CD Sextant Experiments

17 Feb 17

In February 2005 I saved a copy of this page by Omar Reis showing how to make a sextant using a compact disc, some Lego blocks, and two mirrors. I did nothing with it at the time. Now, wanting to experience the challenges of taking sextant sun sights without horizon shades, using a pinhole sight vane, and making sights with an artificial horizon, the CD 'sextant' looked like a quick and dirty way to try them..

I had a few bits of mirror and clear glass stowed away in a drawer in the basement from long-forgotten optics experiments. I found a 20 mm x 35 mm mirror from which I'd removed a 5 mm diameter patch of silvering (how I did it, or why, now long forgotten). I chose this to be the starting point of a horizon mirror. Using a pocket knife at first, then a single-edged razor blade in a holder, I removed the backing from a rectangular patch of the mirror. With the razor blade, and soaks in bleach, I cleared away the 'silvering' so that a 20 mm x 15 mm section of clear glass remained.

18 Feb 17  Morning

I found a selection of Lego blocks in a tool box my son had used as a youngster, an unopened CD war game I'd been given ages ago, and another mirror that I taped off with blue painters' tape to use as a 31 mm x 37 mm index mirror. I assembled these pieces using double-sided cellophane tape to make what's shown in Figures 1. and 2. below.

For sun shades in all initial experiments I used slip-in neutral gray shades (provided to glasses-wearers after eye dilation at the opthamologist's office). A stack of five of these was clipped directly to the index mirror with clothes pins as shown in Figure 4.

Note: throughout these experiments the observed body was the sun. There was no intent or attempt to use the assembled pieces to make actual altitude measurements.

Figure 1.  The horizon mirror is installed at a 45-degree angle to the CD case edges as was done in the link given
                above.  The 2 x 1 Lego block is a 'knob' for turning the CD. As in the original, there is no sighting device.

Figure 2.  The back side, showing the blocks that the index mirror is taped to.

After some fiddling, I got the pieces oriented so that I could see the horizon and the sun together and dashed off to get
a camera to attempt a photo as shown in the final image of Reis' page. I couldn't recreate the conditions. It seemed
that having the horizon mirror at 45 degrees to the sight line was a contributor to the problem. (The original, though
called a sextant, is really a quadrant.)

18 Feb 17  Afternoon

I reoriented the horizon mirror, rotating it 15 degrees CW as viewed from above, so that it made a 60 degree
angle with the case 'lower' edge.

I added a pin-hole sight of card stock, with two holes, one vertically aligned with the center of the clear portion
of the horizon glass, the other with bottom edge of the clear section.

Figure 3.  The 'sextant' with a card stock sight vane taped in place.

Figure 4.  Stack of 5 slip-in sun shades clipped directly to the index mirror with clothes pins.

Though having the index shades directly atop the index mirror is the wrong thing to do, this combination worked reasonably
well at first. There was evidence of two reflections though: the desired, filtered one reflected from the mirror after two passes
through the stack of shades, and off to the right of it, another, larger, brighter, more indistinct image, reflected off the surface
of the stack. This behavior became much worse when I made the change shown in Figure 6.

19 Feb 17 

  Figure 5.  With the direct and reflected images vertically aligned on a ridge five miles away, the separation between
                  the rulers shows that the index and horizon mirrors are not parallel. The pinholes have been enlarged
                  to a diameter of 5/64 inch and blackened with a felt tip pen.

20 Feb 17 

 Figure 6.  As in Figure 4. but with the shade pieces taped together and masked off at the approximate size of the
                 index mirror. This configuration was nearly useless as the index mirror with shade stack acted like a
                 shiny piece of black glass.

Figure 7.  Replaced the card stock sight vane with one made  of aluminum. The pinholes, as before, are 5/64 inch
                 in diameter. In the background at left are four rectangular pieces cut from the unaltered slip-in shades
                 shown in earlier figures.

21 Feb 17 

Attempts at taking sun sights with an artificial horizon

Figure 8.  Four thicknesses of shade material clipped to the forward side of the sight vane for an artificial horizon
                 sun sight attempt.


Figure 9.  0745 EST, 18 deg F, 'sextant' and artificial horizon (frying pan filled with water) in the

This didn't work out. I could see the artificial horizon through either pinhole, but couldn't find and bring down the sun.

Next steps: enlarge the pinholes and try different shade material.

The author of the article linked at the beginning of this page used dark sections from 35 mm film negatives (two thicknesses
for the index mirror shade).  I went through my envelopes of 35 mm prints and negatives and found about 24 total inches
of this material. I did a quick check looking at the sun through two layers of this material and found it roughly equivalent (based
on eye discomfort) to the five layers of neutral gray shade material I used in the above figures. The film material is not neutral
gray. It tends to give a pinkish cast.

22 Feb 17

More artificial horizon tries

I remounted the sextant parts on a thicker CD case to gain a bit more stiffness (in a really flexible construction).

I removed the tape masking part of the index mirror and installed the mirror with the larger dimension vertical.

I enlarged the pinhole diameters from 5/64 inch to 7/64 inch. This nearly doubled the area of the apertures.  After a trial, I
enlarged the lower pinhole to 1/8 inch diameter and taped off the upper pinhole.

