©Lin Yangchen

One of my earliest encounters with the microscope, in the comic strip Ngau-chai (牛仔) by Hong Kong cartoonist Wong Sze-ma. The strip title translates to "cell sketch".

One morning I was sitting in my study when I suddenly wanted a microscope.

I have used state-of-the-art microscopes from the luxury brands—Leica, Olympus, Zeiss, Nikon—in my research. I know what things look like through a 100× plan apochromatic violet-corrected oil immersion Nikon objective, probably the best in the world.

But I was addicted to microscopy and needed a microscope for all times of the day, be it right out of bed at sunrise or straight from the shower. It also had to have the full suite of quantitative polarization capabilities for mineralogy, materials science, art conservation, medical diagnosis and forensic science, with both transmitted and reflected illumination and a trinocular head for photomicrography. And cost less than $1000.

I turned to eBay. And there it was, my dream microscope, on the first page of search results, with free shipping from India.

I had never heard of this microscope maker, Radical Scientific Equipments Pvt. Ltd. They operate from a neat two-storey facility at the 9th milestone along the Ambala-Jagadhri Road, in a landscape of agricultural fields not far from the foothills of the Himalaya. I might well have passed through years ago when I journeyed overland from New Delhi to Manali to climb the Himalayan peak of Shitidhar (5289m).

Ambala Cantonment railway station. Alight here if you're visiting the headquarters of Radical Scientific Equipments. ©Gopal Aggarwal (Creative Commons license)

I gave the mouse a tap, and my new microscope made its way from its birthplace in Ambala through New Delhi-Indira Gandhi International Airport via Guangzhou to Singapore-Changi International and my doorstep in five days.

Radical Scientific Equipments RPL-3T polarizing microscope packed in closed-cell polyethylene foam in a box of total weight 14.2 kg and despatched via FedEx International Priority. According to Amazon, it was first available in 2015. It caters to scientists in developing countries who want to do meaningful research. There is no doubt that one gets the mechanical and optical quality one pays for, but this was the opportunity to find out where exactly the deficiencies are and whether the instrument is adequate for research purposes.

Radical Scientific Equipments RPL-3T polarizing microscope user manual in the 1980s style, showing the eyepieces facing an odd direction. The contents appear to have been typeset in Arial font with Microsoft Word. They are clearly laid out with illustrations in greyscale halftone. The manual contains sufficient details of parts and procedures for assembly and use but assumes familiarity with the general mechanics and operation of research microscopes. There is no quick start guide—you should know what to do.


The base is reasonably hefty but relatively narrow, so the microscope is a little top-heavy and even more so with a camera. It will topple over if you give it a good knock. On the rear of the base is a nice little circular holographic security sticker showing a microscope overlaid with the Radical logo.

Light source

The microscope is supplied with Osram 64250 HLX 6 V 20 W halogen lamps made in Germany. This lamp has a tungsten filament of diameter 0.8 mm that reaches 500 °C in a chamber filled with xenon gas, giving off 460 lumens of photons with a colour temperature of 3350 K and colour rendering index of 100 (something an LED can't match). It has a lifetime of only 100 hours.

I was pleased to see that the filament appeared centered to the collector lens. It is fastened to a metal arm whose position can be adjusted for lamps with different filament positions.

For polarizing microscopy, 20 W is usable but not quite enough. Interference signals are quite dim at 40×. Most research microscopes today run at 30 W to 100 W. A more powerful halogen lamp, however, would require a separate lamp house to prevent overheating. You could put the microscope in a darkroom as they do for fluorescence microscopy in the life sciences.

Massive light wastage.

Actually, there's a very easy way to brighten the lamp. I don't know why they didn't use a reflector. The lamp actually produces lots and lots of light. But it radiates it in all directions, and only a small fraction is captured by the collector lens. Most of the light is wasted. Whatever light that doesn't get absorbed by the completely black bottom panel of the base ends up lighting up the interior of the base like a football stadium!

Olympus BH-2 and SZH microscopes once had the same design deficiency of not having a reflector for the lamp, but it was rectified in a later version of the lamphouse.

