Fluorescent lamps

The action of a fluorescent lamp depends on the discharge of a current through a gas or a
vapour at a low pressure. If a tube containing a vapour has an electrode at each end, a
current will flow through the vapour provided electrons are emitted by one electrode (the
cathode) and collected by the other (the anode). Electrons will be emitted if the potential
gradient from anode to cathode is great enough; the potential difference required to cause
emission decreases as the temperature of the cathode increases, and therefore lamps
designed to operate at normal mains voltage have cathodes which are heated to a dull red
heat. They are known as ‘hot cathode’ lamps.
Even when the cathode is heated, a voltage has to be applied between the electrodes to
start the discharge, and the minimum voltage needed is known as the striking voltage.

After the discharge has started a voltage is still needed between the electrodes to maintain
the discharge, but the maintaining voltage is less than the striking voltage.
A current flowing through a gas or vapour at low pressure causes the gas or vapour to
emit radiation at wavelengths which depend both on the nature of the vapour and on its
pressure. An incandescent lamp gives out light energy at all wavelengths in the spectral
range, whereas a fluorescent lamp gives it out at certain discrete wavelengths only. The
wavelength of the radiation emitted may be in the visible spectrum or above it or below
it, and one of the functions of the lamp is to convert all the primary radiation into useful
visible radiation.
A fluorescent lamp consists of a long glass tube containing a mixture of mercury
vapour and argon gas at a pressure of 2 to 5mm mercury. When the lamp is cold, the
mercury is in the form of small globules on the tube surface, and the argon is needed to
start the discharge. As soon as the discharge starts the temperature rises sufficiently to
vaporise the mercury which then takes over the conduction of practically the whole
current. At either end of the tube, there is an electrode made of a tungsten filament coated
with an alkaline earth metal having suitable electron emission properties. Each electrode
acts as cathode and anode on alternate half cycles of the a.c. supply. Anode plates in the
form of metal fins are provided round each electrode to assist it in collecting electrons
during the half cycle in which it acts as anode. The inside of the tube is coated with a
fluorescent powder. A fluorescent material is one which has the property of absorbing
radiation at one wavelength and emitting radiation over a band of wavelengths in another
region of the spectrum. It emits radiation only while receiving it; a material which
continues to emit after the incident radiation has ceased is called phosphorescent and it is
an unfortunate confusion that the fluorescent materials used in commercial lamp
manufacture are commonly called phosphors.
Thus in the fluorescent lamp the radiation emitted by the current discharge through the
mercury vapour is absorbed by the fluorescent coating which
Figure 7.3

then emits a different radiation. The fluorescent coating is most susceptible to excitation
by ultraviolet radiation, and it is the need to have the radiation from the mercury in this
region that determines the operating pressure. The secondary radiation emitted by the
coating is in the visible spectrum and its colour depends on the material used for the
coating. So many fluorescent materials are now known that it is possible to obtain almost
any colour, including an almost exact reproduction of daylight.
The circuit needed to operate such a lamp is shown in Figure 7.3. As we have
explained, the voltage required to maintain the discharge is less than that required to start
it, and therefore once the discharge has started the voltage across the lamp must be
reduced. If it were not reduced the current through the lamp would go on increasing until
the lamp was destroyed. The necessary reduction is achieved by a series ballast which
takes the form of an inductance or choke. Initially, when there is no discharge through the
lamp, the entire voltage of the mains is applied across the electrodes. As soon as the
discharge is established, current flows through the lamp and choke in series, a potential
difference is developed across the choke, and the voltage across the electrodes is reduced
by the voltage across the choke.
The circuit also includes a starter switch which consists of a small neon glow lamp and
a bi-metal strip. When the lamp is cold, the bi-metal switch is open so that the whole of
the mains voltage appears across the neon glow lamp which discharges. The heat of the
discharge heats the bi-metal until the contacts on the end of the bi-metal close. There is
now a circuit formed by both electrodes of the main lamp, the bi-metal and the choke. A
small current flows through this circuit and heats the electrodes. However, the bi-metal
short circuits the neon, which ceases to glow and, therefore, to heat the bi-metal. In
consequence, the bi-metal cools and after a time it opens again. This interrupts the circuit
which, being highly inductive, responds with a sharp voltage rise across the switch. Since
the switch is in parallel with the lamp the voltage rise is also applied across the lamp and
is sufficient to start the discharge. The choke now takes the normal current and reduces
the voltage across both lamp and switch. This reduced voltage is not enough to start
another glow at the switch, but if for any reason the main discharge fails to start, then
mains voltage again appears at the glow lamp and the sequence starts again. This happens
if the choke is faulty and reduces the voltage across the lamp below that required to
maintain the discharge. When the lamp has started the electrodes are kept hot by the
current through the lamp which also flows through the electrodes.
The starting sequence described takes a few seconds. This delay can be avoided with
the instant start circuit shown in Figure 7.4. The electrodes are supplied from low-voltage
secondaries of a transformer, and carry the full heating current continuously. A metallic
strip runs the whole length of the tube and close to it, and is connected to earth. When the
lamp is switched on a capacitive current flows from the electrodes to the earthed strip and
is just sufficient to ionize the gas in the tube and thus enables an arc to strike
Design of electrical services for buildings 108

