Vout p-p tolerance OR ripple ?

[ez24]

From this site:

http://webench.ti.com/webench5/power/webench5.cgi?app=powerarchitect&lang_chosen=en_US

there is this setting (default 10%)  Lower left corner:

                 Vout p-p tolerance % =  ___ %

The default number is 10.  My questions are:

- is this 10% ripple?

- is 10% acceptable ripple and can I assume this is for a switching regulator?

- if a linear is attached on the output of a switching regulator - what % could be expected ?

- would % ripple matter for different voltages  ie 10% for 3.3 volts would be bad whereas 10% for 12v would be ok ??  ie should different voltages have different ripple limits?


thanks

[MagicSmoker]

10% ripple is the default?!? 

That *might* be acceptable if the output of the switcher feeds a linear regulator, but if it will be used directly then 1% is a more typical spec.

Generally speaking, % ripple rather than Vpp is specified because the latter is broadly applicable over a wide range of output voltages. In other words, the same ripple percentage tends to be specified for 3.3V, 5V, 12V, 24V, etc.

[T3sl4co1l]

10% peak-to-peak is, for example 0.5Vpp for a 5V output, and so on.  So the voltage varies from 4.75 to 5.25V, if the peaks are symmetrical (which need not be the case).

This will most likely be an estimate, at the switching frequency.  RF switching noise can greatly exceed this, and depends greatly on topology, component choice and layout.  Additional filtering may not be necessary for digital applications, but is typically critical for analog.  Additional filtering on the input is also desirable for reducing EMI/RFI emissions problems (depending on situation).

Tim