Testing an old torchlight reflector using crucible tongs. Mind you, the lamp reaches 500 °C! Radical didn’t incorporate a reflector and didn’t leave enough room for one. There is little space below the lamp and one side is blocked by the positioning plate. I had to jack up the microscope like a car mechanic and pry off the hinged door underneath that was prevented from opening fully by one of the rubber feet.

A reflector can make a huge difference. The reflector in Nikon’s Eclipse LV100N Pol microscope makes its 50 W lamp 40% brighter than a 100 W lamp. But the reflector must be accurately shaped and positioned to aim the light through the collector lens.

Cutting an old plastic reflector with a jeweler's saw strung with diamond wire.

My shiny new reflector affixed with Nano Magic Tape.

I expected this to happen but was hoping it wouldn't. It melted.

Using information from the Zeiss Online Campus and my high school physics textbook (The World of Physics by John Avison, which I still keep in my library), I figured out that the reflector should focus the image near but not on the actual filament to get captured by the collector lens without overheating the lamp. Specifically, to produce an image slightly nearer to the collector lens than the actual filament, the filament must be positioned somewhere between the centre of curvature and the principal focus of the reflector. Judging from online images, the reflector in the Olympus 5-S119 lamp socket, which uses the same lamp, appears to be of the right dimensions.

I ruled out changing the light to an LED. Despite an extensive search of online catalogues, I couldn't find an LED of the desired voltage, luminous flux, colour temperature and colour rendering index of higher than 90 that I could buy in small quantity. Most LEDs yield visibly inferior, undersaturated images because they don’t emit as complete a visible spectrum as halogen lamps and can’t faithfully render the true colours of an object. An LED will also flicker at the frequency of the alternating current, giving rise to banding at certain camera shutter speeds. But it does have a longer life, save electricity and minimize heating.

Of course, you could use an external light source with a mirror, but it is cumbersome to set up and makes the microscope less portable for lugging around, say, to a field research station.

I like this old-school analogue rheostat for the incident illuminator (see below).

There are voltage adjustment knobs for both diascopic and episcopic illuminators but no voltmeters. I assume the manufacturer calibrated the knobs to max out at 6 V. It is better to adjust brightness with neutral density filters, since low voltages may not generate temperatures high enough to sustain the halogen (most likely hydrogen bromide) regenerative cycle that deposits evaporated tungsten back on the filament. But it may help to turn down the voltage if your specimen is overheating or, in extreme cases with more powerful lamps, if the cement holding the lens elements together is melting.

The electrical input is 220 V AC. The fuse has its own little screw port at the back of the base for easy replacement.

Diascopic/transmitted illumination

The field lens is housed in a moulded cylinder with ventilation slots for hot air to escape from the lamp chamber. Apparently in an earlier model the collector lens was recessed into the base, but I suspect it overheated so they modified it.

There is no field diaphragm; it is disappointing how often such a simple but important device for minimizing flare is omitted. I eBayed one from China, placed it on top of the field lens housing and it reduced flare and improved contrast noticeably.

The collector lens is secured at the bottom of the housing by a metal ring with a pig-nose screw drive, which I undid using a military-spec SOG multi-tool. This is dangerous and should not be done on a good microscope. The lens is frosted on one side, an unnecessary feature if the filament is precisely centered and you configure Köhler illumination correctly. Worse, the frosting scatters light all over the place, some of which strays into the optical pathway and reduces image contrast. One benefit of the diffuser, however, is that it spreads out the light at low magnifications where the filament itself is too small to fill the condenser aperture. Unfortunately, because the frosting is on the collector lens itself instead of a separate filter, it cannot be removed.

Diascopic polarizer

The polarizers in this and most microscopes are dichroic polarizers that transmit light in one plane of polarization and absorb the rest.

The diascopic polarizer mount has a pig-nose drive inside a snake-eye drive. One of them probably screws the filter into the holder while the other rotates it for calibration. But both of them are immobilized with glue (visible in the photo).