Figure 7.4 Quick-start circuit
from end to end. As soon as the arc has struck, the choke reduces the voltage across the
lamp to its normal operating value and the current in the heaters also reduces as a result
of this. Although it is possible to use ordinary lamps with quick-start circuits, it is not
wise to do so because the heaters carry a higher current than when they are used with
switch start circuits. For this reason, only tubes designed for quick-start should be used in
fittings having quick start circuits.
Figure 7.3 shows a small capacitor across the switch contacts. The capacitor is inserted
to suppress radio interference. Figures 7.3 and 7.4 both show a larger capacitor across the
entire circuit. This one is included to correct the power factor, which would otherwise be
unacceptably low because of the inductive choke.
The capacitors, switch and choke are usually housed within the luminaire which holds
the lamp. It is important that the luminaires used are suitable for the lamps they are
intended to employ. The current taken by a luminaire with fluorescent lamps is inductive
with a power factor of about 0.8. The switches controlling fluorescent lights, therefore,
have to break inductive circuits and must be capable of withstanding the voltage rise
which occurs when an inductive circuit is broken. The voltage rise can cause arcing
across the switch contacts, and if this is serious enough and occurs often enough it will
destroy the switch. Most modern switches are capable of breaking an inductive current of
the value at which the switch is rated, but the older switches designed only for
incandescent lights had to be de-rated when they were used on circuits serving
fluorescent lights.
Lighting 109

There are some occasions when the switchgear for the luminaire, that is to say the
capacitors, switch and choke, is mounted remotely from the luminaire. One case in the
author’s experience was in a swimming pool where the lamps were mounted behind a
pelmet around the perimeter of the hall. The architectural design was such that there was
no room for ordinary luminaires containing the gear, and the lamps were simply clipped
to the pelmet. Each end of each lamp was plugged into a two-pin socket from which
cables ran to a remote cupboard. Inside the cupboard there was a series of trays each of
which contained the control gear for one lamp. Another case was in an air-conditioned
computer room. Much of the heat lost from a fluorescent light is generated in the control
gear rather than in the lamp, and in order to keep the heat load on the air-conditioning
plant as low as possible, it was decided to keep the control gear out of the air-conditioned
room. The method used for doing this was similar to that in the previous example.
Fluorescent lamps have a lower surface brightness than incandescent ones and have a
higher efficiency in terms of light energy given out per unit of electrical energy
consumed.
Fluorescent tubes can be made in the shape of a U or an annulus. The production
process is more difficult and therefore these tubes are more expensive than the commoner
linear ones. However, advances in production techniques have made it possible to
produce tubes so compact that a complete lamp can be little bigger than a large
incandescent lamp. Small diameter tubes formed into an annulus have found application
in domestic and commercial lighting and can be used in bulkhead luminaires not much
larger than the corresponding incandescent ones. A further development is a U tube bent
again and enclosed in an outer bulb which looks like an incandescent lamp. It has also
proved possible to make the control gear so small that it can fit inside the outer bulb
which can then have an end cap suitable for plugging into a standard incandescent type
lampholder. The Philips SL and Thorn 2D lamps are of this type and can be used to
replace incandescent lamps without any change to the fittings.
Fluorescent lamps cannot be dimmed simply by the insertion of a resistance to reduce
the current through the lamp. If a reduced current flows through the tube, it does not heat
the cathode enough to bring about emission of electrons. Successful dimming requires
that the electrodes be permanently heated while the tube current is varied, and this can be
done by supplying the electrodes with a separate heating current from a low-voltage
transformer. A suitable circuit is shown in Figure 7.5 and the similarity with the quickstart
circuit of Figure 7.4 can be noted. The transformer provides a permanent heating
current which is independent of the current through the tube and the latter current is
varied by the resistor. In this way, variation of the tube current does not cut off electron
emission from the cathode.
Design of electrical services for buildings 110

The tube current can also be varied by a thyristor. The arrangement is similar to Figure
7.5, but a thyristor circuit takes the place of the resistor. While permanent heating current
is circulated to the electrodes, the thyristor controls the proportion of each cycle of the
alternating supply during which striking voltage is applied to the tube. This determines
the length of time in each cycle during which current flows through the tube, and hence
the light output. Tri-phosphor tubes save 20 per cent energy consumption when compared
to conventional tubes. This saving is further increased to up to 40 per cent if electronic
ballasts are used.