©Lin Yangchen

[scroll to bottom for full specification]

So, after struggling for a month with the endless mechanical and optical gotchas hidden behind the impressive-looking specs of my new Radical RPL-3T polarizing microscope, I threw down the screwdrivers, sandpaper and super glue and decided to find another microscope.

I narrowed my choices to research-grade models from the 1980s as they were fine instruments, had the Soviet-era industrial look and were affordable second-hand. I settled on the Olympus BHSP for its design, quality, modularity, versatility and track record. The microscopy community holds it in very high regard.

I knew there won’t be no gotchas this time.

The problem was finding one. The standard-issue BH-2 is common. But the BHSP, a full-blown petrographic microscope with all the quantitative optical mineralogy fittings, supersized circular rotating stage with a vernier goniometer that can measure angles to a fraction of a degree, and a lamp five times more powerful, is very rare, especially in complete, working condition. Veteran geologist Gregory McHone’s “magnificent” BHSP was put (back) together from bits and pieces that took years to salvage.

I went on eBay to see if any parts were on the market. There were bits and pieces, most of them visibly in need of repair. But one listing caught my eye. An Olympus BHSP, virtually complete and in “good working condition”, was sitting in Shanghai, its two big eyes looking at me. It was missing the power cord, but my kitchen kettle had the same cord.

The seller was asking a little over 10% of Olympus’ price in 1992, accounting for inflation. There were seven watchers. I quickly checked the seller’s credentials, assessed the risk and analyzed the costs and benefits, and in an hour I had my conclusion: grab it before it’s too late. I clicked Buy It Now.

Two Boeing 767s flew the microscope from Shanghai Pudong International Airport to Shenzhen and onwards to Singapore-Changi International. Maps and flight profiles from Flightradar24.

I didn't know the microscope was made of iron!

It took a long time to extricate the microscope parts which had been disassembled into separate boxes inside the main box. Each part was wrapped in waterproof plastic film and completely immobilized in its own custom-shaped cocoon of closed-cell foam. This microscope was very lucky. Many get damaged en route to their new owners.

The microscope had been given a good cleaning and a new bulb. I plugged my kettle cord into it, flipped the switch and it came to life. I pushed the slider all the way to 12 volts and the 100-watt bulb shone like the sun. The weak signals from my spherulites under crossed polars looked as bright as day. No more 20-watt student microscopes, thank god. I was expecting to do some maintenance, but when I went through all the knobs and pins and diaphragms and sliders and rotating scales, I realized I didn’t have to fix anything.

It was hard to believe a research-grade BHSP was standing in front of me, and it came from China.

But superlative as it was, it was just a standard BHSP, not yet the unique turbocharged instrument with my personal touch of engineering.

For one thing, it is really only half a microscope with just transmitted light. To investigate the full spectrum of microscopic phenomena the world has to offer, one needs both transmitted and reflected light. I chose the BH2-MA illuminator with its black "particle accelerator" tunnel bristling with rotary valve controls and exposed filter slots, over the fuller-featured but boxy and ugly plastic-cased BH2-UMA. The BH2-MA mirror can be swung out when not in use, allowing transmitted-light observation at full brightness.

I found one from a Chinese electronics dealer but it looked strange. It was marked as BH2-MA-2 and the extra-long tube lens assembly had two optical elements or groups instead of one. I could find no record of this variant in any of the Olympus brochures or repair manuals. I began to wonder if it was a top-secret Chinese military knock-off, since it had the same colour as the SR-71 Blackbird and the B-2 stealth bomber. Whatever it was, I reasoned that the extra tube length shouldn’t matter since it was all Infinity Inside (better than “Intel Inside”) and labeled with the usual “f = 180”.

Complete disassembly of the BH2-MA-2 for cleaning and anti-reflective flocking.

Adjusting the filter slot with a spirit level.

The field diaphragm.

The epi-illumination system is a convoluted contraption. After going through ten suppliers from five countries for esoteric parts, with multiple lost shipments and cancelled orders, I finally put it all together, flipped the switch on the massive TGH transformer, and reflected light shone through the microscope.

