How Bright is your Light Part III: LED Conversion

ledsfigure1 incr vibrance
LED Spectra (from Zeiss site)

In Part I of this series, we talked about a simple, home-grown fix for a defective, very basic illumination power supply on one of the most classic of classic microscopes, the Nikon S.  Then in Part II, we moved on to repairing or replacing the more sophisticated supplies  for incandescent lights on modern scopes from about 1975 onward, and met the ubiquitous switch mode power supply, or SMPS.  In this segment, we will review the ultimate, and perhaps best, solution for repairing or updating illumination on a classic microscope:  LED conversion.

Rather than rebuild an older incandescent light source, the most expedient solution may be to upgrade it to LED illumination that provides bright, white-light illumination with a durable system using little current, producing minimal heat, and using no expensive, hard-to-find replacement bulbs.  You can buy a readymade LED illuminator for some microscopes.  For many classic scopes, however, there are no off-the-shelf solutions, and you are faced with modifying the illumination system yourself.   This not necessarily a bad thing – in some cases, the conversion can be done fairly cheaply and easily using pre-built power supplies or a simple battery supply.


Many, if not most, modern microscopes have LED illumination systems.  Aftermarket LED illumination kits or adapters designed for a number of specific microscopes have recently become available.  A Google search for microscope LED illuminators and light sources brings up a variety of retrofit kits and illuminators, a number of which are generic, of Chinese origin, and may or may not be a good fit to specific models.  The most sophisticated yet cost-effective manufacturer of LED retrofit illuminators for older scopes seems to be retroDIODE LLC of Gardner, Kansas.  This manufacturer uses 3D printing to manufacture LED units specific for a number of the most popular classic scopes; this is their unit for the Nikon S-Kt:SKt 02SKt 01At $140.00 US on retroDIODE’s eBay store, this seems to be a reliable and reasonably-priced alternative to building an LED illuminator from scratch.  Promicra also makes some LED illumination systems, but no price is given, and the cost is probably significant.  ThorLabs manufactures very sophisticated microscope light sources in a considerable variety of well-characterized spectra for several major brands of scope, but the price is about $450-600 US, and they are only made for fairly modern microscopes.

Thorlabs Mid-Power_LED_for Zeiss
Thorlabs Mid-Power_LED_for Zeiss

Microscopes with simple, non-Kohler illumination systems are much less of a challenge, as there are a number of simple LED plate-type substage illuminators available on eBay for $20-65:

Simple Chinese LED Plate-style Illuminator - $65 on eBay
Simple Chinese LED Plate-style Illuminator – $65 on eBay

Similarily, eBay carries a multiplicity of ring LED illuminators for stereomicroscopes at very reasonable prices.

The main advantage of buying a commercial LED conversion kit is that you don’t have to mess with deciding on the best LED, choosing a power supply, using tools, tearing out the old system, or making the new system.  However, LED conversion kits are not yet available for all vintage microscopes.  Furthermore, with some ingenuity, a conversion can be done for very little money, and you then have the satisfaction of understanding your equipment better (and deciding what to do with all the money you saved).


Before converting to an LED system, one should really understand what a light emitting diode is and how it works:

The first emission of light from a crystal junction was recorded at Marconi Laboratories in 1909-7 by H.J. Round, using a gallium arsenide crystal with a “cat’s whisker” wire contact, much like a 1920s crystal radio.  The first true light emitting diode was created

(and its theory correctly described using Einstein’s new quantum theory) by the Russian scientist Oleg Lesov in 1927.   Lesov, largely unknown despite being one of the first semiconductor physicists, studied solid-state light emission and actually used primitive semiconductors successfully in amplifiers and radios.  However, in the age of

Oleg Losev, 1903-1942

vacuum tubes, the work of this unsung genius was largely ignored.  Caught in the siege of Leningrad in 1941, he tried unsuccessfully to get a paper on a three terminal semiconductor (possibly the first transistor) out of the besieged city, but the manuscript was lost, and he starved to death in 1942, along with a million other residents of the city.  Lesov spent much of his life working as a lowly technician, and his work was forgotten for twenty years.  Nikolay Zheludev and Tom Simonite have published fascinating notes on Lesov and the original development of the LED.

Siege of Leningrad

Most sources unfamiliar with Lesov’s work credit the invention of the LED to the independent efforts of four American research groups in 1962 (see Wikipedia article); subsequent intensive work by many groups led to the first commercially successful LEDs in the 1970s, and there has been an enormous volume of research and innovation since that time.

Simply explained, a light emitting diode consists of a junction between an N-type semiconductor material (with an excess of electrons in the conduction band) and a P-type semiconductor (with an excess of “holes”, or electron deficiencies, in the lower-energy valence levels).  Application of a current across the junction forces electrons and holes to cross the junctional energy gap.  The electrons and holes jump from higher to lower energy states, and the lost energy is released:

Energy levels across p-n junction in LED (from Wikipedia)
Energy levels across p-n junction in LED (from Wikipedia)

HOWEVER, not all such transitions, and not all semiconductor types, produce light.  Asking “…But…but…why?…” leads the reader into arcane theory and dark places that are better left alone.  Anyone who lacks a strong stomach and a knack for dealing with quantum entities that don’t really exist but still can go bump under the stairs, is advised to take two Dramamine, sign off the computer, and lie down in a quiet place.  The less wary will learn that the energy drop across the junction is referred to as the band gap;  gaps are classed as direct or indirect depending on whether the direction of the crystal momentum vector of the electrons and holes in the valence and conduction bands bands is the same or different, respectively.  If you don’t understand that, don’t feel badly – the crystal momentum vector is only a virtual vector, so it isn’t really there anyway.

Delving further, explanations of semiconductor junction physics descend into pages of little squiggly things that would look nice on a wall, but tend to produce headaches.  For those who would actually like to explore this topic further, the Wikipedia article on electronic band structure is quite good.

What this means in a practical sense is that some semiconductor junctions can produce light and some can’t.  Charge carriers (electrons and holes) in those with an indirect band gap go through an intermediate energy state (rather like a pinball bouncing around) and dribble their transitional energy away into the crystal lattice in the form of heat – this occurs with silicon and germanium.  With a direct band gap, charge carriers make only one transition, giving out a nice photon as each carrier makes the jump – gallium arsenide crystals do this, as do crystals of aluminium gallium indium phosphide.  The color of the emitted light can be changed by doping the semiconductors with impurities to adjust the width of the band gap.  The Wikipedia article on LEDs has a nice table of semiconductor materials and their respective spectra, as does the Zeiss site.

LED Structure (from ADLED site)
LED Structure (from ADLED site)

Once you can produce light from a semiconductor junction, how do you get it out of the junction and focused into a beam that is useful?  Actual LED chips are tiny, ranging from 1/10mm to 1.0mm in size.  Most commonly, they are mounted in a small reflective cup on one of two supports, the anvil, or cathode.  A fine metal contact wire from the post, or anode, connects to the surface contact of the LED.  These internal components are sealed in an epoxy case with an apical lens that concentrates the light reflected from the cup into a beam of predetermined width.

Unlike the simplistic pictures of junctions, the structure of an actual LED is quite complex, and reflects thousands of hours of corporate and academic research, published on thousands of pounds of paper, and wrangled over at hundreds of annual conferences.  Different manufacturers have different designs, and the average LED is a highly engineered multilayer device.

Modern LEDs may use Bragg reflector (dielectric mirror) layers, insulated layers that channel current flow around microscopic central etched cavities, surface grooving to decrease internal reflection,  and a myriad other microstructures, all aimed at increasing efficiency and decreasing light losses.  Terms like evanescent wave coupling are bandied about. All of these little marvels work in the Alice-in-Wonderland world of quantum devices, where the physical laws we are used to simply don’t apply.  A few samples give one an idea of the complexity and diversity of these tiny devices:

Grooves etched on LED surface decrease internal reflections and markedly increase light emission
Insulating layers channel current within the layers of the microchip
Different modes of light emission – lateral versus vertical from etched surface point source

These images provide only a taste of the many variations in LED design, and advances in LED micro-engineering are occurring on an almost daily basis.  Remember that all of this structural work is occurring in an object about the same size as the head of a small pin.

Note especially that the mechanism by which the LED produces light is completely different from that of the old-fashioned light bulb.  Incandescent bulbs emit light due to black body radiation; this kind of radiation results from the increasing vibration of atoms and molecules as bodies are heated to higher and higher temperatures.  Atoms have charges, and vibrating a blob of charged particles faster and faster as the filament heats up generates electromagnetic waves that we see as light, going from longer to shorter wavelengths as the atoms move faster.  The resulting light has a continuous spectrum that shifts from red to blue as the temperature rises:


Any body at a given temperature will emit the same spectrum; for this reason, the color of light bulbs or other light sources is often expressed as a color temperature.  Midday daylight is roughly equivalent to the light from a heated body at about 5600 degrees Kelvin.  Note also that black bodies like light bulbs are hellishly inefficient producers of visible light, with most of the energy output occurring in the invisible and useless, long-wavelength far reds and infrared.  Consequently, most of the energy from the batteries in an old-fashioned flashlight is wasted in producing heat.

For balance, the reader should also review David Walker’s superb article in defense of tungsten illumination.  His images of denser objects taken with near-infrared illumination are striking; these would be difficult to accomplish with the narrower spectrum of present-day LEDs.  His compilation of creative ideas for illumination on the Biolam microscope is also well worth reading.


