The behavior of the sidebands during the last lock is quite strange: sometimes they drop to a very low level and stay down showing a certain instability; sometime they are back to a normal and stable level. This trend is common to all the PDs demodulated at 112 MHz (fig 1), and is seen also by B1p, which contains the sidebands for a certain amount (fig 2). There is nothing similar in the carrier, for example B4 (fig 3). At a first look, I cannot find anything in the alignment signals which could explain this behavior.
The channels
- DAQ_LNFS_56MHz_raw_mag shows the amplitide of the 56MHz line at the level of the LNFS
- DAQ_EOM_56MHz_raw_mah shows the amplitude of the 56MHz line at the level of the EOM readout
The fluctuations at 112MHz of the photodiode signals seems to be directly correlated to a variation of the 56MHz modulation at the EOM level.
Attached is a very long trend (from Christmas) and a recent period (from the beginning of February) of the LNFS 56MHz output signal (DAQ_LNFS_56MHz_raw_mag_mean) and of the EOM monitor out (DAQ_EOM_56MHz_raw_mag_mean).
It appears that a certain drop in the latter signal is visible from February 2 at 10:00 UTC, but the largest change in the modulation index (drops of about 5%) starts on February 7.
Between the LNFS output and the EOM monitor output there are the RF power amplifier and the EOM itself.
For the first one we checked (in lab, on a spare unit) that 1V of the 24Vdc power supply variation create a 0.7% variation of the RF output level.
If these drops persist, we can try to change all RF power amplifiers and their power supply module with spare units, in order to divide the problem between the RF chain and the EOM.
Morning activities:
with Nocera available from remote we performed the following actions:
- 09:00 UTC - replacement of RF AMPLIFIER in IER with a spare one (fig1). No effect on 56 Mhz modulations (plot).
- 10:00 UTC - swap cables CH3 / CH2 on RF DISTRIBUTION rack (fig2) in IER. The behavior seems to be better on 56 Mhz, but the action affected the 6 and 8 Mhz. According with Nocera, we decided to replace also the RF DISTRIBUTION.
- 11:00 UTC - replacement of RF DISTRIBUTION in IER with a spare one and swap back cables CH3 / CH2.
ITF relocking in progress, phase tuning ongoing...
We reached LN3_SQZ quite smoothly. I had slightly tuned the MICH/SSFS phase, the B1p_56MHz phase and the working point.
Parameters tuned are:
- B1p_56MHz_phi0 from 2.635 to 2.67
- DIFFp TY SET (LN3) = -0.056
- DIFFp TX SET (LN3) = 0.008
- BS TX SET (LN3) = 0.01 from 0
- DIFFp TY SET (LN2) = -0.08 from -0.065
The problematic (and yesterday removed) RF_distribution crate electronics was carefully analyzed in lab: we found that one of the 12 channels inside (the 56MHz modulation path) had a faulty SMA connector.
We do not know why and how, after about 5 years of regular and honorable work, he decided to become intermittent (Il diavolo si annida nei particolari); we changed the SMA connector and carefully checked all the other connections. Now the original crate is operational again, with all the signals at the right levels.
Following Alain's logbook entry there seems to be still some problems with the 56 MHz modulation level, so we plan to bring back in operation the original RF_distribution crate during tomorrow morning maintenance time.
Attached are some plots of the spare RF_distribution crate that is now in operation: purple before the noisy period indicated by Alain, blue during. We can see that the noisy period mainly affected the low frequency part of Hrec.
Today (07:50 --> 08:15 UTC) we replaced again the RF_distribution crate, putting the old one (repaired and checked) back into operation.
Attached are plots of trend data before and after, as well as the FFTs