Full power on both illumination systems. The TGH, epitome of the industrial era, weighs a few kilograms and has a big scalloped knob that you can easily grab and turn without taking your eyes off the eyepieces, although best practice is to maintain the lamp's rated voltage.

It took me some time to work out the “dream team” of objectives for the turret. Olympus recommended using the infinity-corrected metallurgical objectives on BH-2-series microscopes with epi-illuminators, as the epi-illuminator converts the microscope from finite tube length to infinity-corrected. However, these objectives show evident strain in polarized light, and the high-powered ones can’t be used with coverslips.

Some users take off the epi-illuminator and swap out the turret for another one with the strain-free finite objectives when they need transmitted polarized light, but I decided I couldn't be doing that every time I needed to switch mode. The analyzer assembly, superwide trinocular head (see below) and camera weigh several kilograms and have to be manœuvred carefully into position after removing the epi-illuminator, like docking the Apollo Lunar Module to the Command Module, with some lens surfaces exposed and at risk of being smashed by the dovetails. This is not how a research-grade instrument should work.

For the epi-illuminator to stay on permanently, the objectives must be infinity-corrected and RMS-threaded, since the BHSP centering nosepiece is only available in RMS thread. Across the four available spaces on the turret, I wanted strain-free glass and transmitted and reflected light with and without coverslips at both low and high magnifications, but not necessarily on the same objectives.

I realized that BH-2-era objectives couldn't do everything I wanted in just four objectives. So I took a look at Olympus' latest UIS2 optical system and discovered an obscure series of objectives designated UPLFLN-P. They are the only ones rated "excellent" for polarized light in Olympus' entire lineup and had stellar credentials overall. I got on the line to Olympus.

Upgrading a Boeing 747 with state-of-the-art avionics. Unfortunately, these objectives are ugly. The glorious days of engraved technical lettering are gone. Today, even the best plan apochromats have their specifications printed on in Arial or Helvetica that seems to have been typed out in Microsoft Word. If you look closely, you can even see the digital aliasing! And because printing is so cheap and easy, the objective barrels are cluttered with unnecessary and repetitive text. To make it worse, Olympus’ latest lacquered bronze finish looks completely out of place on the BHSP.

Fortunately, there is an easy way to fix the problem: gaffer tape with textured matt black finish (see epi-illuminator photo above). It’s not all that inappropriate, since the black barrel is the traditional identifier of a strain-free objective. Best of all, they now look like SAS commandos.

To transform the BHSP from a Nissan to a Rolls-Royce, one must swap out the 20 mm “widefield” eyepieces for the truly widefield 26.5 mm SWHK. McHone says "the view is like flying over my rock section surface in a small airplane". It's like watching an IMAX film through the microscope.

The eyepieces need the rare superwide trinocular head, which I found refurbished from someone with a Russian-sounding name living in a peaceful wooded suburb near the shores of Lake Ontario in the United States. He told me he’d dusted off all the prisms, cleaned and relubricated the binocular dovetails, trinocular slider rails and internal tension springs with polycarbonate roller, and checked the head for image alignment. It was put on a flight that took off from one of my favourite airports, the eight-runway Chicago O'Hare.

WALL-E? The normal and superwide heads side by side. Everybody goes for the sleek Siedentopf binoculars nowadays; my Jentzsch slider looks like a tank and works just as well with its automatic tube length compensation.

The SWH superwide eyepiece (left), patented in 1993 and meant for BX-series microscopes, alongside the SWHK (right). The two are practically identical in field of view on the BH-2 head but the SWH has a higher eyepoint and is a little brighter. The SWHK has a little more green-purple separation, but the difference is so small I could just make it out only in high-relief specimens. The SWH seems a tiny bit warmer in colour temperature. The rotation locking pin of the SWH is on a different circumference than the slot on the BH-2 head tube, so the orientation of the eyepiece can't be locked unless some modification is made or the eyepiece is stuck down with tape.