Production of the daylight-balanced white light essential for most microscopy applications requires even more ingenuity, since LEDs typically produce only a single spectral band.  There are actually several ways in which LEDs can produced white-appearing light (summarized from the detailed but excellent Olympus Microscopy Resource Center article “Introduction to Light Emitting Diodes“).

Probably the most common technique is to combine a blue, violet, or UV-emitting microchip with a fluorescent phosphor.  The short wavelength, high-energy light from the microchip falls on a phosphor lining on the surface of the reflective cup.  The latter absorbs the short wavelength light, then re-emits the light energy (fluoresces) with a broad band of longer wavelengths:

"White" LED with phosphor-lined reflective cup
“White” LED with phosphor-lined reflective cup

Phosphors typically consist of an inorganic host matrix, often yttrium aluminum garnet (YAG), doped with a rare-earth element such as cerium.  The overall emission spectrum then consists of the primary short-wave (blue) band from the microchip, together with the longer-wave and broader emission band of the phosphor:


Other methods include LEDs where all of the emitted light comes from a mixture of phosphors and the blue source light is never projected from the device.  Such devices can use phosphors already developed for fluorescent lights; unfortunately, this method has a much lower efficiency than the mixed microchip/phosphor emission technique.

Instead of phosphors generating the longer-wave spectral elements, another group of LEDs in development employs a second semiconductor layer to absorb short-wave light and re-emit longer wavelengths.  This combination of a current-driven blue LED feeding an optically-driven, longer-wavelength LED is known as a photon recycling semiconductor, or PRS-LED.

In some cases, white light is generated by a three-LED RGB unit whose color can be varied by changing the current driving the red, green, and blue LEDs.  By balancing the intensity of each LED, any color can be produced.

HOWEVER, remember that, even though LED light may appear white to the human eye, it is still only an approximation of the continuous spectrum of white sunlight that floods Death Valley every afternoon.  White LED light is still cobbled together from individual, discrete bands, with peaks and valleys that are not present in the continuous daylight spectrum.  As such, supposedly white LED light may behave differently from true white light in certain microscopic techniques, or may render colors and stains differently.  Chromaticity considerations are discussed in the Olympus LED reference.  This site also has a good general discussion of visible light sources.  The Zeiss web site also has a thorough discussion of light emitting diodes in microscopy.

However, when these theoretical concerns are put to practical testing, as in David Walker’s excellent side-by-side comparison of slides viewed by tungsten light and illumination from a Phillips Luxeon III LED, the images looked almost identical, with the LED colors actually being slightly more vivid.  Mr. Walker also discusses the significance of the color rendering index (CRI) in older and current LEDs.  To my pathologist’s eye, the color differences are slight, and seem functionally insignificant.

Non-Koehler illumination systems are reasonably tolerant with respect to the light distribution from an LED source.  With Koehler illumination, the angle at which the cone of light from the LED spreads, the evenness of the light distribution, and the presence of shadows from small internal structures such as contact wires all become critical.  As part of its thoughtful discussion of microscope illumination, John Walsh’s Micrographia site has a discussion of features to look for in an LED with respect to microchip size, as well as the curvature and positioning of the lens:


Unless you break it, an electric light is really a pretty simple-minded and amiable creature.  The more electricity you feed it, the brighter it gets.  You make it brighter by increasing the voltage applied across the filament, and if you plot the current flowing through (and heating) the filament, it follows the applied voltage in a regular and predictable manner:


(From I B Physics Stuff)

The graph would be a nice straight line if it weren’t for the fact that the resistance of the filament increases as it heats up.

John Powell aptly describes what happens if you try this experiment with an LED:

“…The relationship between current and voltage in an LED is non-linear. As the voltage increases from zero there is only a trickle of current and no noticeable light. At about a volt and a half … the current begins to increase appreciably and the first glimmers appear.

LED current graph

At two volts the LED is bright and with a fraction more it’s very bright. Once over about 2.2 volts, the current rapidly soars beyond safe operation. The LED soon overheats and dies…”

The reason for this behavior is that the applied voltage increases until it forces electrons and holes over the energy barrier at the junction between the two different kinds of semiconductor (the n and p layers).  Initially, nothing much happens until this threshold energy is reached.  However,  once the cascade of electrons begins, small increases in voltage result in a deluge of charge carriers across the junction and, like water eroding a dam, the LED can quickly be destroyed.  Consequently, LEDs cannot be dimmed by increasing the voltage; instead, the voltage must be held relatively constant and, unlike incandescent bulbs, dimming is accomplished by varying the current through the LED.

The water-over-the-dam analogy is really quite good.  With the LED, imagine you’re filling up a dam; the water level rises and rises, yet nothing much happens.  When the water reaches the top (junctional energy barrier), a trickle appears over the edge.  A bit more, and there’s a nice flow down the spillway.  A tiny bit more, and we have a torrent over the top that eats away at the dam and quickly washes it away.  To control the amount of work done by the water behind the dam, we use the gate on the spillway to control the flow (current).  Within limits, the water level behind the dam (voltage) can rise or fall, and we just open the spillway more or less to compensate.

An electric light behaves more like water running through a V-shaped canyon – raise the level or pressure (voltage) at the beginning of the canyon, and more and more water flows through.  The water level rises in the canyon as flow increases, but it is a reasonably steady and linear process.  You will finally reach a point where the waters wash away the walls of the canyon or overflow across the countyside (the point where the lamp burns out,) but the flow has been a fairly regular progression up to that point.

There are two ways of supplying power to an electronic gizmo: through a voltage source or through a current source.  Most sources of electric power with which we are familiar are voltage sources:  our 110V or 220V AC wall plugs, a 1.5V flashlight battery, etc.  A 12V car battery can run a flashlight bulb for a week, or the starter on your car for a minute.  In both cases, the voltage stays fairly constant at about 12 volts, but the current varies from a tiny trickle with the bulb to a hefty 100 amps for the starter.  Putting the wrong gadget across the battery results in a shower of sparks (and a melted gadget) as the current soars and burns it out.  This is a good example of a voltage source – the voltage stays pretty constant, but the current can vary enormously.  This is how we usually think of electricity – we decide the voltage, and the current is just whatever is needed, up to the maximum capacity of the source.  BUT – this is NOT what we need to keep our LEDs happy and unfried.

We are not used to stabilizing the current first and letting the voltage follow along;  this is, however, the nature of a current source – a power source that provides a constant current regardless of the voltage in the circuit.  Note that many solid-state power supplies are pretty happy with a range of voltages, and you can plug  many electronic gadgets into 110 or 220V and they don’t seem to notice the difference.   An electric welding power supply is a good example of a constant-current source, where the amount of metal deposited in the weld is a function of the current.  Similarly, LEDs need to be powered by a source that, at the minimum, keeps current flow from destroying the diode, and ideally, allows the current to be varied in a stable and regulated fashion in order to control the light intensity.

The simplest and most primitive way of creating a current-controlled power source is to take a voltage source (such as a battery or standard power supply), and add a series resistor in the circuit.  This does not create anything like a well-controlled source of amperage, but it does limit the maximum current that can flow through the LED, although in a way that wastes power.  As John Powell notes:

“…The usual approach is to put a resistor in series with the LED. The combination is still non-linear, but in a much more well behaved manner. In fact, over the range of safe operating current, it acts incrementally linear.

LED with resistor

The catch is that it wastes power. If a 12 volt supply is powering a single LED-resistor combination, 2 volts goes to the LED and 10 to the resistor. Only one sixth of the power makes it to the LED.”

However, with an LED, the total amount of power used is small, and resistor losses may not be important.  The following is a simple LED current limiting and dimming circuit  involving only a resistor and potentiometer, working from a 5V power supply:


(From Phil Frost, EE Stack Exchange)

The disadvantage of these very basic circuits is that they are inefficient and may not provide uniform dimming as the potentiometer is rotated.  Better current control, uniform dimming capability, and less power wastage require a more sophisticated circuit.  Searching the literature, as I did for a week, reveals a mind-numbing collection of complex circuits using operational amplifiers, integrated circuits with optical transistors, Zener diodes, etc., etc.  Fortunately, when one understands its relatively simple needs, the LED is actually a fairly amiable and undemanding critter that is happy with a ham sandwich for dinner every evening, and doesn’t require champagne.  However, it is a major project to find a simple, constant-current supply circuit with dimming capability for a single LED or small group of LEDs; many of the available circuits are meant for constant intensity lighting circuits or are designed for residential lighting.

Frank Weithöner (see below) describes a simple variable current two-transistor circuit for powering and dimming a microscope single-LED illuminator:


The two transistors, the MOSFET BD237/243 and the general purpose NPN BC546, are at present commercially available, and conversion datasheets are readily available for equivalent transistors.


Let’s now turn (literally) to the nut and bolts of doing your own LED conversion.  If you decide to design and build your own unit, there are a few references available on the internet.  However, there are as yet no standard techniques for retrofitting incandescent illumination systems with LEDs. LED illumination systems are relatively new and rapidly evolving for all applications, including automotive and residential, and articles on retrofitting classic microscopes are scanty as of 2016.

One important and as yet unsettled question is the number of LEDs in the light source.  Where will the microscope industry land – with a single, high-intensity LED and lens, where one deals with the same problems of uneven center-to-edge brightness found with conventional lamp filaments, or with a multiple-LED source, which is more even between center and edge, but where one must strongly diffuse and evenly spread the light from several discrete sources?  So far, the replacement illumination systems for binocular scopes with or without Kohler illumination capability have used high quality, high intensity single LEDs with focusing lenses.

