For the first time to our knowledge, continuous nonsegmented channel waveguides in periodically poled KTiOPO 4 with guided orthogonal polarizations are used to demonstrate type II background-free second harmonic generation in the telecom band with 1.6% / ͑W cm 2 ͒ normalized conversion efficiency. This constitutes a 90-fold improvement in aggregate conversion efficiency over its free space counterpart. Simulations show that the guided wave device should enable the measurement of timing fluctuations of optical pulse trains at the attosecond level in an optical cross correlation scheme. © 2009 Optical Society of America OCIS codes: 190.0190, 070.4550, 270.2500, 230.7370, 190.2620 Mode locked lasers can inherently generate pulse trains with very low timing jitter-down to the attosecond (as) level or even lower [1][2][3][4]. As a result, they are excellent candidates for applications where timing accuracy is of extreme importance, such as high speed analog-to-digital conversion and precise synchronization.The low jitter characteristics of various kinds of mode locked lasers have been well documented [5][6][7][8][9]. Although the measured jitter has been very low in all of these studies, theoretical limits are still much lower, indicating that either there are unmodeled additional sources of noise, or the measurement techniques lack enough resolution. As a result, in order to verify the ultralow timing fluctuation property of mode locked lasers, and to exploit it for timing purposes, first and foremost, one has to be able to precisely evaluate their jitter performance.A variety of methods have been devised for this purpose. Optical cross correlation is one of them [10]. Recently, it was experimentally demonstrated that a balanced optical cross correlator could show unprecedented precision in measuring the timing jitter of optical pulse trains [11]. In this scheme, the two pulses whose relative jitter is to be measured are projected onto orthogonal polarizations and then launched into a quadratic nonlinear crystal. Due to the group velocity mismatch, the two polarizations travel along the crystal at different speeds. A dichroic coating at the end facet discriminates the fundamental harmonic (FH) against the second harmonic (SH). The SH signal is measured by the first detector of a balanced receiver, and the reflected FH fields travel backwards and generate a new SH signal that will be measured by the second detector of the receiver. The length of the crystal is chosen to provide a one-pass delay between the two FH fields, approximately equal to their pulse widths. If the two pulses are initially arranged to overlap temporally with each other at the end facet, it will correspond to the zero crossing point on the differential output trace of the balanced detector, assuming no depletion of the FH pulses. The amplitude of the electrical signal from the balanced receiver is henceforth proportional to the temporal difference between the center of gravity of the pulses. In the presence of timing jitter, the varying electric...