Bifilar micrometer by Sir Howard Grubb c1895
This Grubb bifilar micrometer is unlike that described in Grubb trade catalogues 1888, 1899, & 1903. The body is brass and phosphor bronze, the scales German silver. The micrometer screws hardened steel.
GRUBB BIFILAR MICROMETER c1895
I re-webbed it using 10 micron quartz filaments in November 1994. The webbing arrangement is shown below:
there are two position webs, theta 0 & theta 1, three fixed separation webs, s0, s2 & s3. s0 is at comb0, s2 is at following rev30, s3 is at preceding rev30. There are two following separation webs f0 & f1 separated by 25.2revs, f1 is inclined 12 divs to f0 @ theta 0. There is a single preceding web p0 moveable between excursion limits comb0 & comb p rev30. It is not possible for the p & f webs to cross over.
Each revolution of the micrometer drum moves the p or f web 1/50-inch (50tpi micrometer screws) or 0".020. Each tooth on the comb corresponds to one drum rev. There are small holes drilled in the comb at 10, 20 & 30 revs for ease of tooth counting. Drum revs are counted off along the comb. One div on the drum equates to 1/100th of 1/50th inch or 0".0002, or 1/5000-inch.
The micrometer screws are held in female collets which can be adjusted to regulate backlash. Backlash is measured with the drum reading increasing and decreasing to determine the zero error. The drums are a friction fit on the spindles and can be adjusted to read zero (100 in this case) when the p & f webs lie at comb0.
The quartz filaments subtend 10 microns or 0".0004, which equates to 2 drum divs.
The webs are extremely delicate and vulnerable to damage when the eccentric eyepiece carrier is removed from the shift plate.
The eyepiece is held in an eccentric carrier, screwed into a shift plate that can be slid from side to side. This enables separations to be measured wider than the eyepiece field of view. This is also why there are 3 fixed sep webs, from which sep measures can be made.
The micrometer screw box carries two forks to which are fixed the p & f webs. Each fork is fitted with a precision ground and lapped female screw bush and a spring loaded trap plate which is spring loaded. Backlash is adjusted by setting the clearance between the trap plate and the female bush. Note if you will the f fork is spot marked with pairs of punch marks to prevent incorrect assembly.
The sep webs are fixed to the back of the cover plate which also carries the comb:
The cover plate is fitted with tin shims and the p & f forks are pressed into the bottom of the screw box by the blued steel springs riveted to them.
The micrometer screw box is mounted on a translation stage driven by a box screw, to enable the micrometer to be adjusted off axis.
The webs are illuminated by red LED's housed in the translation box:
I introduced this illumination system at the time of re-webbing.
The micrometer has been adapted to fit a position circle:
whose scale reads to 30"arc by opposing verniers:
To use the micrometer the observer must first determine the screw constant Ro. An approximate value of Ro may be obtained by determining the ratio of the micrometer screw pitch and the objective focal length. For example suppose the OG focal length is 980mm (38".583), then Ro = 5.1837E-04, and converting to arcsecs, 106".92arc per rev. The limiting accuracy of a single measure is the filament width (2 divs), which equates to 2".14arc. This can be reduced by taking the mean of many independently repeated measures. The screw constant can be refined by measuring the declination separations of known stars, and each determination corrected for temperature.
The PA scale is zeroed by allowing a star to drift in RA, and the position web adjusted using the position circle so the star runs along it. Seps and PA's can then be made with respect to the intersection of the s0 sep web and either the theta0 or theta1 position webs. The north PA is the position circle reading ±90°.
The micrometer originally had three simple Ramsden eyepieces, one is missing. The remaining pair are 1" and 5/8" focal length. I provided a low power plate magnifier 23/8" focal length for a low power acquisition field.
This particular micrometer was originally supplied for use with the 1894 Wilfred Hall 15-inch f/12 Grubb Astrograph, and intended to measure the relative position of a comet or asteroid with respect to background stars.
And what of the bifilar micrometers described in the aforementioned Grubb catalogues? The 1888 catalogue depicts that owned by Armagh observatory. Their 1899 & 1903 catalogues use the same engraving.
That owned by Manchester Astronomical Society, part of the Godlee Observatory is similar:
Both have similar drum heads and thumb wheels but the shift plate is screw driven, not free to slide side to side, neither has a box screw, and both have windows for revolution indicator scales. Grubb describes the bifilar micrometer as being supplied with 5 eyepieces, as is shown in the Armagh observatory model.
The photographs below show my Grubb bifilar micrometer and Position Circle mounted on my TEC140APO. The weight of the micrometer plus Position Circle is 10lbsf, which is within the load capacity of the FT3545 rackmount.