The retro asymmetric pentagonal side profile (left) with no parallel or equal edges is unique in the world. Its avant-garde geometry is derived from the functional layout of the internal optomechanics—a marvel of industrial design, perched near the top of the optical skyscraper with a dizzying overhang. The "vintage Lotus supercar" has evolved over the decades into a "BMW SUV" like the Leica DM750P (right). Also notice the differences in the Bertrand lens, analyzer and waveplate.

But the SWHK has a shortcoming that took some trouble to fix. It doesn’t have a graticule holder, and Olympus didn’t make graticules big enough for it. The world’s widest field of view is all very well but you need crosshairs, micrometers, austenite grain sizing discs and such things for serious work.

In its graticule specification sheet, Olympus instructed users who needed a graticule to buy a special eyepiece similar to the regular SWHK but whose tube was split into two sections with a 24 mm photo mask graticule inside, and to get the graticule professionally changed to one of their choice.

Actually, you can hack into the regular SWHK eyepiece, fit a graticule and be a happy man, although it’s more complicated than simply unscrewing a holder at the bottom. You have to get in from the top by pulling out the plastic rim, loosening three tiny set screws without losing them and unscrewing the focusing helicoid. Then you have to carefully lower and rotate the graticule by eye until the crosshairs are aligned with the slot-screw on the outside of the tube, and glue the graticule to the top of the front focal plane aperture shelf with three dabs of rubber cement.

But there was another problem. A supersized 28 mm graticule was next to impossible to find ready-made with a cross-scale pattern. I also wanted the numerals on the scale in a technical lettering typeface. The Chinese love Times New Roman on their graticules, whose narrow letters with thin strokes and serifs make it hard to read. Meanwhile, a European supplier told me their default was Arial, which is too heavy and obstructs sample detail. Besides, these fonts remind me of Microsoft Word. Desperate, I wrote to two companies that offered a custom graticule design and fabrication service and they each quoted more than $1000.

Then I found Shibuya Optical Co., whose Japanese-English website had hitherto eluded DuckDuckGo. It was like a huge Japanese fish market, with eyepiece graticules in every size and pattern. They actually list the SWHK as one of the eyepieces they cater for. When I wrote in, they said they had stopped making the 28 mm cross-scale, but they offered to make it specially for me. The price was much lower than that for custom fabrication, since they already had the template. I immediately ordered two. It happened that Japan’s Obon festival was round the corner, so I hoped the graticules would be blessed by the ancestral spirits.

The graticules came in classic scalloped twist-cases with intricate raised technical lettering. I was surprised to see “Olympus” instead of “Shibuya” on the cover, as I'd been told by a distributor that Olympus made its own graticules. Anyway, it was a perfect match for my SWHK. I was so close to not finding this graticule.

Removing hardened grease with 99.5% acetone, regreasing the helicoids with MicroLubrol Helimax-XP synthetic low-volatility polyalphaolefin with lithium 12-hydroxy-stearate and polytetrafluoroethylene, and affixing the made-to-special-order graticule with E6000 styrene,1,3-butadiene polymer cement with tetrachloroethylene solvent.

I set the crosshairs in the compass directions by eye, in the spirit of a true master craftsman. This is one of the very few SWHK eyepieces in the world to be fitted with a petrographic-grade cross-scale graticule.

Precise rotational alignment of the stage, polarizers and eyepiece graticule crosshairs to the cardinal directions using an obscure cult accessory, the orientation plate. Most people don't need the precision but I wanted my BHSP to be truly petrographic-grade. I called up several distributors of the Big Four looking for it but no even knew what I was talking about. Finally I found the part number in a BX polarizing microscope user manual and was able to put an order through my Olympus supplier.

Calibrating the graticule. Thank god Shibuya didn't use Arial or Times New Roman for the numbers, but the proper technical lettering typeface. The field curvature is from the camera.

This OB-M stage micrometer (centre), kindly loaned by my local Olympus supplier, bears Olympus' classic logo. Its accuracy was verified by laser interferometry at Singapore's National Metrology Centre.

