[time-nuts] HP5061B Versus HP5071 Cesium Line Frequencies
Donald E. Pauly
trojancowboy at gmail.com
Wed May 31 22:17:07 EDT 2017
Our attempt to measure the low frequency noise to determine the
possible lock improvement was a failure. We will have to get access to
the Loop Gain pot to get a flat signal at low impedance. Beware of
low frequency roll off.
I found out that the feedback amplifier on the A7 board causes the
electron multiplier output to roll off above 19 cps. It places an
8,200 pFd condenser across the 1 M load resistor which causes the roll
off. Here is a linear sweep up to 8 kc.
http://gonascent.com/papers/hp/hp5061/waveform/wideswee.jpg It shows
the 274 cps second harmonic as well as the 16 cps roll off.
Our previous posts involving various modulation schemes were correct
because we drove the scope directly from the electron multiplier.
While the meter driver is a convenient amplifier, you have to
disconnect C1 to prevent the low frequency roll off. The feedback
amplifier has a gain proportional to frequency. This restores
flatness above 19 cps for the loop gain pot. I got a gain of 141 at
137 cps with a lagging phase angle of 90°. R10 is loaded by C1 which
drops the level at that frequency by 7:1. Effective gain is about
20:1 on the Loop Gain pot in the hi gain position of S1. Gain is
about 7 in the low gain position. In both positions low frequency
roll off is about 19 cps.
More information about the time-nuts