Wirt Robinson.

The elements of electricity online

. (page 32 of 46)
Online LibraryWirt RobinsonThe elements of electricity → online text (page 32 of 46)
Font size
QR-code for this ebook

candle-power if the total light emitted were spread uniformly over
the surface of a sphere with the lamp as a center.

510. Photometry. Measurement of candle-power is made by
photometers. Various kinds of these instruments are described in
detail in the electrical handbooks. In brief, they consist of an
arrangement by which a beam from the standard falls side by side
on a screen with a beam from the light being measured. One of
the lights is shifted back and forth until the illumination on the
adjoining surfaces is the same. When this equality of illumination
has been attained, then, since the intensity of illumination varies
inversely as the square of the distance from the source, the candle-
power of the lamps are to each other directly as the squares of
their respective distances from the screen.

511. Life of Incandescent Lamp. The life of an ordinary 16
candle-power incandescent lamp may exceed 2000 hours. How-
ever, the candle-power of a lamp, although slightly above normal
for the first fifty hours, decreases steadily thereafter, and it is laid
down as a rule that the smashing point of the lamp is reached when
its candle-power has fallen to 80 per cent of its rated value. This,
on an average, is after about 600 hours' use. The useful life
depends greatly upon the accuracy with which the voltage is


regulated. It is stated that an increase of three per cent in the
voltage will shorten the life of a lamp one-half. On the other hand,
a decrease of ten per cent in the voltage reduces the candle-power
47 per cent.

512. Efficiency. The efficiency of an incandescent lamp should
be measured by the light produced by the expenditure of a certain
amount of power, that is, by the candle-power per watt. In practice
however, a custom the reverse of this has arisen and the efficiency
of a lamp is given by stating the number of watts required to pro-
duce one candle-power. In this case, the greater the number of
watts, the less the efficiency of the lamp. The hot resistance of an
ordinary 110 volt, 16 candle-power lamp is 220 ohms. The current
through the lamp is therefore one-half ampere, and the power con-
sumed is 110x1/2=55 watts. The wattage per candle-power is
therefore 55/16 = 3.1. By increasing the voltage, more light is
produced and the efficiency may be made 2.7 watts per candle-
power, but in this case the life of the lamp is very much shortened
(Par. 511).

The efficiency of the Nernst lamp is 1.75 watts per candle-
power and that of the tungsten lamp is 1.5 watts or even less.

513. Control of Light. An objection to the incandescent lamp
is that it can not easily be turned down. We shall see later that
if a large number of closely-grouped lamps, such as are used in
illuminating the stage of a theatre, be run by alternating cur-
rent, it is possible to turn them down simultaneously by a simple
piece of apparatus (Par. 621), but it is not practicable to apply
this to individual lamps. It is theoretically possible to insert
in series with a lamp a variable resistance, a rheostat (Par. 302),
by which the current, and consequently the light, may be con-
trolled, but the cost and the necessary bulk of such arrangement
prohibit its use.

514. Grouping of Incandescent Lamps. Assuming that in
transmitting electrical power from the generator to the spot where
the power is to be used the principles outlined in Par. 502 have
been observed, in utilizing this power for purposes of illumination,
the lamps may be grouped either in series or in parallel, though the
latter arrangement is by far the commoner of the two. Among
the considerations which lead to the selection of one grouping
in preference to the other, the principal are the distances by which


the individual lamps are separated and the nature of the current,
whether direct or alternating.

If the lamps are to be located close together, as in the illumina-
tion of the rooms of a building, the parallel arrangement should be
followed. A striking advantage of this arrangement is the in-
dependence of the several lamps and the automatic adjustment
of the current to suit the demands made upon it. The following
will make this clear.

In Fig. 246, G represents a generator constructed, as will be ex-
plained in Part V, so as to maintain a constant difference of poten-

Fig. 246.

tial between the mains A and B. L represents a number of lamps
arranged in parallel -between these mains. Suppose the resistance
of a lamp to be 220 ohms, and the difference of potential between
A and B to be 110 volts. If one lamp be turned on, the current
through it will be / = E/R = 110/220 = 1/2 ampere. If four lamps
be turned on, the resistance between A and B is reduced to 220/4
= 55 ohms and the current is now 110/55=2 amperes, but since
there are four paths, one-fourth of the total current, or one-half
ampere, flows through each so that each lamp gets its proper cur-
rent. So long as the difference of potential between A and B is
maintained, each lamp when turned on will receive its proper
current, and whether it be turned off or on will not interfere with
the remaining lamps.