One new development that may have promise for producing a very uniform light source without the problems of evenly spreading light from a point source LED is COB, or “Chip-on-Board” technology.  In this new design, multiple small microchip LEDs are arranged in an array to form a small, uniformly illuminated lighting panel.  While many of the manufacturers are in China, American lighting corporations such as Cree Inc., headquartered in South Carolina, list a wide variety of small (1-4 cm) lighting panels using this technology.  This new light source is only recently commercially available and has not yet been used in microscopy but has, I think, considerable promise.

Cree CXA2 1.2cm COB LED

As of June, 2016, there are no online references on the possible uses of COB technology in microscopy.  One question, to which I do not immediately have the answer, is how a small flat-panel source fits in with Kohler-type illumination and maximizing resolution.

There a few articles available on LED conversion of older or semimodern microscopes, all of which to date have employed single-chip LEDs with focusing lenses. One excellent site is Frank’s Hospital Workshop, devoted to hospital equipment maintenance in impoverished Third World countries. Frank Weithöner describes conversion of Olympus CH-2 microscopes to LED illumination using a single Luxeon Star/O LED with lens.  He also has a good general page on power supply basics.  This is one of the most carefully crafted sites that I have encountered, with well-written, practical advice on the maintenance of a variety of medical equipment.  The images below, from Mr. Weithöner’s site, show an Olympus CH-2 microscope before and after LED conversion.  The small SMPS 5V power supply is on the left in the converted base, while the potentiometer and current control circuit are at the top:



If the standard power supply can be modified, or a tiny converter supply built into the socket, this might allow conversion to LED without significantly altering the microscope.  A similar mounting of an LED onto a microscope bulb base has been described by David Walker.

The above sites provide details of professional-level conversions.  There is also a place for simple and inexpensive solutions using local materials, similar to spending $5.00 on a cheap dimmer to rebuild a Nikon S illumination system (see Part I).  The cheap and readily-available LED flashlight may prove a simple fix for illumination problems.  When my Nikon S illumination system failed, I removed the lamp and inserted a cheap (10 for $12.00 at the thrift shop) seven-LED flashlight that fit into the lamp housing.  Without a good diffuser, it was nothing that I would use for photography, but it worked and the scope was usable on an emergency basis.

A very simple conversion of an American Optical 150 microscope into a portable scope, using an off-the-shelf LED powered by a 9V battery, was published in the November 2011 issue of Micscape magazine by Bill Resch.  The LED was clamped in place and supported on its two wires in the same place as the lamp filament. Bill’s published images suggest that the illumination was quite even.

In the March 2011 issue of Microbe Hunter magazine, Suphot Punnachaya described a simple LED conversion of a Chinese microscope’s halogen illumination system using an LED array obtained by sawing off the front inch of an inexpensive LED flashlight.  Power was supplied by an old cell phone 5V “wall wart” charger connected to the LED through a series resistor and rheostat.

An even simpler conversion of a Zeiss Gfl scope was describe in the August 2013 issue of Micscape by Franz Schulze; this used a $2.00 single-LED flashlight with lens, purchased from the Dollar Store.  The flashlight was mounted in a simple wooden sleeve machined from a wooden dowel, and placed into the port for the lamp assembly.

It should be noted that microscopes such as the Zeiss Gfl and the Nikon S, where the bulb assembly is inserted  into a port or sleeve in the body, are easiest to convert to LED illumination.  Those models where the bulb is enclosed inside the base require somewhat more work, ingenuity and removal and remounting of  components.  In these cases, assembly and mounting within the base might be further simplified by using a hot glue gun to mount components (or the LEDs themselves) rather than machining clamps.


Powering the LED illuminator can be as simple as using coin, flashlight, 9 volt, or rechargeable batteries, since the power requirements of many LED sources are low enough that batteries are a viable option.  The need for a series resistor is described above.

To power an  LED light source from 110 or 220 volt mains AC, one has the same options as for the incandescent power supplies discussed in Part II:  buy a small LED power supply (known as an LED driver) or build your own circuit. Many inexpensive dimmable LED drivers are available on the internet; they can be wired with a simple Triac dimmer to control the input voltage and hence the output current:


Small Triac dimmer – disconnect the potentiometer connections and reconnect it with three wires, then use it to replace the intensity control


Small dimmable LED driver supply – under $20 online from many sources

Specifications for this dimmable LED driver

 This is probably the simplest option for setting up an LED supply within the base of a microscope.  One might consider combining the small dimmable LED driver with the guts of a cheap dimmer, the latter replacing the microscope’s illumination control and the rest of the dimmer’s circuit components being transplanted as a unit, as described in Part I.  However, you might be better to buy a small dimmer unit matched to the LED driver from the same supplier.

As far as building your own LED power supply from scratch, you can do it if you are more familiar with modern electronics than I am.  Yet in these days of mass-produced circuit boards, this seems like unnecessary complexity. Browsing LED driver circuits online resulted in many pages of complex diagrams using integrated circuits, MOSFETS, small inductors, etc., etc.  I have a PhD, and I felt lost and frustrated.  This seems like a lot of work when I can spend $10-20 and buy a smaller and better-made driver from any one of fifty electronic supply houses, then add on a simple dimmer circuit and figure out how to situate it all within the particular configuration of the microscope base..

One  more thought – the power supply for your microscope’s new LED system may be sitting right in your desk drawer.  Or in a box your neighbor’s garage.  Or at your local thrift store.

PS 02
Wall of “wall warts” at Goodwill store

For the last twenty years, orphan power supplies for all kinds of outmoded electronic gadgets have been piling up in boxes and drawers and finding their way to thrift shops.  Your local computer service store may have a box of disused power supplies in their back room in various voltages.  Some of these are wall warts, while others are inline supplies.  These originally powered computers, printers, cell phones, zip drives, makeup mirrors, cordless phones, games, and multiple other devices that need low voltage DC from a wall socket.

Simple, old-fashioned “linear” transformer-based wall wart

Now, remember that these small power supplies are voltage sources and are of two types: SMPS and linear.  Most of the newer supplies will be switch mode supplies (SMPS) that convert, reconvert and regulate their low-voltage output, with light, high-frequency transformers (see SMPS discussion in Part II).   The older, bulkier units are usually very simple 60Hz linear supplies, with bulky, low-frequency 60Hz transformers.  You can tell the difference by just hefting them in your hand; SMPS are very light, while transformer-based linear supplies are quite heavy.

To convert either to a usable current-based power sources, one can employ either a basic resistive network or a simple current regulating circuit such as that described by Frank Weithöner (see above).  It may also be helpful to Google something like “LED supply from wall wart.”  This will bring up multiple posts with ideas of varying usefulness from electronic hobbyists.  John Bryant’s “A Dummies’ Guide to Working With Wall Warts” is helpful, though it focuses more on voltage than current sources.  I have not cited any others specifically, as they change on an almost daily basis, but they can have interesting ideas.

Be aware, however, that some of the cheap, no-name wall warts lack the protections and sophistication of brand name devices and can be dangerous (see Ken Shirriff’s article “Tiny, Cheap, and Dangerous“):

Genuine iPad wall wart versus cheap copy.  The fewer components in the fake iPad supply reflect the lack of regulating and protective circuits.  Notice that the large ground post in the copy is probably plastic and, unlike its metal counterpart in the real iPad supply, does not seem to be connected to anything.

Do not dismiss the big, heavy, 1980s-1990s wall wart.  Though they are older, heavier, and clunkier, the old linear wall wart of any brand with its heavy transformer is probably safer than any cheap modern SMPS supply.  The big transformer serves as an isolation transformer, preventing the user from directly connecting with line current and risking a serious shock.  A good SMPS is a complex device with safety and isolation features built in, but a cheap switch mode supply with all the safety circuits left out, no ground, and possible internal shorts will still function, and you can’t tell the difference from outside the case.

Whether you use a purchased, ready-made LED driver or put together a supply from inspired scrounging of bits and pieces at the thrift store, this can be an inexpensive but creative project.  LED illumination technology is in its infancy, and new products are becoming available every day.  Consider the conversion a challenge to your inventive talents! Or, if you just want to plug in a storebought box with a cord on one end and an LED on the other and get back to chasing Peranema, that’s fine too.  If either approach helps you to fix your dead scope or restore a classic beauty, it’s worth it.


Note on best general references:  The Micrographia site has a good general discussion of microscope illumination, including thoughts on optimal LED design for microscopic illumination.  The ZL2PD Amateur Radio web site has an excellent basic discussion of switch mode power supplies.  The Olympus Microscopy Resource Site is a very good overall resource on microscopy, with a superb article on all aspects of LED function.  The Zeiss site Fundamentals of Light-Emitting Diodes (LEDs) is also very well-written, informative, and dense with information, especially with respect to LEDs’ potential role in fluorescence microscopy.  I have tried here to summarize just the information necessary for a general microscopist’s working understanding of this new technology.

Part IV deals with the physiology of the eye as it pertains to the many unresolved questions regarding LED safety.


Davidson, M. W.  “Fundamentals of Light-Emitting Diodes (LEDs).”

Elliot, R.  “Electronic Transformers, the Good and the Ugly.”  Elliot Sound Products website, accessed April 4, 2016.

Engdahl, T.  “Light Dimmer Circuits.”

Frost, P. Posted on Electrical Engineering Stack Exchange.  “Using a Variable Resistor to Dim an LED. “

IB Physics Stuff.  “Electric Circuits.”

