April 1959 Popular Electronics
Wax nostalgic about and learn from the history of early electronics. See articles
published October 1954 - April 1985. All copyrights are hereby acknowledged.
Before there were CFL light
bulbs with complicated electronic circuits for generating the requisite high voltages
without a transformer, there were just the familiar straight (and sometimes circular)
fluorescent bulbs that use a simple ballast arrangement and a built-in switch in
the bulb base. As with compact fluorescent (CFL) lights, very few people understood
how they worked. Most knew that the 8-foot-long T-12 bulbs (the large diameter ones
used in commercial buildings) made a really cool implosion sound when they broke
- usually intentionally (by people like me) since the glass tubes are amazingly
tough. I remember many moons ago, between high school and the time I enlisted in
the USAF, while I was an apprentice electrician working on a renovation job at a
public school (I took
electrical vocational courses for 3 years in high school), a co-go-fer and I
spent weeks wiring fluorescent fixtures all throughout the building. We worked atop
a tall scaffolding that was on wheels, pulling ourselves around the room by grabbing
the suspended ceiling grid. When the foreman wasn't around, we relieved our utter
boredom by "accidently" dropping the fluorescent bulbs onto the concrete floor.
If done just right so that the bulb hit the floor in a perfectly vertical position,
it would bounce up two or three feet and implode in mid-air. The acoustics of the
big, empty room, devoid thus far of sound-absorbing ceiling tiles, really enhanced
the effect... but I digress. This article does a nice job of explaining how the
non-curly-Q fluorescent bulbs work - with and without external starters.
The Electronics of Fluorescent Lamps
By Edgar D. Morgan
Fluorescent lamps, once a novelty, are now an accepted part of the American scene.
Their gentle blue-white glow is encountered everywhere, from the tops of skyscrapers
to the depths of the New York subway system. However, the operating principles of
fluorescent lamps are little understood even by technicians.
Invisible Light. Many people know that fluorescent lamps are
among our most efficient light producers, and that they operate with a mercury vapor
arc. But few realize that over 80% of the radiation produced by that arc is in the
ultraviolet region and invisible to the human eye, and that every effort is made
to keep as much energy as possible in the invisible ultraviolet end of the spectrum.
Sounds foolish, doesn't it?
Cutaway view of a fluorescent lamp. All fluorescent lamps are
constructed as shown here. Only the glass tube changes in size and shape from model
The visible light actually comes from chemical compounds coated on the inside
of the glass tube. Called phosphors, these compounds have the property of emitting
visible light when they are excited by ultraviolet radiation. They have been termed
"light transformers" because of their ability to absorb energy at one wavelength
and radiate it at another.
The fluorescent lamp depends upon ionization for the production of the necessary
ultraviolet arc. Here's how it's done. The free electrons in the gas are accelerated
by an applied voltage, and each time a collision occurs between an electron and
a gas molecule, one or more additional electrons are displaced. These electrons
in turn, are accelerated enough to repeat the process on other molecules, and a
chain reaction takes place.
As each molecule returns to a stable state, it gives off its excess energy in
the form of radiation. It's the frequency of the radiation that determines whether
visible or invisible light will be obtained. In commercial lamps, the pressure of
the gas sealed in the lamp is adjusted very carefully so that nearly all of the
radiation occurs at one given ultraviolet wavelength, 2537 Angstrom units. This
frequency is selected for optimum excitation of the tube's phosphor coating.
Each chemical compound in the phosphor coating radiates light at a certain wavelength.
For instance, zinc silicate releases its radiation as green light, cadmium borate
radiates a predominately pink color, and calcium tungstate when excited gives off
blue light. By carefully blending these and other compounds, almost any desired
color can be obtained.
Warming to the Job. The electrons required to facilitate the
starting of the arc are provided by coated tungsten filaments in the ends of the
fluorescent tube - similar to the filament in an ordinary vacuum tube. Once the
arc is achieved, the heated filaments are no longer required and are automatically
switched off. A small amount of argon or krypton gas present in the tube facilitates
the initial arcing, which also serves to evaporate the globule of mercury in the
tube. From this point on, the arc is basically mercury vapor.
There are two pieces of auxiliary apparatus necessary to operate a fluorescent
lamp. One of these is a starter which acts as the automatic filament circuit switch
mentioned above. The second additional element required is the ballast, which serves
as a choke coil to regulate arc current as well as an autotransformer to provide
the high voltage kick needed to start the arc.
An Inductive Kick.
Fig. 1. Basic fluorescent lighting circuit in general use
only on small desk lamps. A manual switch performs the starting function.
Fig. 2. This circuit is sometimes used to correct the poor
power factor in the circuit of Fig. 1. The ballast is shown here functioning
as autotransformer as well as choke coil.
The basic lamp circuit is shown in Fig. 1. When the switch is closed, the
ballast and the two filament windings are placed in series across the applied voltage,
and the filaments heat. When the switch is released a moment later, the inductive
kick of the iron-core ballast coil causes a momentary voltage surge across the lamp
which starts the arc. As the ballast is now in series with the lamp, the arc current
is limited by the impedance of the ballast. This simple circuit is widely used on
small desk lamps, but has too many disadvantages for general lighting use.
One of the disadvantages of the single lamp circuit is its poor power factor.
As the circuit is primarily inductive, due to the ballast, the current and voltage
have a phase relationship which makes for inefficiency in light output compared
to current drawn. A partial solution to this problem is the circuit shown in Fig. 2.
Here the starter is an automatic device although its function is the same as
the switch in Fig. 1. The power factor in this unit is improved by a shunt
capacitor. As a capacitor and a coil have an opposite effect on power factor, one
offsets the other; and the resistor serves to bleed off any charge which remains
on the capacitor. The circuit of Fig. 2 has been drawn to indicate that the
ballast is also serving as an autotransformer. The number and size of the lamps
used determines whether this is necessary or not.
Factors and Flickers. Another objectionable feature of the single
lamp circuit is its flicker, due to the 60-cycle line. Since incandescent lamp filaments
operate at a very high temperature, there isn't time for them to cool sufficiently
from cycle to cycle for the variation of light to be seen. The fluorescent lamp
must extinguish and restrike its arc 120 times per second as the voltage reverses
polarity. This causes a disturbing stroboscopic effect around machinery with cyclical
Probably the most common circuit in use today is shown in Fig. 3. This is
a two-lamp circuit and corrects several of the disadvantages inherent in single
lamp setups. The lamp indicated as inductive is connected in the same way as the
lamp in Fig. 1. The other lamp, however, has a capacitor in series with its
ballast. This serves to change its phase relations so that its current leads rather
than lags the voltage, and corrects the overall power factor. Thus, the two lamps
operate more efficiently as a unit.
Fig. 3. For the control of two lamps, this circuit is in
wide-spread use and has excellent power factor characteristics.
Fig. 4. Typical glow-type starting circuit used in home-type
How a Glow Starter Works
The addition of this series capacitor, though, produces another problem. When
starting, the capacitive circuit sometimes limits the current required to preheat
the filaments. This effect is overcome by adding another compensating coil in series
with the starter on this lamp only.
Use of this type of two-lamp circuit is also beneficial in overcoming flicker.
Because the lamps operate out of phase, they reach their peaks of illumination at
different times and the combined light is relatively free from disturbance.
The Starter Story. One of the most fascinating pieces of auxiliary
equipment is the fluorescent starter. Its task is to close the circuit containing
the filaments when voltage is first applied, and after a preheating period of several
seconds, to open the filament circuit and keep it open as long as the lamp remains
on. A bimetallic strip is used which bends when heated, and serves as sort of a
The most common type of starter is the glow starter. The entire unit is sealed
within a small glass envelope containing neon or argon gas and connected as shown
in Fig. 4. A bimetallic strip controls a contact which is normally open.
When first turned on, the full line voltage is applied across the glow lamp,
causing the gas to ionize and conduct. The heat of the ionized gas is sufficient
to cause the bimetal strip to close the filament circuit. The contact closing shorts
out the glow discharge and the bimetal begins to cool. After cooling (which takes
long enough to preheat the filaments satisfactorily), the contact within the starter
opens and the fluorescent lights. The starter is now across the lamp voltage, which
is not enough to ionize the neon gas, and so the unit is inoperative as long as
the lamp remains on.
The glow-type starter consumes no energy from the circuit after the starting
period is over. Its timing is not accurate, however, and as it is difficult to maintain
the proper gas pressure over a long period, sometimes the timing tends to become
The line voltage produces a glow discharge between the bimetallic strip and the
fixed contact (A); the heat from the glow actuates the bimetallic strip, the contacts
close and the filament preheating begins (B); this shorts out the glow discharge,
the bimetallic strip cools and the contacts open (C). The resulting inductive kick
from the ballast then starts the tube.
A great many of our present-day starters incorporate modifications such as a
manual reset. If the lamp does not light after repeated attempts by the starter,
it ceases functioning until the trouble is corrected and the starter is reset by
pushing a spring-loaded button. There are others that use different contacts for
restarting so that it isn't necessary to wait for a bimetal strip to cool completely
before it can recycle.
The starter shown in Fig. 4 also contains a capacitor across the lamp contacts,
which acts to suppress radio interference.
Manufacturers of fluorescent lamps publish handbooks which elaborate on these
principles. They also give many more lighting circuits and their applications. The
very fact that fluorescent lamps are so readily accepted and so seldom studied in
detail is in itself a fine tribute to their efficiency and dependability.
Posted November 22, 2021
(updated from original post on 11/11/2013)