There are still other parallel arrangements, such as the three-
wire system, the five- wire system, etc., in which more than two
mains are used, but explanation of these is deferred until the
machines supplying the currents for these systems have been

If the lamps are to be widely scattered, as in street illumination,
they should be arranged in series and supplied by a constant current
generator. At the Military Academy the roads are lighted by
incandescent lamps, each requiring three amperes at 50 volts, and
arranged in series, 50 in a circuit. The generator must, therefore,
supply three amperes at a pressure of 2500 volts. Were these



lamps arranged in parallel, the mains would have to carry, for a
portion of their length at least, a current of 150 amperes.
Since the lamps are in series, should one burn out, the remainder
would ordinarily be extinguished. To avoid
this, an arrangement shown diagrammatically
in Fig. 247 is employed. From the lamp socket
proper there extend downward two brass
springs C and D, shaped so that they press
tightly together like a pair of spring tweezers.
They are kept from actual contact by a thin
sheet E of mica, or of similar insulating ma-
terial, which is inserted between them. When
in position, these upper springs make contact
with corresponding springs A and B, by which
the current is brought in and taken out. Should
the lamp burn out, breaking the circuit, the
voltage between C and D, which up to this time
had been 50, immediately mounts to 2500. This
is sufficient to pierce the sheet of mica E, burn
Fig. 247. it out, and re-establish the circuit.

515. The Arc Lamp. The electric arc was described in Par. 485
and later its use in the electric furnace was explained. It was also
pointed out that not until the comparatively recent development
of machinery for supplying the necessary current did it become
possible to utilize it. It was discovered by Davy in 1808. By
means of a battery of 2000 cells and with charcoal electrodes he
produced an arc four inches long and of very great brilliancy.
Thirty-five years later Foucault substituted the more compact gas
coke for the charcoal used by Davy. Carbon is still the principal
material used, although certain other substances have recently
been introduced (Par. 523).

Arc lamps may be grouped in series or in parallel, the same con-
siderations governing as explained in the preceding paragraph.
Since they are most largely used for external illumination, and
also since they require a much larger current than does the incan-
descent lamp, they are usually arranged in series.

516. The Carbons. The carbons for use in the arc lights are
made with the greatest care. They are made from lampblack, or
from gas coke, or from a similar coke produced in refining certain


petroleum products. These forms of carbon are ground to a very
fine powder, passed through a bolting cloth like that used in the
manufacture of flour, and intimately mixed with granulated pitch
which is warmed enough to cause the ingredients to adhere. The
mixture is then cooled and again ground to a fine powder and
passed through the bolting cloth. The resulting meal is formed
into rods, either by being compressed between steel molds by
hydraulic pressure or by being forced through a die and emerging
in a continuous piece which is cut up into the required lengths.
The rods are then placed in layers in a furnace, the layers being
separated and covered by sand, and they are then heated and
maintained at a high temperature for from ten days to two weeks.
In this process a good many are spoiled by warping. The carbons
thus prepared are frequently copper-plated. The coating of cop-
per strengthens the rods, prevents chipping and the formation of
dust, and adds about one-fifth to the life of the carbon, but its
main object is to obtain a better electrical contact. The molded
carbons are the most largely used but, mainly because of the re-
mains of the web along the sides, they are not exactly cylindrical
and can not be used in certain forms of arc lamps described later
(Par. 521). The pressed carbons are perfectly cylindrical and
when necessary can also be made in the form of a tube for the
manufacture of cored carbons. The average arc light carbons are
one-half inch in diameter and vary from six to twelve or more
inches in length. Their average resistance is 0.15 ohm per foot.
Carbons for search lights may be as much as two inches in diam-

517. Requirements of Arc Lamp Mechanism. The mechanism
of an arc lamp must automatically perform the following functions:

(a) When the current is turned on, it must bring the carbons
into contact.

