With traditional logic analyzers, capture time and resolution are negatively correlated. If you adjust the sample-rate to enhance one parameter, it comes at the expense of the other. Increasing resolution causes a corresponding decrease in capture time, and vice-versa. Since you can't optimize both, you must compromise both in search of an acceptable balance. This is particularly problematic with low duty cycle signals or fast-but-sparse data.

The minimum acceptable sample-rate is often much higher than people assume. It is very subjective, but is driven by the maximum acceptable distortion or edge uncertainty. If you are interested in measuring pulse widths, the over-sample rate (pulse-width/sample-period) determines the pulse distortion and measurement error. Obviously an over-sample rate <= 1 results in missing pulses and generally garbage data. An over-sample rate of 10x still results in a 10% uncertainty in the measurements (and display). So the sample-rate must be faster (and usually much faster) than the data rate.

Even with relatively slow signals, you might need a high sample rate to measure the skew or setup/hold times between signals. Again, the over-sample rate (skew/sample-period) will determine the accuracy of these measurements. The sample-period must be less than the setup/hold times to ensure you can even tell which signal preceded the other.

Often, when you use the minimally acceptable sample-rate, you find that the capture time is still unacceptably short.

As you would expect, compression increases the number of samples stored in a given amount of memory, thereby increasing the capture time. Less obvious is the fact that it also allows for improved resolution and all of its benefits.

DigiView's hardware-based compression automatically optimizes both capture time and resolution. It gives you the longest possible capture times by storing compressed samples at the DATA-rate (1 sample per transition.) Note that a traditional logic analyzer must store several samples per transition (over-sampling) to achieve any kind of acceptable distortion levels.

At the same time, it gives you the highest possible resolution by always sampling at its maximum sample rate. This does not fill up the storage because the compressor removes the redundant samples.

Best of all, this is fully automatic. It does not require any user settings or need any advanced knowledge of the data format. The compressor adapts to the data characteristics in real-time, providing optimized results even with highly variable data rates (like bursts or low duty cycle signals.)

See for yourself. The following interactive chart lets you specify different data patterns and then compare DigiView's compressed sampling with that of a traditional logic analyzer.

Set the Pulse Width and Duty Cycle to emulate your data. Then adjust the sample rate and observe how the two approaches compare in the 'Results' chart. We also show the distortion effects of the sample rate in the graphic at the top of the 'Configuration' chart.

To ensure apples-to-apples comparisons, we assume the same storage and maximum sample rates for both units and set the minimum pulse width to a reasonable value for the maximum sample rate.