Jump to content


- - - - -

Findings on a home-built refractor with a 150mm f/10 achromat doublet from ISTAR

Discuss this article in our forums

Findings on a home-built refractor with a 150mm f/10 achromat doublet from ISTAR

By Stanislas Maksymowicz

Why a refractor

When I started with astronomy in the 1970s, owning an astronomical instrument was a dream. This still holds today for many, but in addition it also represents a tool that has to be efficient in catching details under varying sky conditions, determined by parameters like transparency, seeing and local observing conditions which can range from poor to excellent.

In this perspective, and after testing a lot of scope designs with diverse quality levels, it was time to acquire the scope that will help me to monitor the planet activities of Venus, Mars and Saturn.

Since the last 10 years, I had in my hands a refractor of 100mm by Vixen, few C8 starbright coatings, few Maksutov-Cassegrain of 150mm F10 and 15 from Intes, few Newtonians of 150mm F5 Celestron and F8 Orion Optics, a 250mm F4.8 from Orion Optics, and lastly a 305mm Schmidt-Cassegrain from Meade, a refractor 152mm F6.5 from Antares that followed the Cave 203mm F8.

The conclusion was after some long time use:

  • the absence of central obstruction for visual use get better contrasts at the first look,
  • the 6” refractor was working better into seeing with less contrast degradation (sometimes outperforming the 10” due to the seeing conditions).

This were the main reasons, with some others given later at the conclusion.

Why ISTAR-Optical?

This has been a strange quest since in the 70s almost nobody was offering affordable lens doublets or triplets at good quality levels (there was no Internet at that time either). Their price levels were so high that everybody was going for reflectors: this was the 200mm Cave scope personal buy instead of a refractor.

Excellent jobs done with but so subject to the instrumental seeing effects with its open air tube. Except the Jaegers and the Clavé, almost nothing was available, including Lichtenknecker.

Recently on the CN site, somebody was coming, Istar Optical, offering, in first, achromatic doublets exceeding the Rayleigh criterion (Peak to Valley wave front error smaller than lambda/4) and at good prices.

I paid 385USD for a 150mm f/10 doublet.

Their arguments were: each objective in cell is delivered with tests, the tests consisting in a star test on an artificial star, a polarimetry test for proving the absence of any internal stress in glass, and an interferometer test for quantifying the optical accuracy of the lens surfaces in terms of peak-to-valley error, wave-front error and roughness RMS.

The interferometer bulletin was given with a Ronchi test evaluation (refer to the upper right picture of the test report – see below).

I am strongly favorable to suppliers giving tests results for a bought scope, this is always more than a trade name reputation often not characterized or overestimated.

This, in combination with the prospective of easy building of the OTA, was the main motivator for acquiring this 150mm f/10 lens doublet.

Test results on the #40 ISTAR lens doublet

Here is the bulletin given with the objective:

The P/V wavefront error is 0.153 lambda (1/6.5), and the RMS error is 0.022 lambda which corresponds to 1/45 lambda). The performance tests were executed with a laser test beam (that is probably a red beam). These data are evaluated from the worst meridian profile where the Peak to Valley error shows the maximum amplitude.

The Ronchi evaluation report gives almost a straight lines pattern on an extra focus plan: seems to present almost no spherical aberration (the curve amplitude of the line near the meridian gives an evaluation of the spherical aberration in terms of a fraction of the inter-wave separation, here probably less than 1/8 wave of spherical aberration).

The doublet passed the polarimetry test: so internal stress level negligible or absent.

The doublet is delivered spaced and assembled in its cell.


I think the P/V data well acceptable with the evaluation tool in use.

I think the RMS (root mean square) number too exaggerated because with the interferometer bulletin shown, the CCD matrix that capture the image represents 255pixs for a 150mm diameter, so 0.6mm of the objective surface for a pix. This is a little too coarse scaling for giving something representative of the actual status of the optical roughness.

If the Peak to Valley ratio may approach the actual value of the optic system, this is rather more exact than the RMS data. If we take into consideration the Marechal criterion (which is a more accurate view than the Rayleight criterion) the P/V 4 front-wave value shall correspond to 1/14th wave for the RMS so a ratio of 3.5 (respectively to be more accurate 1/13.5 and 3.46).

This ratio is important because for having a relation of a Gaussian kind (that represents the dispersion of industrial products, commonly flaws spread uniformly of the considered optical surface), 3.5 shall be, the Marechal criteria being evaluated by calculation with this consideration and not empirically as the Rayleight criterion was. At that quality level, the strehl value is 80%, which means the peak intensity (also evaluated by calculation) is 80% of the peak intensity of a star captured with perfect optics. So that for each P/V value, it corresponds a calculated strehl number.

That means the more actual RMS value for the #40 lenses should be 1/23rd wave (P/V x 3.5).

Taking into account that fact, the Strehl value should be something as 92% (less than the 98% given with the given raw RMS result of the ISTAR bulletin which is 1/45).

Now when excluding some narrow and light portion surfaces of this objective (near the edge), the data can be largely better and especially the P/V. This is improving but that means also the relativity of the numbers given on such bulletin. The shaping of the optic system with the P/V evaluation with the tests is OK being based on a millimeter scaling but not the RMS with this too coarse scaling test.

