The benefits of partial assist depend on the situation, but typically might be on the order of a 50% reduction in VCO calibration time. This is helpful, but many applications demand more. Full assist addresses this by bypassing the calibration altogether and forcing the correct value. The challenge with this is that these values vary from part to part so one has to lock the device beforehand with the standard calibration and read back the settings. Then one can force them in and bypass the calibration time. Full assist provides the benefit of the fastest calibration time and also consistent values, although it requires a lot of programming overhead. Also, if the write speed is slow, then this will limit how fast the VCO can change frequencies. Full assist VCO calibration can dramatically improve the VCO calibration time, but requires significant setup, so this is why no results are shown for this approach. Another drawback with the full assist mode is that it takes several register writes to change the values of VCO_SEL, VCO_CAPCTRL, VCO_DACISET, PLL N value, and PLL fraction. Between these writes, the PLL will not be at the correct state and will slam the VCO tuning voltage to the rails. To mitigate this, the charge pump can be tri-stated, but this is more programming effort. In summary, full assist calibration is fast and it also has the benefit of always giving the same calibration settings for the same frequency. The drawback is increased programming overhead and also it requires the setup effort of locking the device to all the required frequencies and reading back and storing the calibration values so that later on they can be forced for full assist calibration. For more details and examples with full assist calibration, see the Streamline RF Synthesizer VCO Calibration and Optimize PLL Lock Time Application Report.