I have been trying to understand if my echelle is stable or not in its wavelength calibration. I ran a test where I started the camera and the echelle temperature control and took calibration spectra at 15 minute intervals. At each point I collected 6 spectra and median combined them to give a good S/N ratio.
I then looked at the line at 6531A. Initially I used Bass to locate its position in Px and looked at the variation from sample to sample. Following that I did a similar thing using SPLAT-VO to fit a Gaussian distribution to the lines. Both methods gave similar variation in the samples although there was an offset of about 0.5 Px between them.
In both cases the variation (once the initial settling see later) was about +/-0.1 Px. The average dispersion was 0.16 A/Px (R~9300). Thus equivalent to 0.74 Km/s for 0.1 Px.
I then tried using the CCF function in ISIS and got the results in the attached diagram. Sample 7 was used for all the correlations. Clearly I need to do the trial again over a longer period as it seem it take up to 2 hrs for the echelle to settle down once switch on even though it's internal temperate (measure in the air) quickly comes under control.
There was no obvious relation to the shift in wavelength with the external temperature or pressure during the test.
As always advice and comment welcome.
- Stability test.jpg (37.8 KiB) Viewed 3008 times