Power supply Ripple measurement problem

[kumarghz]

Hi,
I have to automate the measurement of the ripple as indicate between the 'cursor 1' and 'cursor 2' in the figure below.
The ripple is measured by taking the Vp-p. The 'cursor 1' and 'cursor 2' has to be leveled manually. However, we need to automate this for the mass production testing. When automated, the scope measures the Vp-p(max) where it actually measures the switching transient which is very high. You can notice the high switching transients in the figure below.

Do you have any idea on how I can measure the ripple by ignoring the transients. I tried to set the filter BW to 20MHz but it doesn't helps much to suppress the transients.
Please help.

 

[FvM]

At first sight, you are missing a clear specification of "ignoring the transients". Obviously there must be an algorithm or calculaton scheme to derive the value.

Although a statistics text book can suggest you a lot of suitable algorithms, your oscilloscope probably won't support it. On the other hand, if you try to find a technical reasoning behind the ripple measurement criterion, a Vpp value of the low-pass filtered seems to make much more sense.

 

[chuckey]

Put a low pass filter between the wave form and your measuring equipment, try a 1K in series and a .1MF from the measuring side of the resistor and earth.
Frank

 

[FvM]

According to the 45 kHz ripple frequency, the low-pass filter should have a 100 kHz to 1 MHz cut-off frequency.

 

[dick_freebird]

Using one of the BW limit options might clean it enough
to get repeatable. Otherwise, consider using an applied
discrete filter like these guys mention, an "amplitude"
'scope function which tends to ignore spikes (check the
docs) or maybe a true-RMS (AC coupled). For purposes
of acceptance testing it's probably enough to get to a
repeatable screen for "normal", and fit the limits to the
population, and not essential to get the last little bit
of absolute accuracy.

I'd start with looking at the 'scope "measure" functions
list and see what that holds for you.