(b) It must then "strike" the arc by separating the carbons the
proper distance.

(c) As the carbons consume away, it must feed them together.

(d) If the carbons approach too close, it must separate them.

(e) If the arc goes out it must restrike it.

(f ) In a series arrangement, if the carbon burns out or breaks,
a cut-out switch must operate to shunt the current by the dis-
abled lamp.



When it is realized that the mechanical and electrical arrange-
ments by which the foregoing objects are attained must differ
according as the lamps are connected in series or in parallel, and
also must differ according as direct or alternating current is to be
used, it will be seen that the kinds of lamps are very numerous.
We can do no more, therefore, than outline the principle of opera-
tion of a few typical forms.

518. The Clutch. In all ordinary direct current arc lamps, the
positive carbon is the upper one. There are two reasons for this.
The first and principal is because eighty-five per cent of the light
produced by the arc is emitted from the crater at the tip of the
positive carbon and therefore this must be above so as to throw
its illumination downwards. The second is because the positive
carbon is consumed more than twice as rapidly as the negative, or
in open arcs at the rate of about one and a half inches per hour
and by placing it above it is in the best position to be fed by grav-
ity. These considerations do not apply to alternating current
lamps, nor to certain projectors and search lights. In this last
class it is desirable that the crater should face
the reflector and lie in its focus; the carbons are
accordingly often placed horizontally, or one
horizontal and the other vertical, and both may
be fed automatically or by hand. The arrange-
ment by which the upper carbon is lifted and
held at the proper distance from the lower
and by which it is allowed to slide down as
it burns away, is called the clutch. There are
many forms of clutches. Some operate like the
tongs used in hoisting stones and close when
they are raised but open when they are lowered.
^^^^ A very simple form is shown in section in Fig.

248. This consists of a metal plate A pierced
I I with a circular hole slightly larger in diameter

^ ' than the carbon holder which passes through it.

One end of this plate fits loosely in the jaws B


Fi 248

of the lifting apparatus. As B rises, the plate A is canted and
thus grasps the rod. When B is lowered, A strikes the stop C
and is brought to a horizontal position, thus releasing the carbon
which slips down.



519. Constant Potential Arc Lamp. As stated above, arc lamps
may be run in series or in parallel. The series arrangement is by
far the more common, but the parallel grouping is also frequently
employed, especially for interior

illumination. In this case the
lamps are connected across mains
between which a constant differ-
ence of potential is maintained.
One of these lamps is shown dia-
grammatically in Fig. 249. With
the carbons in contact, when the
switch S is closed the current
enters at A, passes through the
resistance R, thence through the
solenoid C to the upper carbon,
down this to the lower carbon and
out by B. The current passing
through C causes it to suck up the
plunger and, through the clutch,
to raise the upper carbon and thus
strike the arc. As the carbons
burn away, the arc gets longer
and its resistance increases. This
reduces the current, and the lift-
ing power of C grows less until finally it can no longer support
the plunger and the carbon and they fall. The clutch strikes the
stop and releases the carbon which slides down, shortening the
arc. This increases the current and the plunger is again drawn
up, and so on.

Without the resistance R, the result of closing the switch with
the carbons in contact would be in the nature of a short circuit
(Par. 306). This resistance steadies the current by preventing
violent fluctuations and it is therefore a "ballast" as described in
Par. 508.

520. Constant Current Arc Lamp. For operating arc lamps
in series, the generator and its regulator are designed so as to
furnish a constant current, therefore, whether the arc be long
or short the current is the same. On this account, resistance in
series with the lamp is not required. Furthermore, the arrange-
ment described in the preceding paragraph could not be used, for

Fig. 249.



