Collimation of a Littrow spectrograph


Here is described how the Littrow spectrograph will be well collimated. Finally, you might exploit the potential resolution of the spectrograph's also, so you can win optimal sharp spectra.

First step:
Thus, the reflection grating can disperse the light, the incident light to be parallel (divergence = zero). This will be achieved if the slit is exactly in the focus of the collimator. To verify this setting, the following procedure is useful: You remove the grating and looks with a telescope in an infinite set through the collimator on the back illuminated slit. The slit is in the focus, if it is to see clearly sharp (including its cutting and the dust contained in the slit). Is this not the case, the collimator or the slit are moved along the optical axis until the slit is sharply focused. That was used to collimate telescope / finder / binoculars should have the widest possible focal length and magnification. The more accurately can be performed the collimation. Of course, the telescope / finder / binoculars must be previously set to infinity. The same eye should be used as later on the spectrograph.
Alternatively you can take pictures of the slit with a digitalcamera equipped with a long focal range lense.

Second step:
The grid will be installed again. Then illuminate the slit with a neon lamp (or other lamp producing sharp spectral lines) and examine the resulting image on the CCD camera. The aim now is to move the camera to the optical axis at the point where the lines are sharply on the CCD. However, usually only succeed in the sub-optimal. Since a few hundredths of a millimeter difference in creating this optimal point been a blur, we will have to be satisfied with a rough edge (which is then optimized in the third step). The sharpness can judged exactly when photographed spectral lines (= slit images in monochromatic light of spectral lines of the calibration lamp) as .fit-files are saved, and then in the .fit file the FWHM of the lines are measured with VSpec or Midas. Example: a theoretical resolution of the spectrograph 3.8 pixels (with SimSpec calculated), FWHM achieved by moving the camera to the optical axis 5.7 Pix. That's OK. The rest is done in the third step.

Third step:
Dieser Schritt ist nur sinnvoll, wenn sich die Kollimatorlinse (wie im Lhires III) sehr feinfühlig mit einem Feingewinde auf der optischen Achse verschieben lässt. Man verschiebt die Linse durch Drehen am Fassungsgewinde um geringe Beträge (Drehwinkel am Gewinde) und beobachtet, was mit den Spektrallinien auf der CCD passiert: Werden sie schärfer oder unschärfer? Ist man nahe dem Optimum, werden die FWHM der Spaltabbildungen wieder mit VSpec vermessen. So lässt im im Allgemeinen die Schärfe der Linien optimieren, so dass > 90% des theoretischen Auflösungsvermögens erreicht werden.

This step is only useful if the collimator lens (as in Lhires III) can be moved very sensitively with a fine thread on the optical axis. Move the lens by turning the thread, a small amount (rotational angle at the thread) and watch what happens to the spectral lines are on the CCD: get them sharper or fuzzier? If you are close to the optimum, the FWHM of the slit images are again measured with VSpec. In generally you can improve the sharpness of the lines so tha t>90% will reached of the theoretical resolving power.

This work recommendation assumes that the slit is fix mounted, sliding the collimator in a version with thread by turning on the optical axis and the CCD camera in its adapters grossly displaced along the optical axis and is then clamped.

back to content