[time-nuts] Trimble Thunderbolt question, splitting its output
csteinmetz at yandex.com
Tue Mar 31 03:46:05 EDT 2015
>I'm not sure how to avoid the change in input impedance with the
>clipper involved unless I were to increase R1 to a very much higher value.
The voltage noise density of a 3.3k resistor is about 7.5nV/rootHz,
for a total noise of about 23uVrms in a 10MHz bandwidth (compared to
2.7nV/rootHz and 9uV for a 475 ohm resistor). I doubt that changing
R1 to 3.3k would raise the jitter measurably.
With R1 = 3.3k and R5/R6 = 33k (to reduce input attenuation with the
increased R1), the input impedance for the first ~600mV (+/-) of
input voltage would be ~50 ohms (100||100|33k), and for input
voltages >600mV it would be ~49.2 ohms (100||100|3.3k||33k). (Again,
with a transition zone of ~100mV as the diodes turn on.)
More information about the time-nuts