Oct., 1896. ] Series Arc Lighting. 303

E L E C T R I C A L S E C T I O N .

Stated Meeting, September 22, 1896:

MR. CLAYTON W. PIKE, President, in the chair.

A N E W SYSTEM OF S E R I E S A R C L I G H T I N G .

BY THOMAS SPENCER.

A n y one who has wa tched the development of arc l ight- ing wi th in recent years cannot have failed to notice the gradual , bu t radical, changes tha t have taken place in the methods of del iver ing l igh t from a central stat ion. Origin- ally the idea seems to have been to break up the source of current supply into a grea t n u m b e r of small units . Whe the r , f rom an eng ineer ing s tandpoint , this idea was a well-worked-out scheme, or s imply the ou tg rowth of circum- stances, is, I think, very easy to a n s w e r ; for i t is usual to begin on as small a scale as possible, so that , in ease of failure, the financial loss will be as small as possible. These first a t t empts proving successful, the conservat ive feeling, which to a grea t ex tent accompanies the i nves tmen t of money, caused a cont inuance in the course which had been found practical.

Perhaps in no branch of the electrical i ndus t ry is this so apparent as in tha t of arc l ight ing. However, at present there seems to be a general t endency towards a more eco- nomical system, as is a t tes ted by the number of large arc l igh t dynamos tha t are now replacing the numerous small mach ines ; fur ther , by the rapid rise of arc l igh t ing from cons tan t potent ia l D. C. circuits, and also by the gen- eral in teres t mani fes ted in a l t e rna t ing arc l ight ing. The last named, which, of course, signifies the use of constant potential , is the direct ion towards which eve ry th ing seems to be tending, and will be the sys tem most general ly em- ployed in the future , unless some efficient method of direct ly conver t ing hea t into electr ici ty is discovered.

There is no ques t ion bu t t ha t a s ta t ion equipped with large uni t s and supply ing one kind of c u r r e n t - - a n d tha t a

3o4 Spencer : [ J. F. I.,

curren t which can most easi ly be cont ro l led- - i s the mos t efficient, Such is the a l te rna t ing current . As far as the a l te rna t ing arc lamp is concerned, i t has some poin ts which make it s l ight ly inferior to the direct-current l amp; b u t exper ience has shown tha t these ob jec t ions are not so seri- ous as at first appea red .

Recent ly , a t t empt s have been made to use a device b y means of which the cur ren t is c o m m u t a t e d into a f luc tua t ing current , hav ing a lways one direction, bu t not wi th conspic- uous success. There are several such plants in England, I believe, and, f rom some of the crit icisms, I should j u d g e tha t the sys tem is by no means as efficient as it should be to be genera l ly in t roduced, especial ly in this country , where , I am sorry to say, a piece of appara tus is ap t to fall i n t o the hands of men who fail to g ive it even the mos t neces- sary care. As for the advan tages ga ined by the use of such a device, there is no doub t b u t tha t the efficiency of the are is improved. F rom m y exper ience I should j u d g e that , as the current is a f luc tua t ing one, the arc would still b e noisy, a l though less so than the unrectif ied a l t e rna t i ng current .

As I have said before, the ob jec t ions to the a l ternat inff a rc have not p roved to be as ser ious as they at first ap- peared. There is a g rowing t endency towards the use of the a l t e rna t ing arc for a11 kinds of l ighting, and th i s is especial ly marked in s t ree t l ight ing. T he sys t em w h i c h , up to the p resen t t ime, has been in use may be d e s c r i b e d as follows : Each lamp is b u r n e d separa te ly f rom a 3o- or 33-volt t r ans former (see Fig. z). The a m o u n t of wire u s e d in this sys t em is not genera l ly greater , and of ten less, than tha t requ i red in the old direct-current series sys tem. T h i s sys tem works ve ry well and has m a n y good features, per- haps the g rea tes t of which are tha t each lamp is independ- ent of the rest, and tha t the pressure on the l amp is low. Fur the rmore , the l amp has only one series spool of coarse wire, and is free f rom shun t spools and cut-outs.

