Peak-to-peak jitter is often described in two ways:
• It can be calculated as shown below using the RMS multiplier. This is predicted peak-to-peak jitter based on a defined BER (usually 1 x 10-12).
• It can be measured using test equipment. In this case the minimum period sampled is subtracted from the maximum period measured. As we have seen before this will depend on how many samples are taken.
The predicted and measured values will not be the same. The measured value will give a feel for the jitter in the system, but is somewhat meaningless in relation to BER. For example, the peak-to-peak jitter measured from 10,000 samples will not tell you the peak-to-peak jitter needed for BER = 1 x 10-12. You would need at least 1,000,000,000,000 samples to accomplish that.
Therefore RJpk-pk is usually calculated from RJRMS at a defined BER. Refer to application note AN-815 for more details. For other questions not addressed by the Knowledge Base, please submit a technical support request.