but they need to be switched at lower than normal frequencies to avoid overheating.

 

[Wedwin]

Thanks, so snubbers it is I guess.

Has anyone tried to store the spike to reuse in in the pulse to come?

I am not trying to make a 100% efficient PWM unit here. But I don't really like to just burn the energy off for nothing.

Wed