:MEASure:JITTer:JITTer:UTJ?

Query Syntax

:MEASure:JITTerN:JITTer:UTJ?

Where N identifies one of four possible jitter/noise analyses {1:4}. For example, JITTer4.

Description

Returns UTJ (p-p), Uncorrelated Total Jitter, which is the peak-to-peak value of the total jitter calculated at a specific symbol error ratio (SER), after removing data-dependent jitter. It is the value of TJ(p-p) for jitter that is uncorrelated to the data pattern.

Stating that UTJ (p-p) = TJ(p-p) - DDJpp approximates the formula that is actually used.

UTJ (p-p) is measured using the same formula used for TJ (p-p), except that it is applied to the uncorrelated total jitter histogram, rather than the regular histogram. The general formula uses the dual-Dirac model TJ (p-p) = DJ (δ-δ) + Q*RJrms, where DJ (δ-δ) and RJrms are found using jitter decomposition and an appropriate value for Q is based on SER.

The uncorrelated jitter histogram is formed by first removing all jitter that is correlated with the data pattern, or in other words, subtracting out data-dependent jitter (DDJ) from the time-interval-error (TIE) data before forming the histogram.

Removing the data-dependent jitter is straightforward when the waveform uses a test pattern. In that case, you first average together pattern-length sections of the TIE data, which will average out all jitter that is not correlated with the pattern. You will be left with a pattern-long DDJ vector that corresponds to the data-dependent jitter at each edge in the pattern. To form the uncorrelated TIE vector, simply subtract the DDJ jitter just computed from the original TIE vector. More specifically, for each edge in the original TIE data, first figure out which edge it corresponds to in the pattern, then subtract off the corresponding jitter value from the DDJ vector. The uncorrelated jitter histogram is the histogram of these adjusted TIE values. If the waveform does not use a test pattern, you can still compute the data-dependent jitter, but it is much more complicated.