With these results nevertheless we can conclude the #40 lenses to be at a good to a very good level  that almost reach the Françon criteria (that is P/V < 1/8 wave  and RMS < 1/28 wave with a strehl value of 98% compared to P/V < 1/6.5 wave and the extrapolated RMS < 1/23 wave. The Françon criterion defines a level where when reached a better quality does not bring better sensitive contrasts on the viewed images).

OTA construction, alignment

Some photos here attached in order to show the different phases of the assembly.

Frankly, this is not a great deal for achieving such.

The main thing is to own a pipe with a good geometry. I had an old Newtonian tube from Meade and did the recuperation of this pipe, appearing to be a resin glass pipe well painted white outside and with a black mate paper cover glued inside wall.

The tube is very straight, well circular at the ends and the ends well square to the pipes ends section (all in a 1mm tolerance).

For the ring support of the objective, a duraluminium ring of 5 mm thickness was cut and glued and screwed on a resin ring poured at the pipe end. This is light and enough for supporting the 2,7kgs weight of the objective designed with push-pull screws.

The rings of the tube were the original ones with a 33cm dove-tail plate under the Vixen standard mount.

A simple Crayford focuser 1:1 was recuperated from a SCT and installed on a 2mm thickness aluminum plate as shown. This intermediate plate is important for the alignment because can be leveled in order to be squared to the pipe end and centered to the lens optical center with a laser beam.

The addition of small spacers on each set point of this plate can make it square to the pipe axis with spacers insertion and then after with a simple lateral translation of the Crayford axis in regards to the pipe optical axis to get a laser beam at the objective centered at the opposite end (the laser device insertion into the Crayford tube).

A check of the laser collimator, when turning it around its axis, is important because the trace of the beam at the opposite end can move with the rotation with a mis-collimation of the laser collimator.

Getting these settlements at 1mm less is enough for the alignment of the Crayford, at its end, then the simple collimation with the push-pull screws of the objective will end the settlement of the OTA to an excellent level.

It is possible also when finished to verify the alignment with a cheschire: the central light dot has to be single (the doublet well assembled in its cell) and centered in regards to the faint light circle of the cheschire appearing (close the objective from the external light with a dark cover).

High power used on the sky can allow perfectly the collimation (did at 405x) to be finished perfect.

The hood is fabricated from an aluminum sheet rolled and bolted; the moisture prevention is performed with the protection coming from a flexible SCT hood for moisture prevention.

The tube weighs 9.5kgs with the tube rings, the 50mm finder, the electric focuser (the Orion model for Newtonians) and the moisture prevention, and can be moved by a simple HEQ5 equatorial mount.

However please note that, although this is acceptable for visual observation, any manual action on the focuser will induce significant vibrations due to the inertia of such a long OTA. When observing at high magnification this is a serious nuisance, which is the reason why I installed an electric focuser.

Observations with the refractor

The first actions was to evaluate the objective on the sky with the observation of a magnitude 3 star on April 24th, 2010.

Just 1/10th of a turn on 1 of the 3 push-pull screws of the cell was necessary to meet perfect collimation with 9/10 images near the zenith. The Ronchi eyepiece (10 lp/mm) was inserted and the images confirmed the Ronchi report given on the control bulletin, not more, not less.

The star viewed at 405x was nice with a nice circular Airy disk, the first diffraction ring being concentric and equal in brightness at quite calm times.

Here are some drawings of Mars performed on the nights of April 23rd and 24th, 2010, using the doublet at 295x magnification.

Some details around the tiny NPC, Syrtis Major lowered by “a haze curtain” cloud on the equator surfaces of the disk, Hellas being bright whitish with a higher level than the NPC. Some haze at the limb and at the terminator on the disk.

One of these views is showing the disk without the use of a filter.

Generally speaking, below 2 x D magnification on Mars, the use of minus violet filters is not necessary.

They involve too yellow colored results without significant improvement about the detail perception. More than that fact, what is shown in white light is difficult for interpretation of most of the atmospheric event captured about their nature.

It is preferable to make use of colored filters in order to facilitate this interpretation. This is what is illustrated here through these few documents where in relation with the color of observation, some different surface albedo on details are shown plus the presence or not of atmospheric event on the disk.

Here, the doublet is well adapted for such exhibits where the residual chromatic aberration, still here but at a so tiny level, does not introduce some disturbance about the views captured. At the time of observations the Mars disk was only 7.60 arc-seconds in diameter so that the 6” aperture remains still enough for catching the main features and the cloud effects.

Saturn was also near the local meridian and the views got with 238x rings were very present, a dark dot seen at each anse of the ring as the accessible portion of the Cassini division, the Creep ring was also stunning.

The rings were almost closed but shadows involved on the disk by the rings as a deep dark grey tone elongated area, hemispheres darker than the equatorial zone, the edge of the hemisphere being a double narrow tropical band distinct to the hemisphere area. The images were conducted with the use of a yellow filter W12 and with no filter. The yellow filter (the red better for) is interesting for the research of the presence of clear dots on the disk (not seen at the times). The views obtained without the filter use exhibits, as well with the yellow, sharp views without noticeable residual chromatism in the 6.3mm ED eyepiece (238x), razor edges especially on the rings.

