Final Test Result of S1500117: https://dcc.ligo.org/LIGO-S1500117
After some staring the schematic and checking some TFs, I found that the DAQ channels for MON2 have a mistake in the circuit.
Differential driver U14 and U15 of D0900848 are intended to have the gain of +10 and -10 for the pos and neg outputs.
However, the positive output has the gain of +1.
Daniel suggested to shift R66 and R68 by one pad, replace with them by ~5.5K and add a small wire from the now "in air" pad to
the GND near pad 4.
The actual modification can be seen on Attachment 1. The resulting gain was +10.1 as the resisters of 5.49k were used.
The resulting transfer function is found in the Attachment 2. ow the nominal magnitude is ~x20.
You may wonder why the transfer function for MON1 is noisy and lower at low freq (f<1kHz). This is because the input noise of the FFT analizer
contributed to the BNC MON1 signal. High frequency part dominated the RMS of the signal and the FFT analyzer could not have proper range
for the floor noise. The actual voltage noise comparison between the BNC and DAQ signals for MON1 and MON2 can be found in Attachment 3.
|