Thanks for some helpful pointers.?
I should have been maybe been clearer. It's a dual PLL based system which I ultimately want to test (to help avoid the reference source noise level being the limiting factor), so definitely need a carrier or it won't lock (and unlocked the transfer function will look very different - even if the loop lock finding circuits are turned off). To make life simpler I'm starting with only one loop in use (ignoring the second mixer).
I'm trying to measure offsets -100dBc down at 10Hz offset with a noise floor below -180dBc. At such close offsets well within the PLL loop bandwdith and there is a need to compensate for the PLL closed loop transfer function (which is measured by injecting band limited white noise into the loop which is coming from a direct digital generator which sounds similar to the E5500). But reading that documentaion has indeed been useful.? I'm familiar with the 3048 code as the HP-BASIC program is readily available (and very well structured and pretty easy to follow apart from two mysteries a) the routine Hp3561_range - which contains machine code for the 3561 to change the range preserving phase relationships which is fine, but why does this matter when all the measurements are in scalar network analyer mode (magnitude ratios) and b) why are there poles in the open loop response at 800kHz and 1.2MHz.
In further hunting I found this paper which feels useful
Alan