The diascopic polarizer assembly wastes lots of precious vertical space. The filter, scale, swing-out arm and knurled ring for rotating the filter are separate components that are spread out way too much vertically. There is even an extra knurled ring that serves no apparent function other than as an excessively thick spacer. Furthermore, there is an unnecessary vertical gap between the polarizer and condenser. Luckily, there is enough space to focus the condenser for Köhler illumination.

The filter rotates continuously through 360° in a smooth and lubricated manner. The etched degree scale is graduated at 5° intervals and hard to read as it is in a dark place and there isn’t enough contrast with the shiny chrome finish. The filter has a greyish-blue cast that with the yellow halogen lamp gives a pale green cast in the image when not crossed, and presumably affects interference colours to some degree.


This is an Abbe condenser, an 1870 design consisting of two lenses, a plano-convex and a bi-convex, that leave spherical and chromatic aberrations uncorrected. It has a numerical aperture of 1.25, two centering screws and a flimsy plastic swing-out holder for 32 mm filters.

The condenser is of the older design that slots upwards into a ring instead of sitting down in it, and is held in place by nothing more than a screw with a tantalizingly large knurled knob that encourages you to turn it, upon which the condenser will fall out and crash into the polarizer below, or if the polarizer is swung out, into the field lens farther down.

The iris diaphragm has battered blades that make it look as if a squirrel chewed a hole out of it.

The condenser focusing knob sticks out farther than both the coarse and fine focusing knobs despite being the least used. Sometimes I accidentally turn it when I want fine focusing instead.


The rotating circular stage, radius 70 mm, has centering screws and a goniometer graduated in degrees with no click stops. The vernier scale ends slightly short of the ninth degree, making it completely useless. Moreover, as you tighten the screw to immobilize the stage at the desired azimuth, the clockwise torsional force of the screw on the goniometer makes it go off a little anti-clockwise.

The ball bearing assembly.

The stage has a threaded hole and a pair of guide holes spaced the standard 35 mm apart for adding an x-y mechanical slide holder.

As I was moving a microscope slide around, I discovered that the stage wasn’t flat! The slide rocks from side to side. The central stage insert bulges slightly, an artifact probably generated during manufacture when the hole was being punched in it. You can remove the insert, leaving a shallow circular depresson, and let the ends of a standard microscope slide straddle the depression.

The insert has a diameter of about 64 mm. Fellow microscopists have suggested that the insert could be exchanged for one without a hole to prevent light from reflecting off the substage condenser during epi-illumination when the specimen isn't completely opaque, or that a glass insert could be used for unmounted specimens. After removing the insert, the condenser hole in the middle of the stage becomes only slightly larger, but it has a screw thread for attaching as-yet unknown devices from either above or below the stage.

Checking if the stage is level … just in case.


Coarse and fine focusing are controlled by completely separate knobs, like in an early-20th-century microscope. Coarse focusing is done with a rack-and-pinion mechanism. Fine focusing is done with a screw-lever mechanism. The left-side fine focus knob is graduated in case you wish to do manual focus stacking or something. The fine focus knob doesn't immediately catch the mechanism when you turn it either way, but this small focus lag is apparent and bothersome only at 40×.

But there’s a bigger surprise. The coarse and fine focus knobs work in different directions! You have to turn the two knobs opposite ways to move the stage in the same direction.

Furthermore, the coarse and fine focus mechanisms have their own travel limits that are completely oblivious to each other, so if you hit the end of one of them before your specimen comes into focus, you have to switch to the other one and remember to turn it the other way which is not how all other microscopes work. If you're on a high-powered objective and run out of fine focus, you have to go down to a lower objective to overshoot safely on coarse focus, go back up to high power and try again. When you’re trying to focus (pun intended) on your research, it’s an unnecessary irritant that makes you cry.

The coarse focus has an adjustable upper travel limit pin to avoid driving the slide into the objective.

With the supplied objectives (see below), a slide-mounted specimen comes into focus only when the stage is very near its bottom limit. This leaves very little space to manoeuvre specimens and makes it impossible to bring into focus specimens thicker than a microscope slide for episcopic observation.

