Author Topic: RCD Snubber design  (Read 1744 times)

0 Members and 1 Guest are viewing this topic.

Offline XaviPachecoTopic starter

  • Regular Contributor
  • *
  • Posts: 243
  • Country: do
RCD Snubber design
« on: June 17, 2018, 10:01:01 pm »
Hello,

I want to design a RCD snubber network like the one attached. I have read some guidelines and books regarding to it. But they all use a VDS/IDS  vs time waveform to determine the turn off time. That graph is not given in datasheets. What is given in datasheets are the turn-off delay and fall times for specifics values of ID, VDD, RG and VDS. So, does the mosfet have to get through empiricals tests before designing a snubber?

This is the mosfet that I want to snubber: https://www.infineon.com/dgdl/Infineon-IRF250P224-DS-v01_00-EN.pdf?fileId=5546d4625a888733015a8bb640137c6e

« Last Edit: June 17, 2018, 10:04:43 pm by XaviPacheco »
 

Offline CopperCone

  • Super Contributor
  • ***
  • Posts: 1415
  • Country: us
  • *knock knock*
Re: RCD Snubber design
« Reply #1 on: June 18, 2018, 02:07:29 am »
has anyone seen an active measuring snubber ?
 

Offline Neilm

  • Super Contributor
  • ***
  • Posts: 1545
  • Country: gb
Re: RCD Snubber design
« Reply #2 on: June 18, 2018, 07:01:23 pm »
I try to simulate it before committing to the snubber. I made a high isolation DC-DC flyback supply a few years ago that had to fit in a really small space. I just measured the leakage inductance in the transformer then ran the simulation. Designed a lossless snubber and the simulation agreed with real life to within one volt (peak was 90V)
Two things are infinite: the universe and human stupidity; and I'm not sure about the the universe. - Albert Einstein
Tesla referral code https://ts.la/neil53539
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 8106
  • Country: fi
Re: RCD Snubber design
« Reply #3 on: June 19, 2018, 07:24:01 pm »
Your missing parameter is parasitic inductance, and since it comes from your PCB layout, it cannot be in the datasheet.

My suggestion is to try your best in minimizing the loop area, and use proper low-ESL (small ceramic) DC link capacitors very close. You need to think about this loop: input capacitor + -> high side switch -> low side switch -> input capacitor -, and mininize that loop.

Modern SMD parts have made this much easier.

Now, you can add footprints for an RC snubber across both FETs for prototyping, but chances are you won't need them.

Run first at lower voltage and look at the scope trace. If your layout sucks, you may easily have 2x overshoot and ringing in the switch node voltage -  this is why you run at lower voltage first so that you won't blow the FETs on excess Vds. OTOH, if the overshoot is below about 1.5x the DC link voltage, your FETs Vds rating is fine with it, and there is no requirement for EMC, then snubbers would only waste some energy.

In modern motor controllers or switch mode designs, most of the time there are no snubbers. Do the layout right.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: RCD Snubber design
« Reply #4 on: June 19, 2018, 11:27:08 pm »
Yeh... what's on the other side of that connection dot?

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: RCD Snubber design
« Reply #5 on: June 19, 2018, 11:34:19 pm »
My suggestion is to try your best in minimizing the loop area, and use proper low-ESL (small ceramic) DC link capacitors very close. You need to think about this loop: input capacitor + -> high side switch -> low side switch -> input capacitor -, and mininize that loop.

This may be possible but it isn't a general solution.

Example, a 5V 50A converter built with DFNs still has too much stray inductance for this to be a viable approach.

For a general solution, the key is realizing, the switching loop is an impedance like any other: a transmission line that must be terminated.  Maybe you aren't going for termination as such -- you generally want reflections in a switching supply, as long as they're in-phase so the energy recirculates back into the supply -- but whatever the case, this observation provides much needed guidance on the situation. :)

A snubber is nice, but also not always possible, what with the stray inductance of two components in series, even.  (On that note, I wish they made SiC schottky in SMT packages... at SMT prices, mind.)  The only thing you can do at that point is make the channel smaller (so that the load impedance becomes higher than the switching loop impedance, thus reducing voltage overshoot), and use multiple acting in parallel.  Multiphase is also a win for ripple current.  This does of course carry the unfortunate side effect of needing more drive and control circuitry.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline XaviPachecoTopic starter

  • Regular Contributor
  • *
  • Posts: 243
  • Country: do
Re: RCD Snubber design
« Reply #6 on: June 19, 2018, 11:50:31 pm »
The application is based on a PMDC motor controller like the one attached for treadmills. I'm trying to figure out if the snubber is strictly necessary in this application.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: RCD Snubber design
« Reply #7 on: June 20, 2018, 11:55:55 am »
Do you have a layout to show?

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf