switching noise/spikes on buck converter / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

switching noise/spikes on buck converter

1 2 3 next

Move Thread

LAN_403

IamSmooth

Tue Nov 16 2010, 08:56PM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

So, what is the best means for subduing switching spikes? Higher filter capacitance? Using a particular, low ESR capacitor, or adding input/output series inductance? Someone must have dealt with this issue.

Thanks.

2Spoons

Tue Nov 16 2010, 10:19PM

Registered Member #2939 Joined: Fri Jun 25 2010, 04:25AM
Location:
Posts: 615

Are these spikes on the output? L-C pi filters are good.

Daedronus

Tue Nov 16 2010, 10:24PM

Registered Member #2329 Joined: Tue Sept 01 2009, 08:25AM
Location:
Posts: 370

I think I said this before...but a common mode choke will do the trick :)
Or at least severely attenuate them.

GeordieBoy

Tue Nov 16 2010, 10:29PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Switching spikes where? Within the converter? on the DC output? conducted back to the input?

In general for reducing switching spikes and ringing in power electronics systems...

In this order:

1. Use fast switching diodes, schottky or FRED
2. Minimise stray inductance in switching and clamping loops
3. Minimise leakage inductance in transformers
4. Minimise stray capacitance across chokes
5. Slow down switching transitions to minimise Ldi/dt spikes.
6. Add dissipative snubber if all else fails.

-Richie,

IamSmooth

Wed Nov 17 2010, 12:04AM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

If my induction heater (the load) switches at 100khz, does it matter if my converter switches at a slower (36khz) speed, or should it switch higher? Does it make a big difference?

The switching spikes I was referring to were seen on the output.

Daedronus

Wed Nov 17 2010, 09:46AM

Registered Member #2329 Joined: Tue Sept 01 2009, 08:25AM
Location:
Posts: 370

A induction heater should not care at all about some switching noise.

Also, if you feed DC to the induction heater switcher, the switching frequency of the buck converter feeding the induction heater, will not matter, DC is DC :)

On a power supply output you have two different things, voltage ripple and switching noise.
The ripple is more or less how much the voltage goes up and down between two cycles. This can be improved by having more bulk capacitance and a bigger inductor on the output.
The noise, the little high frequency spikes will not go away with a bigger capacitor and/or inductor. However a common mode choke will make them cancel themselves out.

I'm working on a half bridge smps that right now works nicely in open loop mode and I will try to post wave traces before and after the output choke as well as a picture of the choke itself which is a simple ferrite toroid with 2 windings.

IamSmooth

Thu Nov 18 2010, 04:13AM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

With regards to the gate drive, is this subcircuit acceptable?

Daedronus

Thu Nov 18 2010, 09:24AM

Registered Member #2329 Joined: Tue Sept 01 2009, 08:25AM
Location:
Posts: 370

The zenners are not really needed, and they should be rated at about 16v for a 15v power supply.

They are needed when you use a gate drive transform, because the output of the transformer is not as well controlled as in the case of a driver IC.

I would also add pull down resistors (10Kohm should do) on the pwm input and on the gate so they are not left floating by accident.

IamSmooth

Fri Nov 19 2010, 08:40PM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

I'm going to add a snubber.

I read this link

which seemed to provide some good information. The output capacitance of one of the igbts that I have is 370pf, so I was going to use something close to 1000pf.

Would this surface mount be acceptable?

My output voltage can be near 750v @ 50A. Using this with a f = 35khz I get a power dissipation for the resistor as 15R @ 20W. This would be a big resistor. Any suggestions for what to do here?

Daedronus

Fri Nov 19 2010, 09:44PM

Registered Member #2329 Joined: Tue Sept 01 2009, 08:25AM
Location:
Posts: 370

750v at 50a? are you sure? that's 37.5KW
What is your input voltage?

Also, buck converters don't normally use snubbers. Snubberrs are normally used to control the voltage rise on flybacks and ringing on half or full bridge smps.

A verry good source of information for building smps is this book:
Switching Power Supply Design by Abraham Pressman.

It does cover buck converters and more, I strongly recommend it.

Also, try to post waveforms for the noise/spikes that you want to filter out.

1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob