Jennie W, Sheila D, Ryan S,
This morning Ryan and co were trying to lock at 200mW to allow the porcupine beam dump for the PRM misalignment beam to be checked at low powers. They were having trouble locking JAC at power below 2W. I improved the alignment of the PZT mirror (H1:JAC-PZT_PIT_OFFSET and H1:JAC-PZT_YAW_OFFSET are the channels) an the JM1 suspension while we were locked at 2W so I could look at the transmitted power.
We seem to be falling out in the 'SCANNING' state during locking at 200mW and if we lock at 2W and try and turn down the power it also falls out.
Sheila checked the JAC guardian and found out that the 'JAC_locked' checker function has a threshold of 0.5 which must be passed before we stop scanning the PZT and jump to the LOCKING state.
It uses the Beckhoff channel H1:JAC-TRANS_A_DC_NORMALIZED which gets normalised by H1:JAC-TRANS_A_DC_NOMINAL to do this check.
H1:JAC-TRANS_A_DC_NOMINAL is a scalar mA value that does not change with input power (Daniel rescaled this value on Monday manually to make the H1:JAC-TRANS_A_DC_NORMALIZED = 1 when the input power is 2W).
This was not happening so the check failed and the guardian jumped to DOWN.
Therefore Sheila has divided the H1:JAC-TRANS_A_DC_NORMALIZED by input power and set the threhold to 0.2, so this check should always come out about 0.5 for a well-aligned JAC and will therefore pass the 0.2 threshold.
I have committed the JAC guardian to the svn at userapps/isc/h1/guardian
Included pic of alignment changes for PZT mirror and JM1 plus input power monitor (H1-IMC_PWRIN_OUT16) and JAC lock check channel (H1:JAC-TRANS_A_DC_NORMALIZED).
There is still some failure mode in the SCANNING state when below 2W. The Beckhoff PZT controller does not trigger on the fringes on the JAC TRANS PD as the inbuilt trigger on the Beckhoff screen for JAC_PZT_SINGLE (lower limit of which is circled in blue is channel H1:JAC-PZT_DRIVER_SCAN_TRIGGER_CHANNEL_1_LEVEL) only works if LSC_CUST_WHITENING screen for JAC-REFL_A PD has a value for H1:JAC-TRANS_A_DC_NORMALIZED (circled in red) above H1:JAC-PZT_DRIVER_SCAN_TRIGGER_CHANNEL_1_LEVEL. We might need guardian to scale these values in the future with the input power if we ever want to regularly lock below 2W.
This isn't really how this is supposed to work. The PD has a normalization coeffcient that should be set so that the normalized power is 1 when the cavity is locked. If we want to automatically scale for different input powers, we should scale the normalization. This could be added to the PD screen using the laser power measured on the PSL table.