| review: The NVLT - a 12 inch f/4 Newtonian for Astrophotography (appr. € 5000,-) |
Inspired by my work for the ESO VLT-Instrument AMBER I thought a nice name for this telescope could be The NVLT - Not Very Large Telescope.
It was constructed and built from scratch but by using prefabricated parts.
Especially there was no mirror grinding envolved.
I am not patient enough for that job and excellent newtonian mirrors are available.
The main purpose of this telescope is CCD imaging with an unvignetted image of 15 mm in diameter and a maximum spot size for a small pixel CCD camera.
The specification is as follows:
- Main mirror of 12 inch (305 mm) f/4 λ/33 RMS, 42 mm height
- Focal length of 48 inch (1220 mm)
- Secondary with minor axis of 89 mm giving 29% obstruction
- Unvignetted image of 15 mm in diameter and a maximum spot size of 0.019 mm allover that field
- Tube with low thermal coefficient of expansion and 360 mm of inner diameter
- Very rigid mechanics with extra strengthening at the critical places
- Very good shielding against straylight
- Lowest reflection in the tube without using baffles
- Best possible (passive) air flow through the tube
- Precision mirror holders for good alignment and stability
- Motorized focuser without shift and very fine movements because of f/4
The NVLT without straylight shield and secondary mirror
Optics and related Issues
The main mirror was manufactured in Russia by an Ex-Intes optician.
A λ/33 RMS error on a mirror as fast as f/4 must be regarded to be very good.
This is corresponding to a Strehl ratio of 0.96.
Any better mirror would have ment an exploding price and very long waiting time.
But the achieved quality is already very promising.
A secondary mirror with a minor axis is giving a linear obstruction of 29%.
This is a very reasonable value for photography and a good compromise for casual visual oberservations.
This secondary is resulting to a 15 mm image field with full illumination and a focus point of 110 mm above the tube.
This relatively big back focus distance is needed for the motorized filter wheel and a quick-change-adapter for T2.
For my CCD camera I need a maximum spot size of 0.019 mm not to oversample the image and gain maximum detail possible in prime focus.
The coma of an f/4 paraboloidal mirror is increasing very fast when leaving the optical axis.
A simulation in OSLO light showed that the required spot size is only available within a diameter of 4 mm in the image plane.
Hence the focus point as well as the filter wheel and CCD camera distance had to be checked for the use of a coma corrector.
I will use the Baader MPCC which is demanding for exactly 55 mm distance from it's T-thread to the CCD detector.
The MPCC is delivering a spot size of 0.012 mm allover the image of a 35 mm film and hence meets my specification.
Click here for my test report about the MPCC.
The OSLO simulation also showed that the depth of focus is very small for f/4.
A linear focus error of only 0.05 mm is translating to a spot size of 0.02 mm on axis.
So a minimum focus movement of 0.04 mm is needed.
The JMI NGF-DX1 motorized focuser is just able to perform that fine movement.
According to that maximum focus error it is clear that the tube must not shrink or grow more than that when the ambient temperature is changing.
I have chosen a special compound material called "Pertinax" widely used in the german industry.
It is not yet very well known to ATMs.
The thermo coefficient of this material is close to 0.
Additionally it is easy to work that material, i.e. drilling and sawing.
The inside diameter of the tube was chosen to a generous 360 mm so that eventually arising tube currents can float well outside the light path.
This is also the reason why I will not use additional baffles inside the tube.
On the other hand stray light due to tube reflections is a serious contrast killer.
Some amount will always end up in the image plane.
In a long time exposure every little flaw will be visible in the final image.
Flat black is still very reflective at low angles.
So I picked a special material with black velvety hairs.
A lot of other ATMs reported good success with it.
After applying the material the tube was so dark black that I could hardly find the drilled holes for screws!
That never happened to me with flat black painted tubes - again very promising.
The best way to deal with stray light is when it never enters the tube, of course.
So I sawed two rings from the tube's end - each 30 mm wide.
