Virgo Logbook

Injection system (General activities)

gosselin, derossi, montanari, carbognani - 15:33 Tuesday 24 March 2020 (48792)

Flip mirror between SL and neovan

One of the critical point of the injection subsystem if operated from remote is a possible break down of the chiller cooling the pumping diodes of the neovan.
When it happens the neovan goes in security mode and switches off but it is still seeded by the 20 W of the slave laser. Those 20 W are focused into the neovan head and could possibly drift because of thermal effects that are differents that the ones in normal conditions.

To avoid that, this morning, we installled a flip mirror between the slave laser and the neovan that send the 20 W on a beam dump (see attached picture)
This flip mirror can be driven from remote. From the VPM go in the TCS section, subsection ttl_inj, button INJ_AMP_FLIP_UP to block the beam.
We are also implemented an automatic command in the metatron to flip it automatically when the DC power at the output of the amplifier is too low. More info to come.

We also took advantage to set the current of the pumping diodes of the neovan to 0, reverse the flow of the cooling system by swapping the main chiller of the injection (like we did in entry https://logbook.virgo-gw.eu/virgo/?r=48683 ). This, again, had a positive effect on the peltier corrections that were slowly increasing during the last days (see the attached plot).

Images attached to this report

Comments to this report:

Montanari - 19:03 Tuesday 24 March 2020 (48798)

The device used to control it from remote is in ELE-ROOM in the INJ4 rack, the same of the TCS HWS installed on the EIB.

The channel used is the second and the cable is the red pair of a snake labelled 'INJ'.

Images attached to this comment

derossi, gosselin - 17:25 Thursday 26 March 2020 (48823)

The DC signal transmitted from the amplifier is PSL_AMP_DC. As shown in the the first plot of the attached figure, in normal conditions the AMP DC value is around 0.45 V (on the left), when both the amplifier and the SL are off it goes to 0 (on the right, green part), while when the amplifier is off but the SL is still seeding the signal is around 0.05 V (in the middle, red part). The last one is the case we aim to avoid by flipping the mirror between the SL and the amplifier.

Images attached to this comment

carbognani - 19:30 Friday 27 March 2020 (48837)

Today I finalized the testing of the automation logic added to the INJ_MAIN node to implement the safety check via the AMP_DC_Check() function and the assert_AMP_safe() decorator.

The check uses the average over a 10 sec buffer of the PSL_AMP_DC signal. The tests did not extend to the effective flipping of the mirror (lines commented out) since this would imply make some preparatory setting on the INJ system. This testing will be done at due time once we are back operational.