Intelligent Controls pamphlet, accessed March 25, 2016.  “How a Dimmer Works.”

Kuphaldt, T.  “Fundamentals of Electrical Engineering and Electronics:  The Triac.”

Majumdar, S.  “How Switch Mode Power Supplies (SMPS) Work.”

Micrographia Site.  “The Microscope Lamp.”

Powell, J.M.  “LEDs and Dimming.”

Spring, K. R., Fellers, T. J., and Davidson, M. W. , Olympus Microscopy Resource Center. “Introduction to light Emitting Diodes.”

Weithöner, F. “Microscope Conversion to LED Light.”

ZL2PD Amateur Radio Website.  “Introduction to Small Switchmode Power Supplies.”

How Bright is your Light Part III: LED Conversion

How Bright is Your Light Part II: Repairing More Modern Illumination Systems

Zeiss Laboval 4, base open to show illumination power supply
Closeup of power supply circuit board – four transistors, one integrated circuit, two fuses, three trim pots, and four round gizmos that I cannot identify (left side – bridge rectifier??)
..and all to light one little bulb

This article continues from Part I, which discussed repairing the illumination control circuit of the Nikon S, as well as similar very  early electronic microscope power supplies.  Here, we discuss the principles behind more modern microscope illumination systems and suggest fixes for blown lighting circuits in scopes of this vintage.

The simple fix described in Part I worked well for the uncomplicated, 1970s-era intensity control circuit on the Nikon S, and will probably be applicable to similar early microscope illumination systems. Unfortunately,  the greatest number of quality used microscopes are of more recent vintage, use more complex electronic control systems, and are more of a challenge to repair.

Removing the base plate from a newer generation microscope will likely reveal a printed circuit board bearing a number of small capacitors, resistors, coils, transistors, and integrated circuits.  The Laboval 4 power supply shown above is complex, with some very Russian-looking components.  Built in East Germany in the late 1980s to early 1990s, the Laboval 4 is built like a Russian tank and has excellent optics – but buying replacement electronic parts would be almost impossible.  You can see why a scope of this type would be scrapped if the power supply blew.

Besides, what do all of these electronic gizmos have to do with running a light bulb?  I must admit to having had almost complete ignorance about electronic microscope power supplies, and yearned for a good old autotransformer that could break a toe if dropped.  However, if your power supply dies and you are faced with repairing it or spending bucks for a new scope, don’t give up.  That small, mysterious printed circuit board in the base uses very standard circuits, and the supply can be rebuilt or replaced more easily than you might think.


In the decade after the Nikon S-Kt, microscope manufacturers worked hard to cram a smaller but equally powerful illumination system into the microscope base and eliminate the external supply.  Since the transformer is the largest and heaviest element in the illumination power supply, this meant reducing transformer size while maintaining capacity.

The solution was the so-called “electronic transformer” – actually a misnomer for a “switch mode power supply” or SMPS.  Don’t let the fancy name scare you; this type of power supply arises from an ingenious but simple concept and is made from basic circuit components.  The essential idea is that the size of a transformer, for a given wattage, is inversely proportional to the frequency at which it operates.  So a transformer operating at 300 Hz (or 300 cycles/second) can theoretically be 1/5 the size and weight of a similar, old-fashioned unit working at 60 Hz, or AC line frequency.

220 to 12 Volt “Electronic” Transformer.  This is only a few inches long and weighs a few ounces.

This type of power supply operates by transforming 110V or 220V line AC into direct current, and then transforming it back into alternating current via a simple transistorized switching circuit – but at a higher frequency.  This “re-manufactured” higher-frequency alternating current can then go through a smaller and very efficient transformer to be stepped down to a lower voltage suitable for powering an incandescent or halogen lamp. Many of the small supplies work at quite high frequencies of 20,000-50,000 Hz; they are thus quite efficient, and any interference signals produced are well above the threshold of human hearing.

The circuitry for rectifying the AC current to DC and then”inverting” it back to higher frequency AC  that can be re-converted into whatever you need (much like the 12V to 110V inverter one buys at the hardware store) is light and compact, and the final transformer is a fraction of the size and weight of an old 60 Hz unit.  The AC ==> DC ==> high freq AC ==> low voltage AC transitions for lamps are handled on a board a few inches long and weighing only a few ounces.  For incandescent or halogen lamps, the low voltage output can be either alternating or direct current – lamps with filaments aren’t fussy about what they eat for dinner.  For a direct current-only device, a final step is added in the form of a second bridge rectifier to convert the low voltage AC into smoothed DC:

Simplified circuit for a DC output switch mode power supply, with second bridge rectifier converting low voltage AC output to low voltage DC

The filter capacitors store energy as the rectified direct current pulsates, taking in energy during the peaks and releasing it in the valleys, much as an old-time millpond provided a constant flow of water to the mill’s water wheel.  The result is a smooth direct current output voltage.


Since solid state components (integrated circuits, transistors, and diodes) can now be mass-produced for a few cents, tiny SMPS power supplies are so cheap and convenient that they are used for supplying most small electronic devices – battery chargers, cell phones, tablets, clocks, lamps, etc. – and most households have at least a dozen hiding in wall sockets and behind bureaus.  A home with many electronic gizmos may have a hundred of these small devices.  In electronic slang, the small, socket-mounted variety is called a “wall wart“. This derisive moniker arose from the tendency of the older, larger chargers to eat up space on wall sockets and power strips, and to be heavy enough to fall out onto the floor at intervals, crashing the instrument they were supplying. Today’s socket-mounted  SMPS chargers are much smaller and lighter than the “linear” designs of twenty years ago, which, using a bulky transformer followed by a rectifier and filter, often drooped from the socket.

“Wall Warts”: Old linear transformer-based power supply (bottom) drooping from the socket, versus contemporary tiny iPhone SMPS (top)

One should not, however, disparage the older “linear” power supply.  The term “linear” refers to a supply that, unlike the SMPS, does not convert and reconvert electricity from AC to DC to AC, changing the type of current AND the frequency AND the voltage.  SMPS supplies have taken over in most consumer applications because of their small size, light weight, low cost, and high efficiency.  However, their components are constantly working as tiny switches; they consequently have a higher noise level and poorer voltage regulation.  The older “linear” (transformer ==> rectifier ==> filter) supply, named because its components are operating in the linear phase of their operating curves, is inherently heavier and produces more heat, but has less noise superimposed on the output current, better regulation, and faster recovery from transient spikes in the line voltage.  They are still used where very pure, low noise, very tightly regulated current output is required.  A chart comparing the features of the two systems can be found on the Acopian site.

From a microscopist’s standpoint, the tiny but powerful SMPS supplies and halogen bulbs resulted in leaner, lighter scopes and brighter lighting without bulky power transformers.   HOWEVER, the downside of this innovation was that those clunky but durable transformers were replaced by electronic components with a much more uncertain lifespan.  These may be difficult to replace after a few years have passed (one service specialist struggles to find parts for some twelve-year-old models), and the first generation of these electronic systems is now approaching the 50-year mark.


One has three alternatives to repair a blown electronic power supply: move to a simple external supply, buy a power supply that will fit, or custom-build a small supply and control unit that fits into the same space as the old unit.  Building an external supply was discussed in Part I – a cheap dimmer driving a 6V or 12V transformer, possibly of type used for a doorbell and available from a local home supply store.  Buying a new power supply is easiest, and these are cheap and standardized – this small SMPS works on 85-265V AC input, puts out 12V DC at one ampere, and costs $8.40 on Amazon Prime:85-265V to DC 12v "Electronic Transformer"One may have to figure out how to tie the microscope’s light intensity control potentiometer into the circuit, but this is a cheap, prepackaged fix if the supply fits the space available in the base.  These small power supplies are available from online stores in a wide variety of combinations of wattage and output voltage.  A few have the dimming function already built in, and are intended as replacement power supplies for single halogen bulbs:

6 Volt, 30 Watt Dimmable Halogen Lamp Replacement Power Supply, $22.99 on eBay

If you do buy a generic power supply, try to pay a bit extra and get it from a reliable manufacturer.  There are many cheap electronic devices available in the online market, but many of these are of overseas origin, poorly-constructed with substandard components, badly designed, and pose a shock hazard (see Elliot).

One very cheap solution is to look for an old computer or printer power supply in an appropriate voltage.  This may take considerable ingenuity and searching, but these can be found readily at thrift shops or Salvation Army stores.  Depending on the supply’s circuitry, its output may be dimmed either with a low-resistance rheostat on the low voltage side, or an appropriate 110 or 220 volt dimmer on the line side.  The supply can either be mounted in its plastic case, or opened and the circuits removed.  An older, linear-type supply may be easier to dim.  It may be necessary to try two or three thrift shop power supplies until you find the one that works.  You might even consider buying an old, variable-intensity halogen desk lamp at the thrift store and transplanting the power supply.


Rebuilding the power supply or building it from start is a bit more challenging, but can be done even with modest electronic experience.  You may be able to clean off and use the appropriately-shaped circuit boards from the old illumination system.  Depending on the illumination system of the microscope, there is a multiplicity of adaptable basic circuits on the internet; these may use two transistors, thyristors, or a small integrated circuit for the switching circuit, as there are a number of ways that this simple circuit can be set up.  The circuit below uses a bridge rectifier (left side) to produce DC from 220V mains AC, then employs an IR2153 integrated circuit and two IRF840 MOSFET (Metal Oxide Semiconductor Field Effect Transistor) devices to chop the DC to rapid frequency AC; the latter is then stepped down by the small transformer on the right to light a halogen bulb.

halogen lamp circuit

Given the many variations in space and circuitry used by microscope manufacturers, there is no single fix that is adaptable to every vintage scope. Building a replacement illumination supply will take a bit of ingenuity and perhaps a visit to your local electronics store (NOT the consumer radio store in the mall – you need a REAL electronics store with racks of equipment and perhaps a bit of dust in the corners).  Which circuit you use will depend on available materials and components, what kind of advice you find, where you live, and how much power your light needs.  Finding a truly simple dimmable power supply circuit for a single halogen bulb is not an easy task; the above circuit is one of the least complex.  But actually constructing one of these power supplies is not rocket science and can even be done as a first electronics project.


This is an example of a tiny, home-made halogen lamp SMPS power supply – look closely and you will see that the wiring and soldering are pretty rough, but it works:

OLYMPUS DIGITAL CAMERAHomemade Switch-mode" Halogen Light Power Supply

This small, home-made switch mode lighting circuit supply uses two thyristors (or possibly MOSFETs, they all look similar) and a small integrated circuit.  Note how compact it is – this should be able to be fitted into the limited space in a microscope base and work with the old bulb and lenses.

Sixty years ago, this unit would have been completely impractical, with vacuum tubes and their power supplies occupying the space of a packing crate, the unit weighing 70-100 pounds, consuming more electricity to heat filaments than it produced, and

emitting enormous amounts of heat.  It is a testament to modern microelectronics that it is practical to convert and reconvert power several times in this manner, and that the process can occur in  a packet the size and cost of a box of lozenges.

For those to whom this is all new information, the concepts behind the electronic power supply are very simply and coherently explained at the Williamson Labs’ site:

Note this especially:  If you think that this is all a long way from looking at small wriggly things and batting mosquitoes in the swamp, you are right.  A quick look at the specification sheets above for components like the IC and the MOSFET rapidly convinced me that I’d rather just sit back and applaud than try to understand what’s happening in all those little black thingies.  BUT – if this lets me buy a research-level Zeiss or Olympus with blown electronics for $200, and then have it sitting on my bench to use every day, it’s worth it.  And you don’t really have to understand it, except in a general sense.  You just have to know enough to solder together the wires and close up the case.

If, however,  you are inspired to learn more about these ubiquitous and amazing little devices, read Ken Shirriff’S dissection of the Apple iPhone power supply (the little white cube in the above image).  His description of how an SMPS (including transformer ), with sophisticated voltage regulation and overload protection, is compressed into a one-inch plastic cube is fascinating.  The images below are courtesy Mr. Shirriff:

The images above show how ingeniously every cubic millimeter was used in the iPhone cube, as well as the sophistication of the SMPS circuitry to handle a variety of loads with reasonable regulation in such a tiny space.  Fortunately, microscope illumination systems can be much more basic and much less space-intensive, with simple circuits that do not need to adapt to such a multiplicity of circumstances.

Hopefully, this information will take some of the mystery out of microscope power supplies.  I wonder how many sophisticated microscopes have gone into the junk bin or been broken up for parts because the owners assumed that a busted power supply was irreparable?  I would like to have suggested one simple, all-encompassing solution that would replace the power supply for all of the “semi-modern” microscopes that are new enough to have electronic lighting supplies, yet old enough that parts are no longer available.

Unfortunately, what will work as a replacement power supply is dictated by space, the power and voltage requirements of the bulb, the materials at hand, your experience with tools, and, ultimately, your budget and patience.  There is no single universal fix.  Ingenuity and backwoods improvisation are often the essential elements to salvaging  a sophisticated scope with bad electronics.  The point is that these power supplies from 20-30 years ago can be replaced or rebuilt, and they are not mysterious, irreplaceable devices. If you have an idea of how these circuits work and are willing to dive in, your reward will be a functioning, classic microscope and the satisfaction of having a fine piece of equipment that you have restored.

  When things were simple: antique microscope light sources – Spencer on left, AO on right

 This article is continued in Part III, which deals with LED conversion.  We will review the construction and function of light emitting diodes, discuss how to power and regulate them, learn why they are NOT light bulbs, and include some of the few articles describing how classic scopes have been adapted to use LEDs.

How Bright is Your Light Part II: Repairing More Modern Illumination Systems

Seattle: The Tiny Desk Aquarium

On one of my many trips to Seattle, I began contract work with the University of Washington, and decided to start a new aquarium in my hotel room.  For $2.99, found an 8″ cubical vase with thin, fairly even side, and stocked it with plants and mud from the U of W lakeside grounds and in ditches around Koll Business Park.


Anyone who has been around marshes knows that oily films on the surface are very common, and result not from pollution, but from substances released from decaying organic matter. Examining these quickly shows that they are a rich and possibly poorly explored environment.

The tiny hotel desk aquarium is doing well, but has developed a film of this type as the leaves, organic debris, and mud stirred up during collection undergo natural processes of degradation; the film can be best appreciated by comparing it to the clear avenue in this image (see image where you can see a clear avenue in the film):

Aquarium with Surface Oily Slick
Aquarium with Surface Oily Slick

Passing a slide through this film brought up a layer of mucky brown substance:SURFACE SLICK ON SLIDEYet the microscope revealed an area of almost unbelievable richness of bacteria, flora and fauna, as shown in the two videos, one of the film of degenerating plant material, and one of the water between islands of decaying fibers. Resolution is not perfect, but best I can manage in the field, and you are looking through a fair anount of guck on many of the levels. I count long strands of acinetobacter, gliding algae, cyanobacteria, ciliates, sessile algae, a rotifer, a small water flea (not shown), Synura or a similar species, several Vorticella, and many flagellates, a few of which I think are euglenoids:

Many questions come up regarding the function of these surface films and their role in the marsh in terms of gas exchange, absorption of sunlight, breakdown and recycling of organic material, possibly aided by solar energy, etc, etc. Fascinating – like having a webcam into a stretch of Amazon rain forest!

Seattle: The Tiny Desk Aquarium

How Bright is Your Light? Part I: Rebuilding the Nikon S Light Intensity Control with a $4.95 Dimmer Switch

Nikon S-Kt illumination system, base removed.  Note old-style “toroidal” transformer at bottom and voltage adjusting potentiometer at upper left.  The toroidal design is actually highly efficient and shows less power loss than square designs.  The thyristor (Triac) is the pillbox-sized rectangular structure at upper right.  Someone has improperly rewired the potentiometer at upper left – note bare wires!


Restoring my beautiful old Nikon S proved to be a continuing saga – and a remarkable learning experience.  In my last article, I described the procedure for replacing a broken nylon fine focus spur gear, a problem that has sidelined many of these finely-crafted old scopes.

I originally bought the Model S-Kt with lenses (in lovely shape from 20 years on a shelf) for only $75.oo because the light would not turn on.  Getting the light working took only ten minutes with a soldering iron to re-secure a loose flange on the light bulb, after which the bulb connected properly.  However, deeper problems lurked within.  After repairing the focus gear and bulb, using the newly-restored scope revealed that the bulb voltage could not be raised over two volts and the transformer was overheating and smelling hot. It was unclear whether the transformer was shorted or the electronics were not working.

This question was solved when, after half an hour of operation, the light suddenly flared to its full brightness and the intensity control stopped working.  Clearly, the transformer was producing full voltage, but the 45 year old electronics had blown.  Removing the baseplate revealed homemade circuits on the intensity control potentiometer; these may have contributed to the overheating and failure of the old circuits.

I knew nothing of light control circuits, but online research informed me that the core of a light dimmer system was something called a thyristor.  So what is a thyristor, and what are those mysterious little thingamagummies on the printed circuits inside the base of any but the oldest microscope power supplies?  I had to educate myself in the history of microscope illumination, and it turned out to be a fascinating journey.

(With thanks to Mark Morris for suggesting this fix)

In the 1950s and 1960s, microscopes typically had external power supplies for the illumination system. These usually had no electronic components, and relied on an autotransformer, or variable transformer, to control light intensity.  The latter is a simple device, with a single winding on a circular core, and a moving contact to take off a variable secondary voltage for the bulb.  Autotransformers are primitive, heavy,

Old Nikon Autotransformer Illumination Control

clunky, and much too large to fit into a microscope base – but they are also simple, sturdy, and almost indestructible under normal conditions.  And when one did (rarely) short out, you tossed it into the trash, bought a new one, plugged it into the scope, and went on with your day.  The insides of the microscope contained a bulb, a mirror and sometimes a few lenses, but little else.  With the exception of dirty contacts or broken cables, these simple systems could be expected to function reliably fifty years after they were first turned on.

As microscopes became more streamlined, power supplies moved into the base of the scope and incorporated newer, more compact electronic control systems made possible by the advent of solid-state electronic devices.

Controlling electric current electronically is not a new idea.  In 1906, Lee De Forest invented the “Audion”, the first functional vacuum tube; this design and its basic theory was later improved by Fleming.  Electrons boiled off a hot filament flowed to a positively-charged second electrode (the “plate”).  A small negative voltage applied to a thin mesh (the “grid”) inserted between the filament and the plate could control this relatively large current, thus allowing the tube to be used as an amplifier and heralding the beginning of electronic communications and controls (see De Forest’s original Scientific American article:  However, vacuum tubes were fragile and wasted large amounts of power to heat filaments.