295x views exhibit some purpled colored background sky near the disk but at a confidential level that did not colored purple neither the disk itself nor the rings.

The moon was present at that time also just on the ecliptic. We could have a look on some famous lunar formations as Copernicus, Plato and the Aristarchus area (not lighted yet at the times). Under these light conditions the images at 238x, and 295x more, were light purpled without filter use, the use of the minus violet filter from Lumicon help for annealing the coloration of the views (shadows initially light purpled became black with the filter). All the views were sharp and beautiful; Plato exhibited the main 4 craterlets inside the arena in spite of not the best light conditions. Not sure that at the full moon, the Lumicon filter will be enough.

All these visual results are frankly not bad, rather promising, I think, obtained on the 23rd under bad seeing conditions of 4-5/10 (chromatism is more present under bad seeing) and on the 24th under almost excellent seeing conditions of 7-9/10 (chromatism so confidential).

Comparisons with other OTA

The OTA tubes used in the comparison were:

  1. Intes M615: 150mm f/15 with P/V6 front global scope,
  2. Antares refractor: 152mm f/6.5 with P/V5.5 front,
  3. Celestron C5 SCT: 127mm f/10 “diffraction limited”,
  4. Meade SCT: 305mm f/10 “diffraction limited”.

Here are some views of Mars performed with under similar conditions of sky with each scope.

Intes M615: images are sharp and accurate but at a lesser contrast level, seems more affected by the “instrumental” seeing, the OTA being at the ground level and more subject to the thermal gradients of the atmosphere during the night (more thermal intertie with).

Antares 152mm f/6.5: in spite of the F ratio; the chromatism remains moderate but higher substantially than the ISTAR. The use of filters is mandatory especially the narrow bandwidth ones. Here are some views performed with a narrow bandwidth green filter that helps to show tiny details in spite of the chromatism normally present but cancelled by the filter used.

The disk size was here 4.8 arcseconds.

Celestron C5: in spite of the moderate aperture and the central obstruction presence, it shows interesting features at the time of observation. However at the time of observations here reported (March 21st, 2010, 10), it shows only fuzzy features without shape contour, the NPC is not viewed. However, I think it, better than the diffraction limited level.

Meade 305mm f/10: well collimated. In spite of the aperture and under 5-6/10 images, it does not reveal significantly more details, shows more but not more significantly. The limitation is mainly lead by the sky conditions. Since half a year that I own it, it was used only 3 times and with, only for the best, average images were met here. It will be and will remain almost newbie for years, except the 2 or 3 opportunities in a year. It does not bring advantages under seeing in comparison with a 6-8” aperture, rather a 6” here.

Here is a graph collecting the seeing levels of my personal observing site over the 2009-10 mars opposition period:

Considering these data, the Danjon scale was selected, because a seeing level can be characterized easily.

For a 120mm aperture with 1 arc-second resolving power, images collected on a star of:

  • 5/5 means a level of less or equal to 1/4 of the resolving power so 0.25” max, diffraction pattern is fixed,
  • 4/5 means a level of 1/2 of the resolving power so 0.50” max, the Airy disk being deformed and the 1st diffraction ring start to brake with condensations turning with a speed in regards to the seeing level,
  • 3/5 means a level seeing of 3/4 of the resolving power so 0.75”, the airy disk is hardly deformed and start to brake, the 1st diffraction ring disappears,
  • 2/5 means a level seeing of 4/4 the resolving power so 1”, the airy disk disappeared, the diffraction rings starting to make a halo,
  • 1/5 means a level of 1.5 times the resolving power so 1.5”, the star has a “planetary” aspect with speckles, etc…

The number of speckles grows with the seeing level. This has nothing to do with the FWHM quotation.

That means for stars above 40-45° above the horizon, a seeing quoted 4/5 in a 120mm shall be 2/5 in a 240mm. Not absolutely exact but so many times.

6-7/10 seeing levels are expected here mostly in a 6”, so there is no matter to push away an aperture for a visual use purpose, except for very few nights.

Final status and conclusion

What can be the conclusion about the ISTAR doublet?

Initially and well reported at the CN forums, I owned a first 150mm f/10 doublet which caused some troubles. Here are some sketches for the doublet in cause.

This doublet exhibits some spikes on a diffraction pattern of a star that come from internal stress of the glass. This involved some glare light on a planet disk so that contrast was affected. In spite of this trouble the doublet gave nice images of Mars as shown on the sketch here.

The second doublet well received does not exhibit such flaw and the consequence allied with a good quality level makes it with potentially excellent images even under average to bad seeing conditions.

A 6” such design allied with a good optical accuracy (it is needed more than the Rayleight criteria), with an optic system placed 2 meters above the ground makes at final an excellent ratio efficiency with contrasted images still when the observing site is average to poor, better in comparison with other designs and bigger apertures.

My Istar objective was delivered with test performance results and flawless insurance.


  • hottr6, Jimmy462 and mihaiG like this


Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics