Waveforms with AD2 - measuring Gaussian noise from switching regulator output

[longboard]

Hello, I am attempting to use Waveforms with the AD2 to measure the Guassian noise from a +5V switching regulator output. My goal is to generate a probability density function (PDF) of the noise amplitude for the output of the regulator. I want to determine if the regulator's output has a noise spec that exceeds 100mV peak to peak for only 0.1% of the time.

Below is an example of a PDF plot, similar to what I am attempting to generate.

The link below describes the process I am attempting to perform regarding the PDF and Guassian noise analysis.

https://www.allaboutcircuits.com/technical-articles/noise-in-electronics-engineering-distribution-noise-rms-peak-to-peak-value-PSD/

I have been experimenting with the O'scope's "histogram" functionality within Waveforms while using one channel from my AD2 connected to the switching regulator's output. I have set my AD2's BNC adapter to AC coupling (via a jumper) and the vertical span is set to 500mV range. I reviewed section 5.3 in the Waveforms reference manual, but am a bit unclear regarding the relationship between '%' along the 'Y' axis and the histogram's resolution setting.  If have noticed that if I reduce the histogram's resolution the '%' increases along the Y axis as can be seen in the images below.

I may have to log a raw time series using Waveforms and process the data to generate the PDF using the process described in the link above, but I would like to better understand the O'scope's histogram functionality within Waveforms to see how I might be able to use it for my noise analysis.

Thank you

Sean

 

 

 

[attila]

Hi @longboard

The histogram Y axis represents the percentage of samples within value ranges reported to the total visible number of samples. The histogram width is the full ADC channel range and the default value range is the ADC resolution.

[longboard]

Hi @attila

Thank you for the explanation. Does the "resolution" setting on the histogram configuration panel effect the number of bins within the horizontal span along the X axis? The image below has the resolution set to 'auto' but I have experimented with many of the other option and included an image showing the 64ki setting as well.

Also, how did you get the bar graph representation on your histogram? My histogram is composed of points connected by lines.

Based on your explanation is it accurate to say that the histogram below is showing that nearly 12% of the samples shown in the 80uS O-scope trace are approximately 0.6mV in amplitude? I just want to make sure that I am correctly interpreting the histogram. 

Thank you

 

[attila]

Hi @longboard

Yes, when the histogram resolution is lower than the ADC resolution there will be probably more samples in hist steps, increasing the %

The bar option is available in newer sw version:

 

[longboard]

Thank you @attila. I downloaded the new version of the software and will experiment with the bar option for the histograms.