1915-20 ad for the De Forest Audion Vacuum tube

The transistor, invented at Bell Laboratories in 1947, uses three layers of semiconductor material to accomplish this function much more efficiently, with a small current applied to the “base” layer controlling a much larger current traveling from

Transistor -courtesy Sparkfun

the “emitter” layer to the “collector” layer.  In solid-state devices, current can travel either as electrons or as “holes” – vacancies in atomic outer electron layers that act as

Transistor - courtesy Corollary Theorms
Transistor – courtesy Corollary Theorms


Current amplification by Transistor (Courtesy Sparkfun)
Current Amplification by Transistor – courtesy Sparkfun

mobile positive charges.  In this way, the transistor, employing the base current like the grid of a vacuum tube, can perform the same functions as the vacuum tube, but with less weight, much less power consumption, lower voltages, and much smaller size.  With choice of physical configuration and semiconductor material, as well as “doping” (addition of trace elements to the semiconductor material), an almost infinite variety of functions can be designed into semiconductors.

Model S microscopes were manufactured just as electronic controls were becoming available.  Depending on the amount of space in the base, these microscopes employed either in-base, electronically-controlled power supplies (Model S-Kt) or older-style, transformer-based external supplies (Models S-Ke and L-Ke) if supplementary lenses and levers in the base left insufficient space for transformers and electronics.  The electronic illumination control system in the Model S-Kt was simple – essentially the most basic of light dimmer circuits – but in the 1970s was considered to be very advanced:

“Modern, advanced semi-conductor technology has provided a facility for changing the flow time of electric means. A so-called thyristor of extremely small size has been developed to enable regulation of the brightness of the lamp. The Microscope Model S-Kt has adopted this type of light adjuster built in the microscope base.”  (S-Kt Manual, Page 9)

The heart of the S-Kt system, wall dimmers, motor controllers and many other electric power control devices is the thyristor, a solid state device very similar to the transistor but with one critical difference:  while the current allowed to pass through the transistor is roughly proportional to the tiny current applied to the base, or control electrode, the thyristor (with leads designated anode, cathode, and gate) is completely non-conducting until a small current is applied to the gate, after which the thyristor flips fully to the conducting state. Thyristors therefore function as fast-acting electronic switches and, unlike the transistor, can easily control hundreds of volts and considerable current, even in high-power electrical transmission systems:

Thyristor valve hall at New Zealand power transmission station

Like the transistor, a thyristor is basically a direct current device.  However, if one connects two thyristors in parallel but oriented opposite to each other with the gate connections wired together, one has a very functional, solid-state electronic AC switch:  the Triac, or bidirectional triode thyristor.  Furthermore, many triacs are not fussy about the polarity of the gating voltage, and can be switched on by either direct or alternating current, thus making the control circuitry simple: a small control voltage can be pulled off the main power supply, applied to the gate, and control a much larger current travelling between the cathode and anode.  Although the theory and structure of triacs are complex, if one is not concerned about the niceties of symmetrical AC triggering, harmonics or very fine control (for example, with an incandescent bulb dimmer) the circuitry can be very simple:

Simple Triac Dimmer. Capacitor helps control phase of gating current.
Simple Triac Dimmer. Capacitor helps control phase of gating current.

This circuit also demonstrates the simple concept of the basic incandescent dimmer, with the AC supply, the light bulb, and the triac dimmer wired in a simple series circuit.  For a low voltage bulb, the primary of the bulb transformer can be placed where the bulb is in the circuit.  Essentially, the triac acts as a switch, cutting out increasingly large chunks of the AC supply’s sine wave as the gating current and phase are altered by the resistor/capacitor combination.  Chopping out pieces of the alternating current reduces the effective supplied power:ThyristorwaveformdiagramsThis system works well for incandescent lights, where the filament stays hot between pulses, effectively smoothing out any flicker in the illumination.


So how does one replace a 45-year old thyristor that belongs in a museum???  Excellent older microscopes (and many expensive and not-so-old models) are often sidelined because of failing electronic components and unavailable parts.  This was becoming a thorny problem until Amateur Microscopy friends contributed ideas, and Mark Morris, who has been fixing these microscopes since I was in elementary school, gave me the key idea.

For $4.95, my local Home Depot gave me a cheap Leviton wall dimmer switch.  Filing off the rivets and removing the case revealed the  simple guts of the controller: a potentiometer, a small inductive load, two mica capacitors, and the tiny, all-important chip of a Triac thyristor:


Dimmer switch, front and with back removed. Note barely-visible, 1 cm square, black thyristor behind mica capacitor (small brown object)

Tracing the dimmer’s circuit shows it to be similar to the basic circuit shown above:

Circuit for the Leviton dimmer. The potentiomer, thyristor, and capacitor comprise the basic control circuit.
Circuit for the Leviton dimmer. The potentiomer, thyristor, and capacitor comprise the basic control circuit.

The potentiometer, thyristor, and small mica capacitor on the right comprise the control circuit, while the small coil and capacitor on the left limit interference from the switching process.  As much as possible, these components were unsoldered from the dimmer potentiometer as a unit and reconnected to the intensity control potentiometer on the scope, thus maintaining the physical appearance and controls of this classic scope  – the electrical equivalent of a heart-lung transplant.  The original circuitry of the Nikon light

Components of the Dimmer - thyristor, small ferrite core inductor, and resistor
Some components of the Dimmer – thyristor, small ferrite core inductor (coil), and capacitor.

controller was very similar and equally basic, but the 1970 thyristor’s volume was around 100 times greater than that of the tiny chip removed from the modern dimmer.

Cutting out the old thyristor
Cutting out the old thyristor

After the components of the dimmer were isolated, the wires to the old pillbox-sized Nikon thyristor were cut, and the hay-wired electronic components added to the Nikon light control system (trim potentiometer, capacitor, resistor, and connecting wires) were stripped out, leaving the Nikon potentiometer in place:

Nikon Inensity control potentiometer, all other components removed.
Nikon Intensity control potentiometer, all other components removed.

The tiny thyristor from the dimmer, which had a metal flange riveted to the dimmer’s front plate as a heat sink, was secured to one of the potentiometer mounting screws inside the microscope base, and the dimmer’s circuit was recreated using the Nikon’s potentiometer:

Dimmer components installed on Nikon potentiometer. Heat sink on the thyristor, the barely-visible black chip, is secured under one potentiometer mounting screw.
Dimmer components installed on Nikon potentiometer. Heat sink on the thyristor, the barely-visible black chip, is secured under one potentiometer mounting screw.

One of the problems with this “transplant” was that the Nikon potentiometer had a resistance value of 100K ohms stamped on one side, but the resistance of the Leviton dimmer potentiometer and the trim pot were unknown;  this introduced a degree of guesswork into matching the two systems.  When this configuration was tested, the Nikon bulb would only dim partially.  Clearly, the amount of resistance provided by the Nikon potentiometer was less than the dimmer circuit was designed for.

With a considerable amount of hope and finger-crossing, the small trim potentiometer was soldered in series with the Nikon intensity controlling potentiometer.  When it was turned on, the system worked perfectly!  The trimmer’s resistance set the dimming range of the original built-in Nikon potentiometer; at one extreme, the bulb would not light at all, while at the other end of the trimmer’s range, the Nikon potentiometer could not dim the bulb fully.  Using the trimmer near midrange, it was possible to set the range of the Nikon potentiometer to ideally cover the full dimming range needed for the bulb.  This is the completed illumination control system:

Completed control system. Range-setting trimmer pot soldered to left of main illumination potentiometer
Completed control system. Range-setting trimmer pot soldered to left of main illumination potentiometer.  Tiny modern thyristor barely visible below the trim potentiometer.

The completed Model S illumination control circuit with the new dimmer components is thus:

Final Nikon S illumination control circuit. Dotted line surround components transplanted from the dimmer switch.
Final Nikon S illumination control circuit. Dotted line surrounds components transplanted from the dimmer switch.

This simple fix should work with older, 60Hz illumination control systems based around a simple power control circuit and a transformer.  It could also be used to replace a missing or non-functioning external transformer system:  a dimmer switch and a 6 volt doorbell transformer in a small electronic case, with appropriate cords and plugs, could form a cheap and simple microscope light power supply.  One could add a voltmeter and a little panel light to make it fancy.

Next:  Part II with some thoughts on repairing more modern systems as well as LED conversions, plus references.


How Bright is Your Light? Part I: Rebuilding the Nikon S Light Intensity Control with a $4.95 Dimmer Switch

The Literary Leech



This has not been a good week. Most days, I lurk innocently near the bottom of my lake waiting for the next meal to come by.  Really, I’m not a fussy eater.  Warm or cold-blooded is fine, just as long as it’s blood.  Though even a leech has its standards.  You hear humanoids say “You are what you eat.”  Well guess what – I am what YOU eat!  Have you ever put a sucker on a pink, luscious humanoid, expecting a warm flood of organically-nourished red stuff, and gotten a snoutful of the big McD?  Chemicals, artifically-digested chicken leftovers, and just plain sludge dressed up as food in a paper box – I’d rather eat fertilizer!  Most people don’t think leeches can barf, but we can – right back into your bloodstream where it belongs.

Enough to make a healthy leech throw up

Here I was, clinging to my usual stem of grass just above the mud, enjoying the warm sun filtering down through the duckweed and waiting for an unsuspecting set of toes and a nice, juicy ankle.  Without warning, the water exploded around me, mud and slime churning up as my favorite stem was jerked rudely upward into hot, dry air and biting, unfiltered sunlight.  Next instant, I was buried under mud in a smelly container with a bunch of other creatures who were just minding their own business.  And the company!  Now, I’m not a snob, but we are members of the ancient family of Hirudinea, and we are NOT bottom dwellers.  We may be may be annelids, but we are NOT worms and, despite what they say about us, you will only find us sliming around in the mud when we ABSOLUTELY have to!  They are down there, and we are up here – we’re blood-suckers, not mud-suckers.

