[time-nuts] Time Interval Measurement Hint
brooke at pacific.net
Tue Mar 8 13:51:14 EST 2005
For a number of months I've been trying to get the C field set on a Cesium standard. This is a continuing learning experience. In order to help in setting the GPS elevation mask angle (now at 50 degrees) I started monitoring a number of parameters, but now have stopped most of them except for the standard deviation of the time interval. I'm taking 1,000 readings of the 1 PPS to Cesium standard 1 MHz zero crossing and the SR620 counter provides not only the mean value but also the standard deviation.
It turns out the the standard deviation is a great indicator that everything is working correctly. If this value gets to 40 nano seconds or higher there's a problem. Recently I tweaked the beam current on the FTS4060 (open loop, center beam current, close loop) the green lock LED turned on in a number of seconds, BUT the SDEV value was many hundreds of ns! Even after giving the FTS4060 overnight in the morning the SDEV was still way high. But unplugging the mains AC power for 15 seconds fixed the problem, probably be resetting the micro computer.
I highly recommend watching the SDEV value when making precision measurements.
w/o Java http://www.pacificsites.com/~brooke/PRC68COM.shtml
More information about the time-nuts