The rest of the tube had just the correct size for the optics.
With these rings I built a stray light shield - again covered from inside with that black velvety material.
The top end aluminium ring is 2 mm longer than the tube and centering the stray light shield.
The NVLT with straylight shield, already under dark skies
Mechanics and related Issues
Though the tube made from Pertinax is much stronger than a Sonotube I decided to strengthen it with aluminium rings.
There are two behind the focuser screws, one behind the diagonal spider vanes and one behind the main mirror.
Focuser with CCD camera, the screws for the aluminium strenghtening are visible from outside
The aluminum behind the focuser was covered with black velevet, of course. The support screws for the mirror cells are M8 and M10.
Main mirror cell with 18 support points
The whole construction is very rigid.
The tube is supported by two surrounding mounting brackets with a distance of 250 mm and screwed to the mount by M8 screws.
The mounting brackets
Commissioning and First Light
The very first light of this nicely planned out and built telescope was - to be honest - a complete disaster!
12 months of planning and construction seemed to be lost.
The first frames were flooded all over by stray light.
A series of frames with the telescope moving in RA for an hour showed strange banana shaped reflections getting brighter and brigther.
The source could be identified soon: a near by street lamp was shining exactly through the small gap of the focusers tube and housing and a special angle of the telescope!
This focuser is black anodized allover.
But again - the low angle is making ist highly reflective.
I built a little baffle from black cardboard to shield that light.
A stray light blocker for the focuser
Also visible is a not so well covered reflective screw.
This was fixed as well.
Unfortunately some amount of stray light was still visible in the frames.
This light was more scattered aroound the frame.
After several tests I found out that light from the rear end of the tube could pass the primary mirror cell and shine directly onto the diagonal.
Though this end is down in the well shielded observatory with the back wall painted black it was enough to ruin long exposures.
The solution was an annular baffle.
The remaining whole is still enough for proper ventilation.
The baffle was again covered with black velvet on it's inside face.
Rear annular baffle, collimation screws can still be reached easily
Now the NVLT's image plane is only collecting light from the sky and it is ready to go.
I picked the 5.8 minutes wide galaxy NGC5005 - beeing almost in Zenith at that time.
A pixel of my CCD camera in prime focus is 1.1 seconds of arc wide.
My mount is able to track to that precision and the differential flexure of the guide scope is allowing for about 300 seconds of exposure time at zenith - so far.
I took 6 l-frames with 120 seconds each and a set of color-frames with 3 frames for each color.
The mirror's performance and the coma corrector's as well is very pleasing.
The smallest stars are appr. 2 pixels in diagonal and 3 pixels wide and are hence meeting the specification for not being oversampled.
Short exposure of NGC5005 showing no optical or tracking flaws.
Short exposure of NGC5005, enlargement of stars at the edge of the field.
Conclusion - Optimizations left
I am very satisifed with the performance of this 12 inch f/4 Newtonian astrograph.
However it is only a NVLT. :-)
The NVLT tracking the skies.
A last problem left to be solved is the performance at angles of 45 degrees above the horizon and below.
I first thought that the primary mirror seems to be pressed against one of the 3 support arms by it's own weight too much.
This is affecting the proper shape of a star from a disk to an egg and hence lowering the contrast of fine detail in the whole frame.
More images at low angles did not show these eggs.
Actually the problem is caused by tube seeing.
The mirror is very thick and hence containing a lot of thermal energy.
When there is a rapid change in the ambient temperature, the mirror is creating warm air flow known as tube seeing.
The air is not floating symmetrically when the telescope is pointed just above the horizon.
This is causing the egg effect.
I will test a little fan in the near future to get rid of that occasional problem.
Exposure of NGC3628 showing the egg effect at low altitudes.
Exposure of NGC3628, enlargement of the egg-shaped stars.
Exposure of M10 not showing the egg effect at low altitudes.
The NVLT from inside the observatory. The guide scope is mounted in parallel with the NVLT.
Search more images taken with the NVLT
|
|
|