I changed the pinhole shades from four thicknesses of the neutral gray material to two thicknesses of the dark 35 mm film.

With these changes in place, the result was no different from the day before: I could see the artificial horizon but couldn't
get the sun visible through the pinhole. I noticed that I was getting a rectangular reflection in the driveway to the right of
the artificial horizon pan. I tried making things align by pushing on the horizon glass--and knocked it off and onto the

Later in the day, while pondering the monster I'd created, I dropped it on the hardwood floor, scattering pieces all over.

Figure 10.  Reconfigured layout for more attempts at an artificial horizon sun sight. Two pieces of dark film are
                   paper-clipped to the sight vane.

Having reassembled the pieces earlier, at quarter to midnight I got motivated to make another attempt early on the 23rd.
I cut out a paper arc marked with radial lines at 10-degree intervals (but labeled 0, 20, 40, etc. to account for the double
reflection) and taped it to the disc.
I looked up solar altitudes versus time for the 23rd, and put a pencil mark at about 37
degrees, the approximate double altitude (2 x 18.4 deg) for an artificial horizon sun sight at 0830 EST, my hoped-for trial time.

I sighted a distant object, aligned its direct and reflected images, and put reference marks on the disc and the CD case to
mark this condition. (I aligned the 0 degree line to these marks when I attached the paper arc).

While set to 0, I replaced the Lego 'knob' with a plastic spirit level, with its axis set parallel to the 'bottom' of the CD case.
The level would serve as a knob and provide for another experimental observation. If I could bring the reflected sun down to
align with the direct image, the bubble in the level should be approximately centered.

23 Feb 17

Artificial horizon finale

The first attempt of the day was successful. A sight with the water-in-pan artificial horizon was taken at 0826 EST. At this time the
solar altitude was 17.8 degrees giving a double altitude of 35.6 degrees. The 'index arm' (disc) was within a hair of being set to the
mark at an estimated 37 degrees that I'd made the previous night.

It is not possible to use this contrivance alone and check the spirit level at the instant the direct and reflected images coincide.
Each time I looked away from the sight vane to check the level, the sextant needed to be tipped slightly nose down to center the

I made a second successful sight at 0940 EST when the solar altitude was 19.9 degrees (double altitude 39.8 degrees).

 Figure 11.  The setup for February 23rd. The shade material is not installed on the sight vane.

 Figure 12.  After a successful sun sight with artificial horizon at 0826 EST.

Index shades - take two

After the morning success with artificial horizon sights, I returned to the issue of index shades. All previous experiments were
done with the shade material clipped to the index mirror.

I suspended four thicknesses of neutral gray shade material between the index and horizon mirrors using a strip of aluminum,
three Lego pieces and double-sided tape. The shade plane was approximately perpendicular to a line joining the centers
of the two mirrors.

This worked. The problems encountered with the configurations used between 18 and 20 February were eliminated.

 Figure 13.  Index shade of four thicknesses of neutral gray material taped together and suspended
                   between the index and horizon mirrors by an aluminum support and Lego pieces.

25 Feb 17

This morning the sun was clear of clouds long enough to make several trial sights between 0740 and 0750 EST for which
the solar altitude at my location ranged from 10.9 degrees to 13.4 degrees..

The photograph in Figure 14, taken with the camera pointed in the direction of the sun, illustrates the conditions. Puddles from
the snow that's melted in the past week is reflecting the sun. (I'd prefer to have had today's sun with the snow cover on the 19th
as a greater glare challenge.)

I used the reflection in Figure 14 that appears as a sliver of light near the mid-field tree line as the horizon. (The larger, nearer,
puddle has eight Canada Geese clustered around it.)

I made sights with four shade conditions. All used an index shade comprising 4 layers of neutral gray material. The variations

1. No horizon shade

2. One thickness of neutral gray material as a shade at the horizon mirror (see Figure. 15 for illustration)

3. One thickness of neutral gray material as a shade at the sight vane pinhole (as shown in Figure 10 with different shade material)

4. Essentially no shades by retaining the index shades but pointing far enough off to the right that the index mirror reflection came
    off the clear glass, not the mirrored, portion of the horizon mirror


1. There was considerable glare off the horizon glass, and might have been useable for a real measurement--but only because my
    'horizon' contrasted with the darker field around it and was brighter than the glare.

2. This was good.

3. This was better than 2. because the sight vane shade not only provided horizon shade but also added to the index shades, making
    viewing more comfortable

4. This might have been useable in a pinch, but the sun reflection off the horizon glass is much less distinct than when reflected
   from the horizon mirror itself. (I tried something like this on the 23rd, pointing just off the sun in the early afternoon with a
   clear blue sky. Under this condition the reflection off the glass part of the horizon mirror was well defined against the sky.)

 Figure 14. Looking in the direction of the sun at about 0745 EST (solar altitude at 0745: 11.7 degrees).

 Figure 15.  I added a gray 2 x 2 Lego brick atop the horizon mirror and slipped a piece of neutral gray shade
                  material between it and the horizon mirror as a horizon shade.

Make a free website with Yola