WTPA2 Noise Floor and Supplies

Hi All,
In building these pre-assembled units, I've had to fix some and poke around with the scope.
One thing I noticed when looking at the noise floor in these devices is that they REALLY care a lot about your DC source.

When run from a bench supply, WTPA2 is dead quiet.  When run from random wall warts, perceptible background and clock noise can be dependent on the wart.

Adding a larger bulk bypass cap (C12) across the DC input can help this.  Just fussing around on the bench, using a 1000uF instead of a 100uF made a big difference.

And naturally, any noise picked up on your input lines gets amplified no matter what, so no sloppy input lines either.

In case you were curious.
TB

Comments

  • Are there any readily available (i.e. from Mouser, Jameco, or the like) wall-warts that have worked especially well?
  • I have a 9v 800mA Radio Shack supply which especially sucks.

    I have something very similar to this:
    which is not bad but not amazing.  If you have no output from WTPA2 and the gains cranked, you can sometimes hear the SMPS kick over when current draw changes (like a light blinks).

    For picking wall warts, look for ripple voltages in a datasheet if you can find them.  You could also build an input filter out of caps/inductors/common-mode chokes.
    In future, I might put space on the jack board for one.

    TB

Sign In or Register to comment.