In PNA-X,the randomized LO is set to avoid erroneous signal.If I check to enable randomized LO,there may be a dither when every LO frequency is set.But how far can the LO dither from it's original value?What's the largest frequency range of the randomized LO frequency?When randomized LO is checked,will this scenario happen:if I use 4 LO overlaps,there could be a frequency range within the span covered by less than 4 LO frequencies.Dose it necessary to add another LO to complete the measurement and change the number of LOs?Really appreciate that.
We always make sure all the LO's cover whatever range is requested. If you use normal, we use 4 or 5 LO's depending on where the "DC" hole lands. Each LO acquisition covers a range of +-34 MHz (newest firmware which ships in two weeks or so allows this to be adjusted) but has a DC hole at around +-1 MHz, so 33 MHz of data are collected on each side of the LO. When we step across the band, if the particual IF frequency lands outside the DC hole we use all 5 LO's for image rejection, if it lands in the DC hole we use 4. The LO steps are approximately (Max-2*min-2)/(N/2) or about 14 MHz (N being 4 LO's, but we always take 1 more for the DC hole). The extra 2 MHz gives us +-1 MHz to "dither" the LO so that it ensures the four LO offsets are not equally spaced. This gets rid of issues where there are signals that happen to land on exactly the LO spacing. But, for some instances a better result occurs with not randomizing the LO, and so we allow that too.