Stacking optical sections for extended depth of focus using the Stack Focuser plugin in ImageJ and generating the first calibrated elevation map from my BHSP. The stage was lowered in 2 μm steps by manually turning the fine focus knob which I had calibrated with the broken edge of a coverslip measured with a micrometer screw gauge. Nowadays you can control a motorized microscope using computer software without ever touching the instrument or even looking through the eyepieces. This is good for high-throughput research and reproducibility but all the fun is gone.

The rotating stage is nice, but its full potential isn’t realised without a mechanical slide holder that not only gives you sub-millimeter translational control but also ensures the slide is immobilized while rotating the stage for angle measurements. I also use the vernier scales to record the exact coordinates of objects of interest on the slide to find them again later. Most petrographic microscopes, even those in research labs, have no mechanical slide holders while most regular microscopes do (but can’t rotate). You can have fine control over either rotation or translation but not both.

Olympus (as did others) made a special mechanical slide holder with top-mounted knobs for rotating stages. It is much rarer than the rectangular ones, and even more so being an optional accessory. I found only one for sale on the whole of the Internet, and it was in perfect condition, waiting for a new home on a BHSP. I bought it.

Ironically, one of the smallest parts of the microscope was also one of the most interesting and time-consuming.

Olympus furnished a pair of wrenches for the centering screws on the stage and nosepiece. It was a simple but clever design of a slotted screwdriver bit encased in a cylindrical sleeve to prevent the wrench from dropping off the microscope when you take your fingers off it.

People didn’t appreciate these things, and most are now lost. The unique design makes it almost impossible to find replacements. You could make do with a regular screwdriver and put up with some inconvenience, but I wanted to savour the elegance of the original design.

In 2021, Larry McDavid and Wayne Moorehead designed and precision-machined limited numbers of replacement wrenches. But I didn’t have the tools for cutting hardened steel, so I improvised two less-glamorous but working wrenches.

Materials for my first homemade centering wrench. It takes only a few minutes to make. You need a sufficiently rigid tube of inner diameter no smaller than 3 mm and outer diameter no greater than 4.7 mm, a short slotted screwdriver whose diameter is between 3 mm and the inner diameter of the tube, and, if necessary, a strip of office paper to roll around the shaft to get a tight fit with the tube and center the screwdriver bit in the tube. Twist the tube as you push it in hard. No glue is needed. Paper has just the right compressibility and friction for the job.

The perfect tubing was found in a pack of frozen Shishamo.

The microscope looks like a triad boss smoking a gold-plated cigarette. The tip of the slotted bit should be recessed about 2.7 mm into the tube (inset).

I made the second wrench out of a power screwdriver bit that cost about $1 apiece. The hex shank makes a good finger grip.

One of the last things to go in was my specially designed and machined "experimental nuclear physics" swing-out filter holder with better functionality, ergonomics and mechanical robustness than the Olympus original which proved impossible to find. With two filters of different neutral densities I get four brightness levels without having to dim the lamp below its rated voltage and risk upsetting its tungsten-halogen cycle.

The Burj Olympus/ Large Photon Collider/ Saturn V/ Soviet Mir space station.

After three months of retrofitting and modifications with components made in the United States, Germany, China, Japan and Singapore, the microscope finally reached full specification at a gross weight of about 30 kilograms. The stacking of the analyzer assembly and the extended epi-illuminator takes it to three quarters of a metre in height—possibly the tallest BH2-series microscope in the world.

McHone regards petrographic microscopes as “humankind's greatest expression of art and science”. Not least in the Olympus BHSP, with its glass-and-metal optomechanical components mounted on the utilitarian die-cast superstructure instead of being smothered by the sweeping plastic curves on the new designs from Leica to Nikon to Zeiss to the Eurostar. The BHSP's khaki, black and silver colour scheme has a timeless aura that sets it apart from the bright white finish popular in the latest microscopes, of which those by Leica and Zeiss are accentuated by loud red and blue marketing stripes. A microscope isn't just a tool but also a marvelous engineering showpiece that nerdifies my living space, and I want it to embody my engineering and architectural ideals.