Anyway, I endured this ignomius treatment and absolutely LOW company with as much grace as possible, then was rudely squashed between two enormous, absolutely FILTHY fingers and roughly dropped into a glass box.  As I plummeted through a thick layer of duckweed (really, some people just do NOT care about housekeeping), I could see20150829_170211 an incredibly ugly humanoid staring at me through the glass. I wish I could have sucked him dry when he pulled me out of that smelly plastic container and dropped me into his tank, but he was hiding like the coward and bully that he was, safe behind that glass20150918_104830(1) wall.  So I took myself off to a quiet corner to think (no, I was NOT sulking!  Leeches do not sulk.  We are just quiet, contemplative creatures that find our own company quite sufficient at times, thank you.)


After some thought and exploration, I may not have landed so badly.  Been talking with the locals – not a bad lot for having been raised in ditches and puddles, though no literary or cultural giants.  Bunch of Brachiura worms spend all day in the mud, waving their their gills.  Pretty, sinuous and graceful creatures.  I know they’re exotic, but I get tired of hearing how their great-great-grandfathers rode over from Australia in a shipment of guppies. Really, invasive species should know their PLACE and not be giving themselves airs as if they belonged here and had any class.

The hydras tell me about a lovely cloud of hay water that descends and how the whole world blooms with schools of bacteria and juicy fat copepods, but it sounds like a bunch of religious claptrap to me.  I work for my living – do you think it’s easy hanging by your suckers all day in the hot, sunny water, waiting for the moment when dinner splashes by?  Besides, they’re a bit weird and not too bright.  What do you expect from a creature with a neural net about as complicated as the spark plug on a Sopwith Camel???  And there’s one of them that spends his time getting high on every oxygen bubble that floats by.  Says he’s “In Recovery”, whatever THAT means.  When I saw him, he was sprawled all over his latest bubble and couldn’t even keep his stalk straight.HYDRA AND AIR BUBBLE MAG 20150826_123901

And that other one!  Just because she got “discovered” and spent an hour being photographed and got her image on the internet is no reason to put in airs! Spending your day squashed under a microscope slide and having every tentacle laid out for everyone to gawk at is nothing to be proud of, and it’s not as if she did anything to earn it.  Just happening to be lolling about of the right duckweed root when the tweezers come down doesn’t mean you did anything to DESERVE all that attention!child-crooss-eyed-with-snailsSpeaking of food, a nasty realization just percolated through my ganglia:  there’s nothing with BLOOD in this tank!  Hemolymph doesn’t cut it – thin, nasty stuff, and even though you’re a simple critter, you can’t do everything properly with ONE bodily fluid!  And you think I’m going to running after some scrap of protoplasm  or eat half-dead organic junk like some rotifer?  It’s undignified!  Even though they make those weirdohippie hydras fat and happy, I can’t eat copepods!  I have a problem!


Heads Up in Hostile Territory

Things are looking up a bit.  When that big, ugly biped (dubbed him The Proprietor, though I could run things better with one sucker tied behind my back) wasn’t looking, I went out exploring.   Those snails aren’t doing their job, and spend half their time falling off plants instead of keeping the windows clean, and…  Anyway, I digress.

Over the edge of the tank,  and out into enemy territory.  Across a desk, and found just what I thought I saw through all that algae on the glass.  It’s a computer, with those lovely, sensitive keys that work if you just touch them!  Alright, you might wonder what a leech would do with a computer?  Well, let me tell you, there is a long history of literate and intelligent invertebrates.  Unfortunately, they could never get the word out until the biped vertebrates invented the writing machine.  But true creativity was (just barely) possible, in the early days before digital came along.


The first true voice of the invertebrate world called himself Archy, and though the world would call him a cockroach, he was a noble and philosophical soul (, and sacrificed so much to be heard!  As if spending all night bruising your carapace on typewriter keys wasn’t bad enough, think about being isolated in a dusty office with a bunch of Socialist newsies, your only companion a spaced-out kitty who thought she was Cleopatra and nothing to eat for years and years but old library paste! The biped named Don DID help Archy in his bumbling way – at least he kept the paste pots filled.

Such sacrifice to let the world know of your genius!

Now there’s none of this clambering up a typewriter, trying to keep from getting your antennae or suckers stuck in all those rods and levers, then jumping off the top and hoping that you come crashing down on just the right key!!!  The old way might be OK for insects, but we leeches just aren’t built for jumping and crashing – we just tend to plop messily in a nasty little puddle of guts and, well, brains if we had them (but that’s another story).  But I do run on…

None of that today, though.  Particularly for us slender and silkily dampish, sinuously flexible lovely beings, just a flick on the key or a touch on the screen and we’re off to the races and on to the big time.  Today, even a leech of modest literary capability can find himself with a national audience.  After all, if Fox News can do it, why not a leech with just a LITTLE more brain and much more class?

The Road to Literary Fame

Time to go over the top! So I strap on my imaginary tin helmet, gather courage and all of my dignity, and charge up the glass into that dry, biting air (SO much nicer when it’s dissolved in lovely warm water), over the edge of my new home, across the desk (Dust and WHO KNOWS WHAT all over my mucus!  I’d fire that housekeeper!), up a cable, and onto the keys!  A little lunge, click, click, and words appear!  I’m into his system!  The great hacker leech has arrived! And there’s even music!  I do a lovely, slippery, sinuous Charleston with “%5” as my own little dance floor!LEECH 05So you think you’re pretty tolerant because you don’t openly snigger at fat people?  Let me tell you, you don’t know what it’s like to be green till you’ve BEEN green!  You have prejudices! The same folk who vote for integrated classrooms and put tofu on the menu  and recycle their newspapers would step on a slug or squash a snail without thinking and believe they were doing a GOOD thing by mushing another hapless invertebrate!  World, look out!

The little folk of the Valley are going to have a voice!  It will ring out from beneath the waters of every swamp, pond and puddle!  Humanoids wonder about what’s beneath Jupiter’s clouds, but they drain a green puddle and don’t even THINK about it.  The only people among those bony behemoths are children (if they don’t squish our soft parts

Small friend:
Small friend: “Look, Mommy!”

too hard) and they forget when they grow up. They lose the joy of befriending a nice snail or starfish when they start to get big and worry about a new bicycle. I’ll be off to work every night to tell the planet about our world beneath the green layer, and next time YOU see a blob of green Spirogyra, you’ll pause and say, “Oh, how beautiful!” instead of scraping it off your rubbers!  GetImage.aspxAnd we’re hard working professional creatures – we’ve been helping heal your disgusting big bodies for CENTURIES!  And this is the treatment we get in RETURN!

(Still haven’t solved the food problem.  I’m starting to get a bit too slender.  I’m even starting to dream about McBloodburgers!  How could a leech sink so low?)


The blood (I mean dinner) situation has been resolved!  I am so full – and what a great meal!

It’s really all hard work and careful watchfullness, and though I don’t like to boast, I am one of the best!  I could have been snoozing on a leaf, or wasting an evening waving back and forth with those empty-headed worms, but I was on alert like the good, hard-working parasite that I am.  And when The Proprietor fell asleep at his microscope, I was waiting and ready! Over the edge, a little slither, slither across the desk, then so, so quietly up on his finger and oh, what a tasty meal!  Mrs. Proprietor has been filling him up with something called the Mediterranean Diet – lovely olive oil, good Italian whole-wheat pasta, and lots of organic fruits, and not a whiff of the Big M!  I will be so slim and healthy, and my slime will positively glow!  What a lovely leech I will be, feeding like this!  And just a few tiny, tiny drops that he’ll never miss!  Then back home (yes, it has become home) for a a little swim off the diving board…SHOULD I TRY A DIVE CEthen a nap in the sun…A NAP IN THE SUN

So that’s life these days.  Oops!  Aquarium light just went out and all is dark.  Time to get to work.  Keep connected and I’ll share the world of the pond with you….

Proprietor’s Note:

Little stinker!  So he thinks I don’t know what he’s up to?  Wet tracks all over my screen, sucker prints on my keys, and that little tail disappearing over the edge first thing in the morning?  No wonder he’s never around in the daytime!

Oh, well, he reminds me to dust my screen, and he’s kind of amusing.  He better watch his step, or it’s back to the pond, and it’s COLD outside!

Funny how this finger keeps itching….



Archy and Mehitabel at Work

The author offers sincere apologies to the ghost of Don Marquis.

See “Invertebrates In Ink” and:

The Literary Leech

Microscopes and Other Equipment: Restoring the Nikon S

Nikon S-KT
1971 Nikon Model S with KT base, providing Kohler Illumination

The Nikon Model S microscopes from the 1970s, with their quality all-metal construction, excellent craftsmanship and optics, shiny black finish, and multiplicity of options, are still some of the best general-purpose and research microscopes ever developed:

“…the black Nikon S scopes that were first released in the 1960s … are still hallmarks of versatility and quality and have developed a faithful following despite their age. The Nikon SBR was the first, the Nikon Ske and Nikon Skt, were research grade microscopes, and the Nikon Lke with constant Kohler illumination and changeable nosepiece was the top of the line in its day. Nikon designed these scopes with many interchangeable components so that a scope could be customized to every application…”
(Courtesy Fred Durette,

The image above is of my own, recently-restored Nikon S-Kt.  However, this wonderful line of microscopes suffered from one serious design flaw, ironically resulting from one of the few plastic (nylon) parts on this otherwise superbly-crafted instrument:  the infamous Nikon S fine focus spur gear.

