NavList:

I have recently acquired a Kollsman MA-1 Model 1972B-02 aircraft sextant. This is the "pendulous mirror" version (not the bubble chamber version). This was acquired from an "antique co-op", so there was no information available regarding its operational state or service history. I was able to confirm a reasonable image through the index prism - eyepiece light path, but without 28V power, I couldn't check the artificial horizon image path before the sale.

Once home, I set the eyepiece focus to optimize sharpness of a distant object and noted that the index reticle was also sharp. After confirming continuity of the horizon reticle lamp and rheostat, power was applied. It was apparent that the horizon reticle was illuminated and the mirror was motion responsive (too much so?), but the AH reticle focus with my eye at the eyepiece was VERY poor (pretty much just a blurry orange blob). The eyepiece focus was swept over its range with no improvement to the AH focus, but with obvious degradation of the distant object and index reticle focus.

Reviewing the illustrations in the MA-1 Overhaul manual, removal of the (horizon) mirror unit from the bottom of the sextant seemed the least invasive way to make an initial diagnostic assesment. Once removed, looking up through the body casting showed that the horizon reticle was intact, uniformly illuminated, and presented a sharp image to the naked eye (only the pelicle beam splitter remains in the light path). A small tear at the edge of the pelicle and significant wrinkling were apparent. The wrinkles would be expected to degrade the image reflected from the mirror to the eyepiece but not the transmitted image from the reticle to the mirror (consistent with my observation looking up from the bottom).

The mirror unit was inspected and there was no obvious signs of leakage. The unit was tilted to all angles and no evidence of gas bubble was found. The position of the bellows was consistent with the fill procedure in the overhaul manual (ie: bottom of #3-48 tread is 1/2" from bottom of unit). At this point I assumed the unit to be properly filled. I did not consider that the unit could be completely empty.

I removed the left side cover and removed the pelicle /frame. The film was indeed torn, so it was scraped from the frame. The pelicle measured .0004" thick. I made a temporary fix with plastic food wrap that also measured .0004" thick. I did not use adhesive, just the "cling" from the stretched film.

Installing the now flat pelicle and reinstalling the mirror unit did not produce any notable improvement to the AH focus when viewed with eye at the eyepiece. Moving the mirror unit through it full position range failed to find an improvement.

At this point it was noticed that looking into the eyepiece from about 18" away, with my corrective (for myopia) glasses on produced a sharp image of the AH reticle! Without the (-4D) glasses, the reticle was sharp at about 10". It was also found that the sextant produced a sharp projected real image to a white card placed 4" downstream from the eyepiece. This is all interesting (and promising) because it would appear to rule out surface condition of the optical elements (ie: scatter and reflections) and gross errors in axial alignment.

Comparing a photo in Morris' Kollsman pelicle replacement (with a microscope cover slip) blog showed that my horizon reticle was positioned about 4mm further up than in his picture. I proceded to add this adjustment to my focus search matrix to no avail. Note the Overhaul manual states to adjust the reticle position to optimize the tilt tracking of the mirror and then to follow with the mirror unit focus. These were of insufficient mechanical range to reach an acceptable focus result.

Given that the normal range of element positions couldn't fix the focus, the other mathematical parameter is focal length of the mirror unit! There are only a few unique elements in the AH light path, and one is the lens at the top of the mirror unit. Its hard to imagine how this could be wrong, but its specs are unknown. Light passes from the AH reticle (ignoring the pelicle) through the lens, through the damping fluid, off the mirror, back through the damping fluid and lens again. Thus the "unfolded" mirror unit can be modeled as a positive lens of index of refraction N.glass followed immediately (no gap) by a negative "lens" of N.fluid with thickness equal to twice the lens to mirror gap, followed by another lens (identical to the 1st lens, but flipped end for end) of N.glass. Approximate thickness and distances in the mirror unit were scaled from "Figure 3-1 Optical System" of the Operating Instructions for the sextant (drawing scale is approximately 1:2).

At this point an experiment was performed making a surrogate "mirror unit" from a dentist mirror and a roughly 2" focal length lens from an eyeloupe. The lens was fastened 1/2" above the mirror and this assembly hand-held outside the mirror unit port in the sextant body. It was quickly proven that a sharply focused image of the AH reticle could be produced in the Index reticle plane (thus in focus of the eyepiece). The relatively long focal length of the chosen test lens insured that the conjugate lengths would occur outside of the sextant body.

The next question was determining the effective focal length of the mirror unit in its current condition. A microscope slide having one end frosted (made so it can be written on for ID purposes) was obtained. A line about 1/4" long was written on the clear portion of the slide with black Sharpy marker, just beyond the frosted portion. The slide was held over the mirror unit in such a manner that the line on the slide was visible as well as the reflected image of the line coming back from the mirror unit. An overhead lamp provided illumination. The reflected image was projected onto the "ground glass" part of the slide. A scale was used to measure the distance from slide to mirror unit lens when the line target and its reflection were both at equal length (ie: magnification of 1:1). This distance turned out to be ~ 15mm.

Measurements scaled from Fig. 3-1 yield approximate distances of ~49mm from Ah reticle to mirror input lens and 23mm from mirror output to Index reticle. The lens is ~ 6mm thick and the distance from lens to mirror is ~5mm (10mm for total reflected path). The Overhaul manual addendum for the 1972B-02 states that the fluid is DC200 (silicone oil) with viscosity of 7 cSt. Data from Shinetsu gives index of refaction of ~ 1.399 for similar material. I took a wild guess that the lens would be something like N-BK7 glass.

The "unfolded" optical path was modeled in WinLens3D per the distances above and the surface radii of the glass lens tweaked until the object distance and image distance matched the finite conjugate solution (and the actual sextant geometry). The resulting lens has a effective focal length of 17.2mm and the silicone fluid "lens" has an effective focal length of -15.92mm.

With the lens properties established, the simulation can be rerun for a goal of 1:1 magnification to mimic the microscope slide practical test, with a result of object = image distance of ~ 32mm. Note that this required distance is about 2X what was previously measured!

A final run of the simulation with the index of refraction of the fluid set to 1.0 (ie: air) yields a 1:1 mag object distance of 15.2mm which amazingly matches the practical measurement! So the "smoking gun" is very likely that ALL of the fluid in the mirror unit has leaked/evaporated and this looks remarkably the same to the naked eye as a full unit (ie: no bubbles). A possible give away is the (lack of) damping behavior of the mirror. Eyeball estimate of the natural frequency of the mirror oscillation is 2 - 4 Hz and rings for at least several (5-10?) seconds. If anyone has a similar functional sextant, or past experience with one, and can confirm if their unit behaves differently, I would very much appreciate a reply.

Of course I'm reluctant to break the seals on the mirror unit. Mostly that's because I don't have replacement fluid on hand and the refill procedure involves submerging in fluid, pulling vacuum in a bell jar for several hours, installing the plugs while still submerged...etc.

Comments and suggestions welcome.