Olympus BH-2 series
technical documentation and repair manuals

Olympus BHSP
brochure | user manual

‘Special Forces’ Olympus bhsp microscope



high-performance transformer with stabilized voltage output
sliding voltage control lever
red LED voltmeter
circuit breaker
field diaphragm
45 mm filter mount on field lens housing
adjustable stage block height

swing-out filter holder with two trays
304 stainless steel with CNC machined finish
designed by Lin Yangchen
manufactured by Rapid Direct, Shenzhen, China

earthed power cord
Line Tek LS-60 ICE 60320 C13 10 A 250 V female connector
Line Tek LP-61L BS-1363/A SS 145/A British and Singapore standard three-pin plug
made in Taiwan

Diascopic illumination

BHS-LSH lamphouse
B2-CLS aspherical collector lens
infrared filter
45 mm filter slot

ABO898 45 mm filter holder
DB-80 clear daylight blue filter 45 mm

Philips 7724 12 V 100 W GY6.35 tungsten halogen projection lamp
luminous flux 2800 lm
colour temperature 3300 K
colour rendering index 100
life to 50% failures 2000 h
flat filament 5 × 3.1 mm
quartz bulb
temperature 900 °C

Nikon ND0.6 neutral density filter 45 mm
Nikon ND8 neutral density filter 45 mm

Episcopic illumination

SoundTech FC-80 220–110 V converter 80 W with thermal fuse

TGH transformer
LED voltage indicator
input 100/115 V 0.7/0.6 A 50–60 Hz AC
output 12 V DC 50 W
7 A fuse

BH2-MLSH lamphouse
built-in heat-absorbing filter
lamp centering knob

Osram 64610 HLX 12 V 50 W G6.35 tungsten halogen lamp
xenon filler gas
luminous flux 1600 lm
colour temperature 3350 K
colour rendering index 100
lifespan 50 h
filament length 3.30 mm
filament ⌀ 1.6 mm

BH2-MA Mark II epi-illuminator
swing-out half-mirror
tube lens (f = 180 mm) for infinity-corrected objectives
aperture diaphragm
field diaphragm

filter holder ⌀ 20 mm for 6.5 mm slot
nylon PA12 Hewlett Packard Multi Jet Fusion 3D print dyed black
designed by Lin Yangchen
manufactured by 3D Print Singapore

filter holder ⌀ 20 mm for 3.9 mm slot
nylon PA12 Hewlett Packard Multi Jet Fusion 3D print dyed black
designed by Lin Yangchen
manufactured by 3D Print Singapore

LBD daylight filter

Edmund Optics linear polarizing filter ⌀ 20.00 mm
dichroic polymer film
Schott B270 low-iron crown glass
fire-polished surface
thickness 2.00 mm
spectral range 400 nm–700 nm
transmission 30%
polarization efficiency 95%
extinction ratio 100:1

Substage condenser

BH2-POC strain-free Abbe condenser
focal length 7.2 mm–28.8 mm
NA 0.9–0.25
lever-actuated iris diaphragm with graduated scale
for 4×–100× objectives
swing-up top lens
centering screws
built-in glass-mounted polyvinyl alcohol polymer film linear polarizer
polarizer rotatable 360° and graduated in 5° increments
adjustable 0° click stop


BH2-SRP circular 360° rotating stage 170 mm with brake
goniometer with vernier reading to 0.1°
lever-activated 45° click stops starting at any position
centering screws
stage insert with 22 mm aperture
holes for stage clips and mechanical stage

centerable condenser holder with rack-and-pinion focusing

mechanical stage AH-FMP
working range 30 mm × 30 mm with top-mount controls
vernier scale to 0.1 mm
spring-loaded slide holder

Centering wrenches

Lin Yangchen Mark I
3 mm slotted screwdriver with cross-hatch knurled handle
plastic frozen fish packaging tube cut to length
rolled paper strip for friction fit

Lin Yangchen Mark II
Broppe 3 mm slotted power screwdriver bit with hex shank (S2 high-impact steel)
plastic frozen fish packaging tube cut to length
rolled paper strip for friction fit


coaxial coarse and fine adjustment
tension adjustment ring for coarse focus knob
calibrated fine focus knob graduated in 2 μm increments
upper travel limit lever


