Neon Lamp
 |
At the end of the first
build I tried to calibrate the spectrograph. 6hrs later - no luck
and I was terribly frustrated. My mistakes were to try to do this
with the ST-8E attached AND had no idea what I was supposed to actually
see. Phil (from the UK) took pity on me.
He had built an L200 from the first batch and had successfully completed
and calibrated his spectrometer. Equipped with drawings provided
by Phil, I set about calibrating the L200 after the second build.
So rather than the camera I jerry rigged an EP holder and placed my
Meade 26mm EP at the focal point (the idea is to focus on the slit).
This was done by setting the spectrometer under a fluro light bulb.
To my complete surprise, the fully open slit was in view, almost
perfectly centred in the EP. I immediately took my Olympus
digital camera and took a photo through the EP (second image from top).
So the next step is to fine tune the position and angle
of the grating then lock them down with the locking screws. I
narrowed the slit and the ran the micrometer through the gratings full
rotation to get 0 order through blue, green and red parts of the
spectrum.
I then turned on the inbuilt Neon Reference lamp
and started looking for a spectra. Since the Olympus digital
camera is limited (1/2 second integration time), there was only so much
it was able to detect - but it did capture a good portion of the Neon
spectrum as shown in the bottom 2 images.
A quick explanation of the
actual spectra. The Neon Lamp image was loaded into the software
package Iris
then VSpec.
Since I had a plot of the Neon Spectra I was able to calibrate the
extracted Spectral lines and generate the plot. The Green spikes
represent the actual Neon Reference lines from the spectral library.
The broad spectral lines in my plot indicate a wide slit (relatively
speaking and not quite being in focus)
 |
Full Calibration Process
Step 1 - Camera focus
For a spectroscope to work, the Camera needs to be focused on
the slit and the target observed through the telescope also needs to be focused
on the slit. (If both conditions hold, the the target will also be focused
on the camera. My first attempt was to try to perform a focus with the
slit light from the outside. Although the camera could 'see' a clear beam
of light, it was not possible to see a crisp image of the slit. I could
see when I was out of focus but the area 'in focus' covered a lot of turns on
the helical focuser. Second option was to light the slit from the inside.
For this I put a Red LED on a long wire inside the spectroscope in front of the
slit. This worked well and I was able to image the slit and focus
sufficiently well to see the tool marks on the tiny slit jaws (see Figure 1
below).
Post note:
It became apparent after initial use, that the Camera focus on the slit
varied with ambient temperature. To this end, I will modify the
Spectrograph to allow a permanent internal light fitting to enable
refocus at the telescope.
Figure
1: This is what the inside of the L200 looks like from the CD camera when
it is internally lit. Yes, somewhat off centre but good enough to focus the
camera on the slit. One thing I do notice is that vertically, the camera
does not stay in focus with the slit. Here the slit is open 450 microns.
Step 2 - Slit Calibration
In order to use a slit spectroscope, one needs to know how
wide the slit actually is. One can measure the slit width directly and by
the FWHM of emission lines. This, of course, will only work if the
spectroscope is properly focused. First step is to take a very bright
light and a diffuser and shine it down the spectroscope then slowly close the
slit until the light can no longer be seen. If the L200 was fitted with a
high quality adjustable slit then that's what I should have seen.
Unfortunately this was not the case here. As can be seen in the images
below, the Slit jaws are never fully closed due, most likely, to imperfections
in the manufacture of the slit jaws and pits in the edge of the teeth.
This is acceptable for this level of instrument but it does limit the minimum
practical slit width.
Figure 2. Here you can see the view with the slit set at 0, 20, 25, 30 and
40 microns (left to right)
To date, all my testing has indicates that the unit
will not generate a slit width of less than 40 microns. This,
however, is possibly an indication that the unit is not focused
perfectly or that there is some odd or non linear movement in the slit
jaws. Again, not totally unexpected for a unit in this quality
bracket.
Step 3 - Micrometer Calibration
Neon (Ne) Reference Lamp Calibration
The next task is to test and calibrate the L200 against a Neon
Reference Lamp. Making various adjustments with the micrometer I was able
to generate spectra;
As you can see, there are overlaps between the spectra which
made it easier to calibrate the right 2 images. According to RSpec, my
system has a resolution of 0.7A/pixel. I have yet to determine this level
of information from the above spectra. So much more to learn. Of
interest is that adjustment of the micrometer is not linear (as expected).
Measurements taken in VSpec revealed a spread of FWHM values
across the image but a resolution of between 0.7432 and 0.7446A/pixel
Mercury (hg) Reference
Lamp Calibration
Well I don't have a Mercury Calibration Lamp so for
this test I used a commercial Fluro energy saver bulb fitted into a desk
lamp swung over the top of the L200 with a diffuser in place. This
allowed me to calibrate the Micrometer over a much wider range than was
available with the Neon lamp.
By providing almost 1/2 the spectral band
width in overlap, the identification and assignment of the emission
lines became easy. It is surprising how accurate the calibration
can be even with just a single emission line. The VSpec calculated
resolution for the L200 has remained in the 0.7432 to 0.7456 A/px range.
The result of these tests shows a not perfectly
linear relationship between the micrometer settings and the spectral band
observable at that setting as indicated below;
Further Analysis
Slit image shift
After the first night on the telescope, it became
apparent there were still some issues to iron out. One is focus and the
other (and possibly related) is image shift based on orientation of the
Spectrograph. I did a handheld trial with 4 basic positions and measure
the position of the slit.
As you can see there is a 30 pixel shift in the position of the slit from one
extreme position to the next. Could this be flexure? Could this be the
'weight' of the grating tension spring - perhaps a combination of the 2!
BUT, this test does not reveal an y significant difference in focus since the
FWHM of the slit did not vary appreciably.
Shifting Focus
Another issue that needs to be replicated is changing
focus. On first light there was a change in focus as the unit
cooled but there was also a change of focus (back toward the original
bench position) when I moved to the second target. Based on this
and be slit image shift, something seems to be loose in the unit.
This will need to be rectified before serious spectroscopic imaging can
be undertaken.