the pull of the solenoid upon its plunger being constant, the
carbon would not feed. For such lamps the so-called "differen-
tial" mechanism is employed. This is
shown diagrammatically in Fig. 250.
With the carbons in contact, the open-
ing of the switch S causes the current
entering at A to pass around the sole-
noid to the point C, thence to the upper
carbon, thence to the lower and out by
B. The passage of this current actuates
the clutch and strikes the arc. To
cause the carbon to feed, a differential
coil is taken off at the point C and con-
nected at D, that is, it is in shunt with
the arc. This coil is of many turns of
fine wire and is wrapped in opposite
direction to, and inside of the first, but
for clearness is represented in the dia-
gram as being below. The two coils
being wrapped in opposite directions,
the pull upon the solenoid plunger is
due to the difference of the ampere

turns in the two. With the carbons in contact, the difference of
potential between E and D is very little, therefore, a very
small current flows through the differential coil. As the car-
bons draw farther and farther apart, the resistance, and
consequently the difference of potential, between E and D in-
creases. This causes an increasing current to flow through the
differential coil and weakens more and more the pull on the
plunger. A point is finally reached when the plunger drops and
the carbon feeds.

521. The Enclosed Arc. The wasting away of the carbons in
the ordinary arc lamp is mainly due to the combination of the
white hot carbon vapor with the oxygen of the air. It is not
practicable to enclose the carbons in air-tight globes but in recent
years there has been introduced a form of arc lamp in which the
arc is surrounded by a globe so fitted that the admission of air is
reduced to a minimum, and in these the life of the carbons is very
greatly prolonged, the consumption being reduced from 1.5 inches
per hour to less than one-tenth of an inch. In addition to the sav-


ing in carbons, there is a very great saving in labor since the lamps,
instead of having to be "trimmed" or supplied with fresh carbons
daily, average over 100 hours and may be run as long as 200 hours
without attention. Other advantages are a steadier light and
absence of the hissing noise of the open arcs. The principal chan-
nel for the admission of air to the arc is the space around the carbon
since this latter must be free to be moved by the lamp mechanism.
To reduce this, the carbons must fit the opening very accurately,
for which reason, as already mentioned (Par. 516), pressed car-
bons are used instead of the molded.

522. The Flaming Arc. With the common arc light, the carbons
are from one-sixteenth to less than a quarter of an inch apart and
the greater part of the light is emitted from the incandescent car-
bons, although the maximum heat is developed within the arc
itself (Par. 485). If it were possible to suspend within this arc
a non-combustible solid, like the mantle of the Welsbach burner,
it would be heated to incandescence and the heat energy of
the arc would be converted into light energy. This object is
partially realized in the so-called flaming arcs. In these, the
positive carbon is either impregnated with certain salts of cal-
cium or of magnesium or has a core filled with these salts. The
vapor produced when these salts are volatilized is highly heated
and emits a powerful reddish yellow light, and since it con-
ducts it also permits the carbons to be separated by upwards
of an inch. They need not be raised to such a high temperature
as in the common arc lamps and therefore their life is longer.
The efficiency of these lamps is at least three times that of the
common form.

Instead of the carbons being in the same vertical line, they are
sometimes arranged both pointing downward like the letter V,
the arc being at the vertex. In this way, neither carbon screens
the other and both tips throw their light down. There is a tend-
ency, however, for the arc to ascend between the carbons. This
is corrected by arranging a magnetic field, similar to the magnetic
blow-out (Par. 485), but only strong enough to keep the arc down
at the tips of the carbons.

An additional advantage of this arrangement is that the slag
formed by the fusion of the impregnating salts drops off and does
not clog the tips of the carbons with a non-conducting glassy


523. The Magnetite Arc Lamp. The magnetite arc lamp, but
recently developed and used with direct current only, resembles
the flaming arc lamp in that the chief source of light is the arc
which is an inch or more in length. It differs from other arc lamps
in that little or no light is given off by the electrodes, also that the
maximum amount of light is developed at the negative end of the
arc. The positive electrode is of copper and is of such size that
the heat developed is conducted away so that the electrode is not
consumed. The negative electrode is a thin steel tube, the size
and shape of an ordinary carbon. It is packed with a mixture of
powdered magnetite, Fe 3 4 , and oxides of chromium and titanium.
The magnetic oxide renders the electrode a conductor, the remain-
ing oxides not conducting until they have been heated. The oxide
of titanium imparts the luminosity to the arc; the oxide of chro-
mium increases the life of the electrode. An eight-inch electrode
in such a lamp with a current of 4 amperes at a pressure of 80 volts
will burn for upwards of 200 hours. Since the constituents of the
electrode are oxides, there is no combustion and the arc is not
enclosed. These oxides, however, are volatilized and condense
immediately beyond the limits of the arc in a reddish soot which
if not removed soon covers globes, reflectors, etc. It is therefore
necessary in these lamps to provide some form of chimney with a
strong draught by which this deposit is carried off.

