Post by avtoelek on Jan 9, 2014 20:17:10 GMT 1
I'd like to share my experience regarding switchers and audio circuitry, especially when they are soldered on the same board. It seems that careful choice of power components, especially passive ones, is very important when designing power management with audio. I had recently some problems with an audio board that had experienced unusually high values of floor noise. This came out during a revision phase and was quite surprising...well the microphone was mounted on the same board that the switchers were. After researching I found that a possible cause of the problem could be the piezo effect linked to MLCC (ceramic) capacitors used for switchers. The first thing to try was replacing all ceramic caps around switching converters with tantalum, electrolytic or poscaps/polymer types. This helped only a little, so I decided to look for other possible causes of noise. Incidentally I then found that if I replaced the power inductors with others, in my case low acoustic inductors from Wurth Electronic the noise came down again. The results were quite good but not excellent though. The only possible cause of noise were all those little decoupling MLCCs across micro controllers, codecs and so on...a bunch of capacitors to remove and impossible to replace with film, electrolytics or others because of size issue. The last thing was to remove the microphone and place it on a separate board connected through some sort of "soft" connection to the main board...why soft? Because it seems the root of the problem was VIBRATIONS caused by piezo effect by MLCCs and magnetic flux inductors!Well I came across low acoutic noise MLCCs made by Murata by unfortunately didn't have the chance to give a try.
Did anyone try the above ceramic capacitors and could report the results?
In the end it seems it is not sufficient to replace ceramics caps and use better inductors, but one should mechanically separate the microphone from the rest of the digital/power circuitry to have a very (near zero) noise floor...talking about -65dBFS. The connection should also be made in such a way to damp the mechanical vibrations before they reach the microphone.
Any good ideas how to design high perfomance power management for high performance audio?
Did anyone try the above ceramic capacitors and could report the results?
In the end it seems it is not sufficient to replace ceramics caps and use better inductors, but one should mechanically separate the microphone from the rest of the digital/power circuitry to have a very (near zero) noise floor...talking about -65dBFS. The connection should also be made in such a way to damp the mechanical vibrations before they reach the microphone.
Any good ideas how to design high perfomance power management for high performance audio?