An upconverted "square" signal will create a spectrum with two side bands. One should be rejected by a filter. Your "simulation" makes no sense at 410 MHz.[/QUOTE
This is a "
Harmonic Balance" simulation, and does not show the entire frequency spectrum, but only an impulse with the power amplitude value on the carrier frequency; Agilent ADS doesn't give you the possibility to see the signal the usual way we are used to, probably because it requires too much computation. Here is the image of the schematic with the simulation tools: **broken link removed** .
The bias tee is needed to remove the DC signal from the output (and some other purposes), and doesn't interfere with frequencies higher than 50MHz (I have tried working without, but still had the same results. The 50ohm load replaces the next component.
While the frequency view may be misunderstood since it's unorthodox view, I can assure you that it's very dependable when you hame many tones; the transient view is the same as usual.
Here is a hint about Harmonic Balance:
https://en.wikipedia.org/wiki/Harmonic_balance ;