The noise floor for cable delay method, and SR560 noise level are measured. The result is plotted below. We are not sure about the

Volt to Hz conversion yet. Once we know the calibration factor, we can decide if we have to replace SR560 with something less noisy or not. 

The Red and Blue plots are taken through daq (16 Hz)and FFT, then divided by the gain setup (1e4) and corrected for AC couple.

Green and Purple plots are taken via SR785 for fast measurement.

From the plot, we learn that,

1) SR560 turns out to be the dominating noise source for the current cable delay technique, from dc to 0.2 Hz, since the noise level from SR560 alone

and the whole setup are comparable.

2) The flat part from 200mHz to 2 Hz is probably the noise from analog to digital converter (ADC)

3) SR560 has corner frequency around 100 Hz.

Peaks from 60 Hz and its harmonic become smaller if we disconnect the power cord, and run SR560 with its battery, see fig2.

%fft 
%[y,f]   = myasd('C3:PSL-GEN_DAQ14',981430365,3600*10,1024,16);
%[y16,f] = myasd('C3:PSL-GEN_DAQ16',981430365,3600*3,1024,16);

load data_2011_02_14.mat   
%[y,f]   = myasd('C3:PSL-GEN_DAQ14',981430365,3600*10,1024,16);
%[y16,f] = myasd('C3:PSL-GEN_DAQ16',981498948,3600*3,1024,16);

%correct for 30mHz AC couple on SR560
ycorrect = y .*...