BH2-PRE quadruple nosepiece
three holes with centering screws
RMS objective thread (⌀ 20.32 mm, pitch 0.706 mm; ISO 8038, DIN 58888)

UPLFLN plan semi-apochromat (fluorite) 4×/0.13 Pol ∞/- f = 180 mm
UIS2 Optical System with wavefront aberration control
working distance 17.0 mm
super wide flat field (field number 26.5)
parfocal distance 45 mm (DIN)
back focal plane position −1.3 mm
resolution 2.58 μm
RMS thread (⌀ 20.32 mm, pitch 0.706 mm; ISO 8038, DIN 58888)
European Union RoHS-compliant
glass free of lead, cadmium and arsenic
metal parts free of hexavalent chromium

UPLFLN plan semi-apochromat (fluorite) 10×/0.3 Pol ∞/- f = 180 mm
UIS2 Optical System with wavefront aberration control
working distance 10.0 mm
super wide flat field (field number 26.5)
parfocal distance 45 mm (DIN)
back focal plane position −19.1 mm
resolution 1.12 μm
RMS thread (⌀ 20.32 mm, pitch 0.706 mm; ISO 8038, DIN 58888)
European Union RoHS-compliant
glass free of lead, cadmium and arsenic
metal parts free of hexavalent chromium

UPLFLN plan semi-apochromat (fluorite) 40×/0.75 Pol ∞/0.17 f = 180 mm
UIS2 Optical System with wavefront aberration control
working distance 0.51 mm
super wide flat field (field number 26.5)
parfocal distance 45 mm (DIN)
back focal plane position −19.1 mm
resolution 0.45 μm
RMS thread (⌀ 20.32 mm, pitch 0.706 mm; ISO 8038, DIN 58888)
European Union RoHS-compliant
glass free of lead, cadmium and arsenic
metal parts free of hexavalent chromium

MSPlan achromat metallurgical objective 50×/0.80 ∞/- f = 180 mm
working distance 0.47 mm
super wide flat field (field number 26.5)
parfocal distance 45 mm (DIN)
RMS thread (⌀ 20.32 mm, pitch 0.706 mm; ISO 8038, DIN 58888)

Analyzer assembly

magnification 1×
slide-out analyzer graduated in 2° increments with vernier scale reading to 0.1°
rotatable 180° with locking pin
20 mm × 6 mm (DIN) compensator slot
focusable Bertrand lens in rotating turret

AH-TP530-2 λ retardation plate (530 μm)


BH2-SWTR30 super widefield trinocular head
rotatable 360°
Jentzsch binocular with fixed 30° incline
automatic tube length compensation for interpupillary distance slider (56 mm–75 mm)
30 mm eyepiece tubes
3-step light path selector pin (100% eyepieces, 100% camera, 20/80 eyepieces/camera)
high-transmission coated glass prism
camera port with 38 mm dovetail


SWHK 10× L
super widefield 26.5
eyepoint 15.6 mm
focal length 25.00 mm
focusing front lens
tube ⌀ 30 mm

calibrated Shibuya Optical Co. 28 mm eyepiece graticule R1401-28
Schott B-270 low-dispersion soda-lime crown glass, thickness 1.5 mm
ultraviolet to near-infrared transmittance
crosshairs with horizontal scale
100 divisions in 10 mm, accuracy ± 20 μm
line width 10 μm
pitch 0.1 mm, accuracy ± 2 μm

U-PJ orientation plate


NFK2.5× compensating projection eyepiece
low distortion
field number 21.6
focal length 50.42 mm

parfocal photo tube (anodized aluminium)
adapter 38 mm dovetail to 28 mm thread
RISE (UK) 28–42 mm step-up ring
adapter T-mount to Nikon bayonet F-mount

Novoflex adapter Nikon–Canon lens mount
exact compensation of difference in flange focal lengths
made in Germany

Canon EOS-1D X
35 mm-format full-frame 18-megapixel complementary metal–oxide–semiconductor sensor
mirror lockup
1080p high-definition video

Canon RS-80N3 cable release

The Sarcophagus.
Powered by SmugMug Owner Log In