This nylon gear, used probably because it never needed lubricating, was constructed as a 107-tooth nylon ring surrounding an inner brass collar pressed onto an eccentrically-mounted steel ball bearing on the fine focus shaft.  Since molded plastic materials were not as structurally sophisticated in the 1970s as they are today, with time this plastic ring contracted and broke, disabling the fine focus mechanism and causing thousands of service calls.  Brass replacements for this gear could be purchased as aftermarket parts for a number of years, but are almost completely unavailable today.  Hundreds of otherwise excellent Model S microscopes have been shelved or broken up for parts for no reason other than this one defect.  This problem is well known in microscopy circles.

I purchased this Model S-Kt scope with lenses on eBay for $75.  It had not been used for 20 years and was advertised as “…not turning on…”  This problem was fixed in 10 minutes by re-soldering a flange on the bulb, and, apart from some dirt on the knobs and a few paint chips and dust, the instrument was in beautiful shape.  However, the fine focus mechanism moved unevenly, revealing a broken fine focus gear and a potentially serious problem.

Changing The Model S Spur Gear

After a week of fruitless phone calls all over the United States, Bob Lair of ScopeDoc, a fellow member of the international Amateur Microscopy group, suggested that I contact an old associate, Mark Morris:

Mark Morris, Micro-Maintenance Inc.
2648 Simpson Circle, Duluth, Ga 30096
770 449-6896 (Fax 770 446-5761)

Mark, the only North American source for these gears, works with a machinist who hand manufactures replacement Nikon S brass spur gears,  and had two remaining from the last lot – probably a significant fraction of the world’s remaining supply.  With $50 and a few days’ patience, I had my gear .  Replacing it turned out to be a relatively simple process, as described below.  Do NOT lose any of the irreplaceable small parts!

Note: This gear was one of the last of a group manufactured in 2004.  The latest batch of hand-crafted gears is more expensive to manufacture and will sell at $75 US (Joan Morris, personal communication).

First, viewing the microscope from the back, remove the right fine focus knob by removing the single screw holding it onto the shaft.  Then, using a specialized spanner inserted into the two small holes in the underlying cover (a lens wrench works well), unscrew the circular cover that forms the center of the larger coarse focus knob:

Removing the fine focus mechanism cover
Removing the fine focus mechanism cover

After the cover is removed, the spur gear is visible, with the crack appearing as a dark space in the white nylon gear:

Cracked spur gear and brass replacement
Cracked spur gear and brass replacement

Then remove the left fine focus knob, being careful to retain in the proper order the two small washers beneath the knob:

Left side of fine focus shaft with two small washers
Left side of fine focus shaft with two small washers

Once the left knob and washers are removed, the whole fine focus shaft and spur gear assembly should easily pull out from the right side:NIKON SPUR GEAR 10

Spur gear assembly
Spur gear assembly

The toothed nylon gear should pull off easily, but this leaves the underlying brass collar that has been firmly pressed onto the eccentrically-mounted steel ball bearing.  This can be cut off carefully with a small carborundum wheel mounted on a Dremel tool.

Cutting the brass inner collar of the spur gear
Cutting the brass inner collar of the spur gear

Cut carefully through this brass ring until only a thin bridge of brass remains, then crack the bridge by inserting a screwdriver into the slot and twisting it.  This protects the bearing from being cut by the grinding wheel.  The collar can then be readily slid off:

Spur gear collar after removal. Note the thin inner bridge of brass that has been cracked to remove the collar.
Spur gear collar after removal. Note the thin inner bridge of brass that has been cracked to remove the collar.

The final challenge is to press on the new brass gear.  This can be done by one of three methods:  Tapping the carefully-sized new cylindrical gear onto the bearing (recommended by Mark), using concentric rings to press or tap on the gear, or employing the thermal expansion method.

A combination method proved most effective.  I cooled the bearing for an hour in the freezer, then heated the cylindrical brass gear on a stove element until it was just uncomfortably hot.  Holding the gear with a tea towel, it could be pressed half way onto the cold bearing with firm finger pressure before temperatures equalized and it seized in place.  This oriented the two pieces.  I then placed the assembly flat on the edge of a counter and tapped the gear gently with a small hammer, rotating with each tap.  This proved effective but slow, so I grasped the edges of the gear and bearing between the jaws of a pair of plumber’s pliers.  Applying gentle pressure at intervals around the gear and bearing, rotating the assembly between each gentle squeeze, quickly but gently pressed the gear onto the bearing and leveled the two surfaces.  The gear proved to be very carefully machined and was easily guided to a smooth and precise fit.

The whole shaft and gear assembly was then carefully cleaned, lubricated with a sparse amount of grease, and reinserted into the scope.  The washers, right cover, and knobs were quickly replaced.  The reassembled fine focusing mechanism now moved silkily, with absolutely no hesitation.

NIKON S-Kt 02 BW.jpg
The finished scope



Note:  Original Nikon S System advertisement 1963:
(Anal. Chem., 1963, 35, pp 91A–91A)Free first page

Microscopes and Other Equipment: Restoring the Nikon S

Moss, Rain, and Rocks: Pond Life Outside the Pond

Desmid Gel on Moss, Mt. Richards.
Desmid Gel on Moss, Mt. Richards.

Not all of water-based life lives in ponds, ditches, and lakes.  In this damp climate, there is water everywhere except in the driest season – within cracks in tree bark, in moss, on rain-drenched leaves, and in the damp of the forest floor.


While hiking around midnight in our constant rain on misty Mount Richards, I noted this gelatinous mass on a dripping mossy rock face:

Desmid Gel. Note water droplets on lower surface of the mass - this is a constantly dripping rock face in winter and spring.
Desmid Gel. Note water droplets on lower surface of the mass – this is a constantly dripping rock face in winter and spring.

I believe that this is the desmid Mesotaenium, which is known to grown on mossy rocks with a constant water supply:

It has been documented in both England and the Netherlands, and consists of dispersed cells in a mucilaginous matrix forming a gelatinous mass among moss stalks. Note that there


Mesotaenium in Gel Matrix, 40X. Note smaller desmid species barely visible in background.

is also a population of smaller algae, likely also desmids, seen at 400X  together with one Mesotaenium cell.

Mesotanenium and Smaller Desmid Species.
Mesotanenium and Smaller Desmid Species.

Images of the larger desmid species show shape and chloroplast structure typical of Mestaerium:

Mesotaenium, 400X. Note chloroplast structure and surrounding cell wall.

In some images, the refractile mucilaginous sheath surrounding individual clusters of desmid cells is visible:


This is a link to an older but informative article discussing mucilaginous algae as well as 1980s theorizing about diatom movement – 1980s knowledge and one of the more pompous and verbose articles in the literature (“…palmelloid cell masses of coccolithophorids…”), but also very infomative:


Note on pages 120-122 the comments on the significant role played by mucilaginous gels in Mesotaenium and other algae:BONEY MUCILAGE 01

as well as the potential size of some of these sheets of algal gel habitats:


Nice result from my mountainous wanderings with a headlamp at midnight in the cloud layer.


If you explore all possible things that are common and unnoticed, the world will never a boring place.

I have one old car –  a beloved 1993 Mercury Topaz bought for a dollar from my uncle’s estate, 22 years old and with only 60,000 miles on the odometer.   From the era when Detroit thought it could build cars from rubber bands and old sardine tins, it is grossly underpowered, occasionally blows a tired old engine seal, and has a back window that repeatedly leaks – yet it has proven to be a faithful and undemanding servant.  After a few months of continuous rain, its rubber window seals turn green with a film of algae.

This note is for Kelly Averill Savino, to answer your question regarding my picture of ugly little desmid blobs in a pot.
The implication behind your question was “Whatizzit? What desmid mass? Surely not those ugly little blobs that you can barely see??? Wipe them off with a Kleenex, PLEASE!”
That’s the point. Desmids (see Wim’s lovely images: are some of the most beautiful objects that you can see under the microscope. However, in their appearance in nature, they can take the prize in the category of “ugly-little-masses-that-only-a-mother-could-love.”
The point of my post on Triple Tree (and one of the least appealing pictures ever posted on this site) is that they are easily missed and not very appealing when you find them.
Gel formation and mucilaginous coatings are very important for algae, and some desmids seem to take this to the extreme. I find them on midnight rainy mountain walks, forming ugly little snotty blobs on shaded, drippy rock faces. Or, as in the last note, forming equally ugly, snotty little blobs in an overwatered nursery pot.
However, under the microscope, these nasty little blobs form an amazing and complex little world. The one in the images, formed on lichen on a drippy rock face, contained gels within a gel – a mucilaginous blob, containing multiple tiny gel spheres of higher refractive index (looking rather like a tapioca pudding at low magnification), each containing its tiny cluster of bright green desmid cells.
This gelatinous little world also hosted other organisms, in this case a tiny nematode and a branching black mold.
The Boney article, though dated and verbose, contains much thought-provoking information on algal mucins:
There is much interest in algal mucins in the health and biotech field because they apparently have medicinal and nutritional properties:
Moss, Rain, and Rocks: Pond Life Outside the Pond