• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Snubber design calculations

Status
Not open for further replies.
Hello everyone. I'm designing a three phase inverter using 3 IGBT modules. I have used the SPWM technique to create the sinewave. The switching frequency is 6 KHz. Everything is working fine just the voltage spikes are being created across CE when the IGBTs are turned off. I've attached the images for the high side as well as low side IGBTs in which Vce is measured. I've read many articles online which explain the types and application etc. but not the actual calculations. I've decided to use "Discharge supressing RCD snubber". Can anyone suggest what should be the starting point and calculations?
 

Attachments

dknguyen

Well-Known Member
Most Helpful Member
You never ran into any calculations? I find them in quite a few places:
http://www.cde.com/resources/technical-papers/design.pdf

Admittedly, they are rather sparse and a lot of it seems to be "napkin calculations" to ballpark the snubber values, after which point you proceed by trial and error while observing the switching waveform until you get what you want. Calculations aren't too specific because they would require measurements about your actual system which can be tricky to get (such as the inductances and capacitances causing the ringing). Getting things really spot on is really more for optimizing efficiency.

If you're already running it with no snubbers and things aren't exploding, then I would just go trial and error. Increasing the capacitance dampens things more at the expense of component cost, some efficiency, and more heating in the resistor. When I jump in blind, I try a 0.1uF capacitor and a 2W or 5W resistor that is 1 to 5 ohms.

I just start with 1uF because it's large-ish but not so large that it's expensive and difficult to find. The only drawbacks are unecessary cost (since you might overdamp the ringing which isn't too big a deal), potential excessive heating in the snubber resistor, some loss of efficiency (none of which are really a problem if I'm testing). And I just use a smallish resistor to not mess with the absorbing the spike too much and trial and error from there.

You can also just run the system low and use some resistors and ceramic caps you have laying around if the voltage rating allows for it. Even if they are too small to handle the power, just run the system at low power and only for a very short time and get a snapshot of some of the waveforms. Seeing the effect of two or three combination of RC values should give you a gut feel for what actually need.
 
Last edited:
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top