I managed to increase the clearance by 9 mm by removing a retaining screw near the bottom of the rack-and-pinion assembly. But if the rack is lowered all the way, the pinion reaches the very end of it and rolls ever so slightly off the end of the rack, requiring extra force on the focusing knob to bring it fully back on the rack. Try not to go there. I used a slice of cork to stop it.

You can get another 1.5 mm above the stage by removing the stage insert, but at the expense of substage space if you maintain Köhler illumination.

The microscope’s rather thin arm seems rigid enough but actually flexes several micrometers when you press down on it. That's enough to make it go out of focus. Be careful with high-powered immersion objectives with very short working distances, although the 40× is not in danger.


The revolving turret has four objective mounts. A ribbed rubber ring provides grip. The indexing mechanism is a metal tab at the back with a protruding bump that sits into the grooves in the turret circumference. The drawback of this external mechanism is that when you turn the nosepiece, your fingers get pinched by the metal flap. It can be quite painful. The benefit is that it is simple and easy to repair.

The nosepiece doesn't have the centering screws for each objective that are found on higher-end polarizing microscopes. This means you have to keep re-centering the stage when you change objectives if you want to do quantitative polarization work, unless you use objectives that can center themselves.


The unbranded objectives are achromatic, but only semi-plan, 4/0.10, 10/0.25 and 40/0.65. The Amazon listing of the microscope says they are strain-free. They are not parfocal but not too far off. Only the 40× objective has a parfocality adjustment ring, which can be reached by unscrewing the outer casing.

The objectives conform to the DIN standard with a parfocal distance of 45 mm, and Royal Microscopical Society standards for mechanical tube length (160 mm) and mounting thread (diameter of 20.32 mm and pitch of 0.706 mm; ISO 8038, DIN 58888). Since the DIN parfocal distance is the largest of the major standards and this instrument is designed with enough space for that, most objectives for RMS tube length and thread will fit regardless of parfocal distance.

Actually, the supplied objectives are wrong. After happily stacking on the epi-illuminator and analyzer assembly (see below), the manufacturer didn't realize or didn't bother that the mechanical tube length had become much greater than 160 mm and would cause spherical aberrations. Since the illumination system and optical components aren't that good anyway, it doesn't really matter except in one situation I describe later with a high-end objective I installed myself.

The lens assembly looks misaligned. The rear aperture (not shown) is plastic. The annotations on the casing are printed, not engraved, in unappetizing word-processing sans serif font with inconsistent kerning and excessive letterspacing. The casing spans the entire length of the objective, so you invariably end up unscrewing the casing from the objective when you really want to unscrew the objective from the nosepiece. You have to unscrew each objective twice. Unbelievable.

These objectives are no good for epi-illumination as they are optimized for coverslips (0.17 mm) and are not epi-rated, which means they probably lack adequate anti-reflection coatings.

Radical says all the optical components have an “anti-fungus, anti-reflection and hard coating”. Possible manufacturers of the optics include Seiwa Optical (headquarters and assembly lines in Japan) and Nanjing Jiangnan Novel Optics (China) which also makes some Nikon optics, according to discussions on the MicrobeHunter Microscopy Forum.

I tested a Nikon CF N PlanApo 20×/0.75 on the microscope. With thin sections it was alright, but with thicker sections such as paper in transmitted light, the out-of-focus parts of the sample flared up badly. Experiments revealed that the microscope's analyzer assembly and half-mirror assembly were the culprits; the flare disappeared when I connected the head directly to the turret. The excessive mechanical tube length is partly to blame. I have long used plan apochromatic objectives to examine paper samples in transmitted light with no problems.

Optical train wreck. The fat column of light from the plan apochromatic 20/0.75 is scattered by the undersized and scratched half-mirror of the epi-illuminator and by the poorly constructed analyzer assembly.

Episcopic/reflected/incident illumination

The incident illuminator brings you back to the pre-Köhler days of the 19th century, when the filament could be seen in the image of the specimen, although it is out of focus in this microscope. The lamphouse has a few narrow slots cut into its thick casing instead of a proper heat sink with lots of surface area, and gets almost too hot to touch. Although there are two lenses to direct the light, there is no variable-aperture diaphragm. When the light floods through the aforementioned inappropriate objectives you get a terribly washed-out image.

