FIRE – Updates

Coldhead temperature instability

20240822 We have observed an increase in coldhead temperature under cold ambient conditions. To maintain proper control of the detectors temperature, staff is continuously monitoring and adjusting the temperature set points in response to these ambient conditions. So far, no impact on the science data has been detected. We are coordinating with the instrument PI to plan the replacement of the coldhead.

FIRE on AUX1:

As of September 10th, 2022, FIRE was moved from AUX2 to AUX1 folded port.

To orient the slit along the parallactic angle, in fields 8 and 9 in the catalog file, you should use the following values:

Echelle mode: -30 HRZ

Longslit mode: 60 HRZ

If you would like to orient the slit along a particular position angle, replace these fields with the following

Echelle mode: [PA-30] EQU

Longslit mode: [PA+60] EQU

In addition, due to this move, there are some changes until all relative code is updated in the nominal path:

The offset A-> B feature in the Telescope-Control GUI does not work and ABBA offsets need to be performed by the command line in IDL. Details in the manual (Dithering your target along the slit).

For offsets in the longslit configuration /ls should be added (fatv_abba,/b,/ls and fatv_abba,/a,/ls).

When using the command fatv_abba for offsets the position A/B is not updated in the header keyword.

To open the FATV slit viewer execute fatv in any terminal.

FIRE noise pattern:

Resuming operations after the uncontrolled warm up of FIRE in the March 2020 due to the compressor failure a light noise pattern has been noticed on the detector. This pattern does not seem to affect data reduction or the overall instrument efficiency, provided all standard calibration files are properly obtained and used.

Time keyword on headers:

Currently the header keyword ACQTIME1 is the one that most accurately reports the starting UT time of a single exposure. 

It has been noted that there is a discrepancy between the keywords UT-TIME and ACQTIME1. When a loop is executed, the ACQTIME1 is updating correctly while the UT-TIME keyword remains the same for all images obtained during the loop. 
To correctly update the UT-TIME keyword, the following formula could be used: 

• UT-TIME correct  = UT-TIME header + exposure time + readout time  (for single frames)
• UT-TIME correct on loop N = UT-TIME of + (ACQTIME1_loop(N) - ACQTIME1_loop(1) )    (for loops)

Readout times differ according to the readmode - refer to the manual for more details. 

Acquisition Camera:

The correct pixel scale for the acquisition camera is 0.147"/pixel 

Slit size: 

The correct length of the echelle slit is 6 arcseconds  and for the long slit is 40 arcseconds with the exception of the 12" slit that is 56" long. 

Software access: 

The fire software can currently be accessed also from the Llama work station 

Older Updates: 

FIRE intermittent behavior - solved:

As a result of work done on the Jade, we no longer detect lost images.

As of Feb 2023, FIRE has shown intermittent behavior with images not being properly saved. Although the issues is usually easily solved with a detector reboot, the PIs should have backup plans ready, just in case.

FIRE operational:

As of Dec 5th 2022, FIRE has been declared operational and is being offered for science.

FIRE not available for operations:

Due to the move of the instrument from AUX2 to AUX1 there are issues with the IGUS chain. The instrument team is working on a solution. In the meantime, starting Sep 21st, 2022, the instrument will not be available for science operations until further notice.

Light pattern on FIRE images:

UPDATE: As of April 17th, 2022, the FIRE front window has been cleaned and the issue has been fixed

In March 2022, a light pattern has been reported in the FIRE acquisition images. Further investigation revealed that the pattern is due to residual on the dewar window. This means that images obtained with the Spectral Array are also affected.

Afternoon calibrations should correct for this and the effect on reduced spectra should not be significant. Long-slit data could be affected more than echelle, since the offsets are larger. The instrument team is working on a solution.