Reducing ground bounce in dc/dc convertors

[marce]

As we all know these nasty little SMPS are dirty beasts, polluting the audio world with their noise!!!
Well Ok a slight exaggeration, though if the layout is not done correctly, and the end result EMC tested you can have problems. I know it's hard for home designs to be EMC tested and at some point I think a thread on home testing should be attempted as even if you follow all the guidelines you can still get caught out...
Anyway an interesting article by Jeff Barrow on reducing ground bounce in dc/dc convertors, with some excellent pictures that illustrate the points much better than dry text

http://www.analog.com/library/analog...und_bounce.pdf

[marce]

Woops, could a moderator change the spelling of the title for me please...

[Osvaldo de Banfield]

Hi Marce. I actually have and DIY audio equipment designed and build of my own. It has a +- 14V and +12V SMPS Flay Back, switching at 50KHz, designed around UC3842, STP7N60, TL431, BYW98/200 of about 80W with an EE core recovered from PC monitor. It also has 3 * TDA2006 (BASS OUT), TDA2003 (TREBLE OUT) driving 3 Karlson Speaker System (build of my own too) 2: 15" + 4" cone tweeter, and 1 12" +4", and lots of TL084, TL082, etc. including TDA1220B and TDA7211 (AM/FM FI and FM front end). I never has problems by irradiating signals from SMPS, this is working since about 11 years.
Being careful with power lines and cables, there is no trouble with SMPS´s. Neither in AM ferrite reception.

[marce]

Osvaldo, the comments at the top were satirical in nature...
I work as a proffesional PCB designer, and do a lot of SMPS layouts for Mil gear and this information, along with my other thread on EMC switcher designis to put what I consider useful information out there for DIYers.
This information is specificly aimed at best layout practice, and reducing the switching loops, and I liked it cos it has lots of pretty pictures that illustrate the points much better than reading some verbose application note.
Unlike DIYers I have the luxury of my layouts being tested to the nth degree, and getting critiqued by the controller chip manufacturers guys...
And layout for both SMPS's and Class D is critical, as it is for all layouts, analogue or digital

[Osvaldo de Banfield]

What happens is that there is a popular missconcept about that ALL SMPS are in nature interference generators. And I could test by my own that it is false. Which I couldn´t never understand is why commercial audio equipment rarely uses this stuff. A good designed Fly Back SMPS with its inherent wide bandwidth and dynamic range aren´t expensive nor difficult to design and could offer cost, space, and performance much better that iron transformers. I´m also Ham Radio, and there are about 3 meters between the SMPS and radio HF equipment and no interference is noticeable as bigger as the PC monitor or TV power supply.

Do you know why?

[marce]

We use SMPS to power our radios and have no problem. It just seems SMPS's get a bad press, probably from their very early days. What I have found is you get more noise generated from numerous digital devices switching than you do from SMPS's in digital systems. In fact we have had problems with interference getting through a linear supply and moduclating the audio, move to an SMPS and all went quiet.
That said I have sen a container full of SMPS's bought at a good price from overseas go in the skip because they failed EMC tests, the initial units passed with flying colours!!!

[Osvaldo de Banfield]

OK, interesting your comments.