The mechanism for centering the episcopic illuminator is more complicated and strenuous than it needs to be, possibly because this design gives more room for imprecision in construction. The up-down adjustment is controlled by a ring that has to be centered in two dimensions, just like the stage and condenser. There is too much friction in the assembly so the small centering screws are ineffective and you have to grab the ring and move it with brute force. Next, for the left-right adjustment, you push in or pull out the lamp housing and lock it in position with another screw.

The half-mirror is too narrow and has scratches, fingerprints and glue on it that are visible in the optical path. The mirror chamber is powder-coated white. What you get is a disco party of stray light and loss of image contrast.

The drop-in filter slot of the episcopic illuminator resembles those found on Nikon supertelephoto lenses. But the supplied generic filters have only a protruding stem for handling, instead of a tab that extends out from the tube to seal the opening and immobilize the filter. The filters match the diameter of the slot but once in there’s a cavity for them to roll, pitch and yaw. It also lets in dust. It's usable but the shoddiness drives me crazy.

Sanding down the metal edges of a Chinese-made iris diaphragm for the epi-illuminator using French-made electro-coated 220-grit silicon carbide abrasive paper to make it fit in the drop-in filter holder.

Episcopic polarizer

The rotating assembly of the episcopic polarizer is very loose-fitting along the optical axis but heavy grease reduces the rattling. It looks more like heavy machinery than precision instrumentation. The in-out slot of the polarizer itself fits snugly within the rotating assembly. The mark indicating the azimuth is just a line scratched by hand into the powder coating of the illuminator tube! The line isn't even straight and is misaligned with the degree scale so the polarizer annoyingly rotates from −1° to 89°.

Wave plates

A gypsum plate gives full-wave retardation at 560 nm, while a mica plate gives quarter-wave retardation at 145 nm. The former fits tightly into the slot and clicks into position, while the latter fits loosely, with rattling, and doesn’t click despite looking exactly identical to the former. The slot opening is of the German DIN standard of 20 mm by 6 mm.

I later tested the mica plate using my calibrated Olympus BHSP and found the crystal more than 10° off alignment. Fortunately it is user-adjustable, being fastened with a snake-eye drive instead of glued down. I eventually converted it to a de Sénarmont compensator.

The metal bar holding the compensating plate isn't of the better-designed long type with a second, empty hole. When you pull the bar out to deactivate the compensating plate, the bar remains slotted but leaves an opening on the other side that lets dust and stray light into the tube.

The direction of the slot was slightly off 45° and I had to adjust it as well as I could with a primitive plastic protractor.


It rotates about 100°. The scale ring is made of aluminium, for some reason not the chrome-plated metal of the polarizer scales. The vernier scale ends at about 9.5, rendering itself completely useless. The polarization cross is not oriented north-south-east-west, which can be adjusted, but the polarizer and analyzer scales are off by several degrees with respect to each other.

You push a pin into the tube to move the analyzer into the optical path, and pull out the pin to remove it. That is intuitive enough. But it’s the opposite for the incident polarizer—you pull out the pin to activate the polarizer and push it in to deactivate it. I don’t need this additional confusion that, like the coarse and fine focus knobs turning in opposite directions, could have been avoided in design. To annoy you further, the incident polarizer automatically slides into the optical path by gravity if the pin is pointing downwards.

The circular dovetail of the analyzer assembly is too small for its bracket and can't be centered precisely. There are no markings such as red dots to align it at the correct azimuth, so you have to use your eyes to estimate by comparing the parallelism of the Bertrand lens slider and the epi-illuminator, which could itself be out of orthogonal alignment, or by using an awkwardly placed protractor to establish the perpendicularity of Bertrand lens slider and microscope arm, the latter of which slopes along the y-axis.

Bertrand lens

The Bertrand lens is centerable but its focus can't be adjusted. I had to clean the lens as it was coated with a layer of dirt. It's an off-center glue job that gives a slightly distorted image of the back focal plane of the objective. Metal shavings are still present.