524. Efficiency of Arc Lights. The efficiency of an arc light is
much greater than that of an incandescent lamp. The common
arc lamp, carrying a current of about 10 amperes at a pressure of
about 50 volts, develops 2000 candle-power in the zone of maxi-
mum luminosity, or, in round numbers, one candle-power per 0.25
watt. The mean spherical candle-power (Par. 509) is, however,
considerably less than 2000. The larger search lights, taking 200
amperes at 60 volts, develop nearly eight candle-power per watt,
but it must be noted that there is a lack of agreement and much
uncertainty as to the measurement of the candle-power of these
powerful lights.

525. Luminous Vapor Lamps. Suppose a high voltage, such as
that produced by an induction coil, be applied to two platinum
wires sealed into the opposite ends of a glass tube, and suppose
that at the same time an air pump be set to work to exhaust the
air from the tube. If the wires be not too far apart, sparks will


pass between them, but as the air is exhausted, these sparks lose
their definiteness and finally take the form of an effulgence or glow
completely filling the tube. The color of this glow varies with the
nature of the gas enclosed in the tube. For air, it is rosy pink; for
nitrogen, yellow; for carbon dioxide, white. At this stage the rare-
fied gas has great conductivity. If the exhaustion of the tube be
continued, the conductivity decreases, the luminous column begins
to break up in striae and finally disappears. When the pressure
has been reduced to about one-millionth of an atmosphere, the
glass itself begins to phosphoresce. Beyond this, the resistance
becomes so great that no current can be sent through the tube.
There is therefore a stage of rarefaction in which gases conduct
electricity and in doing so emit light, and these effects diminish
if the pressure be increased or decreased from what it is at this
stage. Explanation of this will be given later; for the time being
it will suffice to say that when highly rarefied these gases ionize
and therefore conduct (Par. 276). If the exhaustion be complete,
there are no ions left and consequently a vacuum is a non-con-

The foregoing is the principle of the luminous vapor lamps, two
of which we shall now describe. Their luminous efficiency is very
high, for while in the ordinary carbon filament lamp less than one
per cent of the total energy expended is developed as light, in
these luminous vapor lamps twenty per cent or more is so develop-
ed. They have not yet been made in small units but are rather
used for general illumination of large spaces.

Fig. 251.

526. The Moore Light. The apparatus for producing this
light, shown diagrammatically in Fig. 251, takes the form of an
exhausted glass tube one and three-quarters inches in diameter and
of any length up to 200 feet. It is usually suspended along the


ceiling of the room to be illuminated. When in operation, it emits
a soft, diffused light, without flickering or unsteadiness, the color
varying, as stated in the preceding paragraph, according to the
gas contained in the tube. To produce a light of fifteen candle-
power per running foot, about 70 volts per foot are required, the
corresponding current being about one-third of an ampere. By
increasing the voltage, the candle-power can be raised to a maxi-
mum of thirty per foot. A tube 100 feet long requires 7150 volts.
This high voltage is obtained from an alternating current by means
of a simple step up transformer, as shown in the figure above.

As the lamp is used, the gas in the tube appears to be consumed
and the rarefaction increases. This causes the resistance to in-
crease. It therefore becomes necessary to introduce from time to
time minute amounts of gas, and a simple and effective automatic
valve has been devised for this purpose.

527. The Cooper Hewitt Mercury Vapor Lamp. If in a glass
tube, otherwise vacuous, there be introduced a small amount of
mercury, the vacuous space would quickly become filled with the
vapor of mercury (Par. 277). An electric current passed through
this vapor would cause it to glow with a greenish light. This
arrangement would not differ in principle from the Moore light,

Online LibraryWirt RobinsonThe elements of electricity → online text (page 32 of 46)