The brasswork incidentally is not lacquered. It is protected by a micro-crystalline wax, known in the trade as Renaissance Wax, after polishing with PreLim. <http://www.restorationproduct.com/preliminfo.html> Renaissance Wax is supplied by Picreator. <http://en.wikipedia.org/wiki/Renaissance_Wax>
The restoration process entails removing the patina of oxidized lacquer with successive grades of emery paper from 600 thru 2500 (600; 800; 1200; 1500; 2000; 2500), used dry, and then buffed with wadding. (The alternative is to soak in Methylated spirit overnight). The components are then washed in warm water and a mild soap solution, and dried prior to polishing with Prelim, and then waxing. It is important to polish out all sleeks from the previous emery grade before proceeding to the next. When polishing off the patina, the emery paper is laid on an old newspaper, which in turn rests on a heavy, dense piece of flat timber. The component's sides are polished against the flat sheet of emery paper. Epicyclic motions similar to those used to grind a lens or mirror, are employed. Pressure is applied evenly to the workpiece, and the process continued until all trace of patina has been polished away. The same applies to all the edges. The workpiece is never placed on the work surface, and the emery paper torn into pads and used for localised scrubbing. Keep the emery paper whole, laid down flat, and polish the workpiece against it. The grade is done once the surfaces are smooth, matt or shiny, uniformly all over. Sleeks must be random, and definitely not polished in any particular direction. The top sheet of newspaper is discarded before starting the next finer grade. The final 2500 grade will leave a sleek free mirror finish. PreLim is less abrasive than 2500 emery, and is applied with a fine cotton cloth, or soft paper, using small circular strokes on large flat surfaces, and linear strokes on straight narrow parts, until a uniform glass like finish is achieved. Intricate, and steeply curved surfaces are polished using strips of emery paper wrapped tightly around a needle file, before buffing. Plates and eyepiece caps, dust caps &c maybe patterned using 2500 grade emery paper tightly wrapped around a flat needle file, applying the edge of the file against a straight edge, drawing across the surface away from the body, and repeating line by line until one obtains the desired pattern, for example:
These micrometers were only ever made to order. A skilled instrument mechanic would have made the micrometer and probably the Ramsden eyepieces, perhaps aided by an apprentice. The mahogany case made to match afterwards by a skilled cabinet maker. It would have taken weeks to make the micrometer, each part being a one off. The basic components of the micrometer screw box are cast brass, the screw housings phosphor bronze, the precision micrometer screws & bushes, case hardened precision ground steel. The micrometer drums are brass, and the engraved scales German silver (a copper-zinc-nickel alloy, Cu60%-Zn20%-Ni20%). German silver was used for precision scales because it was readily cut with an engraving tool, and was resistant to wear. It tarnishes like silver, and needs periodic polishing. Real silver alloy scales would wear down rapidly. The drum divisions would have been engraved using a dividing head.
The rest of the box is brass, sheet and flat stock, each piece profiled by hand filing. All the separate parts only fit one way, symmetric parts are not reversible. Many are pop marked to identify the correct way to assemble them. Circular parts with several fasteners on a common pitch circle diameter are not equispaced, although they may appear so at first glance; one hole is intentionally slightly offset making it impossible to reassemble the part other than the intended way. All the screws are hand finished; they may have been turned on a Capstan lathe, in bulk, but each screw was finished by hand, and the clearance and tapped holes done by hand. Different sized screws are used to assemble each sub-assembly. It is not possible to use one set of screws intended for a particular sub-assembly on a different sub-assembly. The screw threads are all Whitworth threadform, although neither BSW or BSF.
It may not appear to be but the most difficult part to make for this style of micrometer is the comb:
which is used to count off the screw revolutions. It is a thin plate whose edge is a section of the micrometer screw thread that is fastened onto the cover plate, into recesses, so that it's surface contacts the fixed sep webs which in turn just clear the moving p & f webs carried on the forks. It has to be quite thin, otherwise both sets of webs and the teeth would not lie within the depth of focus of the objective. Given that the refractor this particular micrometer was made for had a depth of focus 0".0127, and the webs are 0".0004, the comb would have to be no thicker than 0".0118 allowing 0".0001 web clearance. The comb was made by turning a brass bar held between centres, and a facsimile of the precision 50tpi micrometer screw cut with a single point tool. A section was then milled out along the middle of the bar, and then 1/100" holes drilled at ten tooth intervals, 1 @ 10, 2 @ 20, & 3 @ 30 teeth. The centre of the comb had to be located with sufficient accuracy so that a fixed sep web bisecting the 1/50" zero hole was precisely on the mechanical axis of the micrometer (I have pushed the So web to one side so the zero hole in the comb can be seen clearly).
To work on these instruments requires painstaking, consummate skill, and the right tools, jeweller's screwdrivers, ring and C spanners. The slots in the screws are very fine. If a screwdriver with too thick a blade is forced into the screw slot, the screw head will be damaged, and part of it may split away. The screwdriver blade must be the same width as the screw head, otherwise there will be insufficient purchase. On no account should the screwdriver blade be wider than the screw head, otherwise the component surface will get scored. Note in the photograph of the eccentric eyepiece carrier, that the screw slots are unmarked. The correct type of screwdriver has been used to remove and then refasten the carrier, in this instance a large jeweller's flat bladed screwdriver with a precision ground and case hardened blade. If a screw is jammed, ease with penetrating oil ['PlusGas'], and try gently tapping the end of the screwdriver with a pin hammer. Apply a light mineral oil to each screw using a cotton bud before screwing into the tapped hole. This will ensure ease of assembly, and prevent it jamming in future.
Originally my Grubb micrometer was webbed with 8 micron platinum wire, which nowadays is almost impossible to buy. I purchased a 10m reel of 10 micron nickel chromed wire from The Scientific Wire Co. London
Most bifilar micrometers are smaller, and intended to measure close doubles. This type of bifilar micrometer was intended to make measurements of the motions of asteroids and comets with respect to background clock stars, prior to the widespread introduction of photographic plate measurement. The Wilfred Hall Astrograph had a 15-inch f/6 Petzval camera which took 15-inch square plates covering ~10°x10°. Although it was feasible to measure many asteroids on a single plate, it must be born in mind that plates were very slow, and blue sensitive. Exposure times were measured in hours, and by the time a pair of plates had been taken, processed, dried, measured and reduced, similarly accurate measures and reductions could have been made using the bifilar micrometer on the 15-inch f/12 refractor. Judging by the lack of wear to the micrometer screws and bushes, it is evident this particular bifilar micrometer was little used.
This page was created by SimpleText2Html 1.0.2 on 09-July-2012