Hi Jo,
Thanks for reaching out.
The propagation delay (tpd) variation specified in the datasheet is due to process variation between samples and temperature variation of the application. The load capacitance CL doesn't directly affect propagation delay, it primarily affects the rise and fall times. This increase in rise/fall times could lead to higher propagation delay measurement as the 50% point in the rising and falling edges will move in the time scale. This is a measurement error only while the device will still be producing the stated propagation delays.
Hence, to show propagation delay measurements, we have used a typical use-case load capacitance of 15pF and estimated tpd. The rise/fall times are approximately 2ns for 15pF load, a decrease in CL doesn't decrease these times much and hence, tpd values won't change much. If CL is increased, then rise/fall times can increase which can lead to higher tpd measurements.
You can refer to the below propagation delay vs temperature plot from datasheet, it shows how propagation delay varies with temperature. If customer's application doesn't operate upto 125C and only operates upto, let's say, 85C then they may see low propagation delay values. Please note that the below plot only shows variation with temperature and voltage, it doesn't show variation due to process. Process variation is a variation in performance due to manufacturing differences, this is covered as a statistically data in datasheet specification limits.
Let me know if you have any further questions, thanks.
Regards,
Koteshwar Rao
