Memory Depth and Oscilloscopes

Have you ever wondered about memory depth or memory point specs shown in today’s digital storage oscilloscopes? Most people know about bandwidth and they know about sampling rate but often overlook the importance of memory depth. Or perhaps we should say deep memory depth. Today most users depend on their oscilloscope to help debug digital signals. To do so effectively, adequate or even more than adequate memory depth should be considered when selecting a new digital storage oscilloscope (DSO).

The real time sampling rate spec on an oscilloscope only indicates the maximum sampling rate that can be achieved. To calculate the actual sampling rate, users should know the memory depth of their oscilloscope. By knowing the memory depth and its typical timebase setting, the true sampling rate can then be calculated.

digital storage oscilloscope stores samples into the oscilloscope’s memory. The larger or deeper memory allows for more samples to be stored. The more samples stored, the higher the sample rate. In other words, deep memory allows users to maintain the DSO’s maximum sampling rate across a vast selection of timebase settings. With this higher sustained sampling rate more reliable and more accurate measurements are obtained.

A typical challenge when using a digital storage oscilloscope is to capture adequate cycles of both fast and slow signals simultaneously while maintaining enough data points and resolution to zoom in and closely view signal details. Without sufficient resolution between data points, it is virtually impossible to determine what is actually going on with their project or design. Insufficient resolution means the user could be completely missing events like glitches and anomalies. Without an oscilloscope with fast real-time sampling and deep memory, serious issues like these can take hours or even days for an engineer or technician to eventually discover. An oscilloscope boasting fast real time sampling but with a shallow memory depth compromises real-world sampling performance and provides an incomplete picture of the digital and analog interactions in the subject design.

If you use your oscilloscope to analyze a faulty circuit during power up, even a very powerful triggering capability won’t answer the question of where to start looking for the cause of the anomaly. The solution would be the availability of deep memory which allows the user to observe a full start-up cycle with high resolution for longer periods of time. This then allows the user to trace the path between the “issue” as well as identify the root cause.

When using an oscilloscope to analyze the spectrum of your signal, access to deep memory plays a critical role. Frequency resolution is directly related to the amount of real time displayed on the screen. The more time displayed the finer the resolution. Additionally, the maximum frequency that can be viewed is directly impacted by the oscilloscope’s sampling rate. A high sample rate allows the user to observe at a higher maximum frequency. Without adequate memory depth it is not possible to capture and view at longer periods of real-time with the necessarily high resolution.

So, in conclusion, memory depth provides things that end users are looking for and need from an oscilloscope. This includes confidence associated with deep memory including the ability to deliver a high resolution waveform capture attributed to the high level of sustained sampling rate, the ability to trace symptoms back to the primary cause when a good trigger event cannot be defined, and the ability to view longer periods of real time data (especially when viewing both digital and analog signals at the same time.)


George Leger has a Masters in Electrical Engineering from Stanford University, worked in private industry pioneering surface-mount technology and in government research labs for twenty years, published several papers on surface-mount technology, co-authored papers published in national symposiums on accelerator technology, was past president of SMTA and an adjunct professor at the community college level, holds a patent, and is a certified microchip design partner, serving as a consultant to many companies developing electronic circuits.

Leave a Reply