LgswDichroicSystem check

In the last days we found many issues with the movement of both the dichroic and fold mirror on the LgswDichroicSystemSX while on the LgswDichroicSystemDX the movement apparently goes smooth in both directions. In the past days we already verified that the movement programs loaded on the 2 systems are the same (14-12-02 version).

15:00: we climb on the telescope and we have a visual check on the Dx and Sx systems. We found several issues with the LgswDichroicSystemSX:

Mechanical issues

1. one of the dichroic cart profile rails is missing a nut to keep grease. The rail shows a large accumulation of grease in correspondence of the parking end stop. This will increase the torque required to deploy the dichroic at "cold start" and can be responsible of a movement failure triggered by the I0 switch that is not deactivated at the end of the indexed run (check lines 123 and 124 in the dichroic movement program)

00WN,123,31,030,38,+6000,x4,124
00WN,124,63,I0,0,158,125 

1. if the dichroic movement is aborted in between the 2 end stops occasionally it fails to restart when a new movement is triggered. This can be due to the large accumulation of grease in the endless screw.
2. the fold mirror seldom fails suddenly to exit the I1 end stop while going to park position.

These issues were not found in the LgswDichroicSystemDX and indeed any movement failure has been reported.

Possibly these mechanical issues can be solved re-programming the dichroic movement without a sensible impact on the deployment time:

• the indexed runs in the dichroic program can be performed with X1 multiplier instead of X4, this can solve issue #1.
• the ramp code to start the dichroic movement (especially the free run) can be increased to 42, this can mitigate issue #2.
• the ramp code to start the indexed run of the fold mirror can be increased to 42, this can mitigate issue #3.

Electronic issues

We found also 2 issues related to the wiring of the LgswDichroicSystem:

1. the LUCI cal unit interlock seems somehow coupled to the I0 signal from the end switches.
2. several times when the movement fails because of the above mentioned issues the control GUI disconnects. Could be due to a reset of the controller?

LgswDichroicSystemSX test

2016-10-22 (MB, TM): tested 2016-10-21 version of the SW (at horizon), both dichroic and fold mirror fail to exit both parking and working limit switches. We must apply 3 times the command to have the movement completed.

We increased both the "Go to parking position" and 'Go to working position" indexed runs to +/-50000 steps:

00WN,123,31,030,42,+50000,x1,124
00WM,143,31,030,42,-50000,X1,144

01WM,123,31,030,42,+50000,X1,124
01WM,143,31,030,42,-50000,X1,144

new movement programs saved as: rta_Dic_01_16_10_22_LBT.txt, rta_Mir_01_16_10_22_LBT.txt and uploaded on SX system.

Tested 4 times and working flawlessly. Deployment/parking times are approximately 1 m 18 s for the dichroic and 33 s for the fold mirror.