I flipped the slider upside down so the circular aperture the Bertrand lens rests on now lies between the lens and the eyepiece, acting as a baffle to reduce stray light.

The springs holding the slider in position are very tight and you have to brace the microscope while pushing or pulling the lens in or out, otherwise the microscope will topple!

Trinocular head

It’s rotatable 360° but you have to loosen three fiddly hex nuts to rotate it. Again the circular dovetail connecting it to the analyzer assembly below is too small. You have to try to center it by visual estimation, not something you should be doing with the optical train. If you mis-center this, the analyzer assembly and the epi-illuminator all in the same direction, you'll get something resembling the Leaning Tower of Pisa.

Wherever setscrews are used to secure dovetails and connections in the microscope, the unsmoothed ends of the setscrews scrape out metal slivers from the tube as you tighten them. When I disassemble and assemble the joints, I have to use a vacuum cleaner to suck up these fragments!

The Siedentopf head, originally made by Zeiss, is a nice surprise since the microscope was advertised with a sliding head for the binocular eyepieces. The Siedentopf head allows the interpupillary distance to be adjusted about a single axis, which keeps the tube length constant and avoids altering focus. The eyepiece tubes are of the standard diameter of 23.2 mm and are fixed at a 45° incline, which is steeper and less ergonomic than the 30° of the Olympus BH2. The left-hand tube has a diopter adjustment ring graduated from −6 to +6, but no marking on the tube to read off the value.

The camera port has a standard 25.4 mm C-mount thread that connects to a tube with a standard 23.2 mm port. The inside of the tube is darkened but still reflects stray light.


Two pairs of wide-field eyepieces came in the box although the user manual lists three, so they shortchanged me a pair. One eyepiece has a 19 mm ERMA (Tokyo, Japan) micrometer graticule with 100 divisions, while another has a crosshair graticule for centering. Why didn't they just use a dual crosshair-micrometer graticule like in most other polarizing microscopes?

The graticules are secured with screw-in rings with snake-eye drives, making it difficult to install or remove the graticules. The eyepieces that came without graticules don’t have a graticule holder at all. There is no pin on the eyepiece assembly or notch in the head tube to orientate and lock the crosshairs in the compass directions for optical mineralogy, so you have to use adhesive tape or something.

The eyepieces are of the non-compensating, Kellner type. The design comprises two plano-convex lenses one of which is an achromatized doublet, a modification of the Ramsden eyepiece. The lenses don't seem to be coated. The inner collar is plastic and you won't know until you dismantle the eyepiece.

The eye point isn't high enough to easily see the full field of view with glasses on. Together with the underpowered light source, you feel nostalgic, as if you’re squinting through a microscope built in the 18th century. There is no rubber eyecup or rubber padding, just bare metal which will scratch your glasses if you’re not careful. If you view without glasses, you’ll soil the lens with your eyelash oil. It’s a lose-lose situation.


When you're done playing, the manufacturer provides a dust cover of woven nylon, like that used in camping tents. The fabric has tiny pores that let the instrument "breathe" while keeping dust out. In this regard I consider it better than the nonporous PVC ones provided by the big brands, although I’m not sure about anti-static properties. My lab is in the humid tropics and has no air conditioning, so ventilation is important. Another problem with PVC covers is that they have been softened with plasticizers that will leach out with age, leaving a greasy residue on your hands and on the microscope.

This instrument manages to produce usable images without falling apart. A geologist and a soil scientist have testified on Amazon that it suffices as a home stand-in when they are away from their labs. But I realized how many things I took for granted in a microscope, such as a flat stage. It may not be for cutting-edge research institutions, given its mechanical and optical limitations, lack of local customer service and unanswered emails. But it has a full set of specialized features for a price that will only get you a basic student brightfield microscope from other vendors. It's really not bad.

But ultimately, it's an amateur microscope. It has too many mechanical and optical deficiencies that will compromise the productivity, accuracy and reproducibility of serious research. Even after a month of troubleshooting, I was still finding hidden flaws. Finally, I decided it was beyond salvation, and turned to a true research workhorse with an impeccable track record: the Olympus BHSP.
Powered by SmugMug Owner Log In