There are a g rea t m a n y sys t ems of this sort opera t ing in this country, and some, I believe, have been in t roduced in England. The only ob jec t ion to this plan is its high

Oct., I896.] Series Arc Lighting. 305

first cost, occasioned by the necess i ty of a separa te trans- former wi th each lamp. To overcome this objec t ion and at the same t ime be in posi t ion to use the same kind of lamp, Mr. Wi l l i am Smi th Hor ry has devised w h a t he calls his " R e a c t i v e S y s t e m " of arc l ighting, which I wish to b r ing before you this evening. Mr. H o r r y couples his l amps in series direct ly in the p r imary circuit, doing away with the separa te t ransformers . Now, any one who has a t t emp ted to run arc l amps in series, which regu la te only by var ia t ions in cur ren t ( that is, have only a series spool), knows very well that they will not operate. The reason for this is readi ly apparent . An arc lamp should depend for its regula t ion on

T T T

F ~ .1 ̧.

someth ing pract ical ly affected only by the burn ing of its own arc. This is ev ident ly not the case wi th the current where lamps are pu t in series across a cons tan t potent ia l circuit. In general, when lamps are run in series, we mus t r e g u l a t e b y the change of E.M.F. a round the are; tha t is, r egu la te wi th a shun t spoo 1. A lamp, of course, s imilar to the regular series arc lamp, migh t be used if it were pro- vided wi th some device to keep the cur ren t constant , or a lamp similar to tha t used on s t ree t ra i lway circuits could be devised, bu t in all of these cases we would have a much more compl ica ted lamp than that u sed by Mr. Horry.

The principle govern ing Mr. Hor ry ' s sys tem is briefly this : VOL. CXLII. No. 850. 20

306 Spencer : [J. F. I.,

In s h u n t wi th each lamp is placed a small coil of the type known as au to- t rans former (Fi E. 2). The coil A is in ser ies with the lamp, and B in shunt . W i t h this device the cur- rent in the l amp will a lways be g rea te r than tha t in the line. Cons ide r ing the cur ren t in the line as constant , which is prac t ica l ly the case where a large n u m b e r of l amps are in series, we must , as shown above, by some means out- side of the l amp itself, cause a change in the cur ren t pass- ing t h r o u g h the series spool of the lamp, in order to enable the l amp to feed. This is accompl ished by mak ing the a m o u n t of iron in the coil small, so tha t it will soon b e c o m e s a t u r a t e d and cause the coil to leak, lowering the pressure ,

Co~.t. ,5 Co%.L ~ Coil.

5 which in turn decreases the current and feeds the lamp, This can be more easi ly under s tood in the fol lowing way : The re la t ion

in a t ransformer , where there is no leakage be t ween the p r imary and secondary , is well known. (Fleming, Al ter - na t ing Curren t Transformer , Vol. I, p. 273. ) Here N1 and /1 are respec t ive ly the n u m b e r of p r imary turns and the m a x i m u m current , and N~ and /2 represen t the same for the secondary, B the m a x i m u m induction, Z the equ iva len t per- meabi l i ty , ! the length of the magne t i c circuit , and a i t s section.

Oct ., 1896 .]

Series Arc Lighting.

307

Now the secondary E.M.F . is directly proportional to B,or B = Ky, where y is the secondary E.M.F and K a con-stant. If in the first equation we write x for Iz and substi-tute the value of B we have

and

N1 2 11 2 -Nz' xz + ( 1Ky lz\4z11al '

l

,a a

JNow, if we suppose I1 constant, and write

a and b being

a- N,I,Nz

b= 4-rN,I,K

constants, we have_ xz

yaI

z +a

This is easily recognized as the equation of an ellipse-Plotting this (Fig . 3), it is plain that, when the current issmall, the E.M.F . varies very little with an increase of cur-rent ; that is, it behaves in this region as a constant poten-tial circuit . Referring again to our last equation, we seethat when

It s

becomes greater, that is, when the equivalent magnetic resist-ance becomes greater, the minor axis of the ellipse becomessmaller. This is accomplished, as before stated, by dimin-ishing the amount of iron in the core of the coil, and, as aresult, we will have a coil whose characteristic is an ellipse,

xzz +

yz

r 4z N,I1 1 z

_Nz~ --K-

11-

308 Spencer. [ J. F. 1.,

as r ep resen ted by the dot ted lines. In this case the maxi- m u m E.M.P. will not be so large ; but , as the ellipse is near ly flat, there is a large region over which the coil p roduces . prac t ica l ly cons tan t potent ial .

T h e r e are a few o ther fea tures in this sys t em w o r t h y of ment ion. The coil is so propor t ioned tha t when the carbons are consumed, and the arc in consequence breaks, the whole cur ren t is forced th rough tha t por t ion in shun t wi th the lamp. T h e choking effect in the coil is abou t the same as the E,M.F. taken b y the lamp when burn ing ; in fact, it is so ,close to this tha t 3o per cent. of all the lamps in circui t may

Y

rl.

0 0

• \ \

6econdar~ Gurren~c. X

be pu t ou t w i thou t sensibly affect ing the amp~remeter in the circuit . The ~oil then not only acts as a cut out, bu t also as a regulator .

The advan tage which this sys tem shares wi th that in which l amps are run from separa te t ransformers , is tha t i t is poss ible to run l amps of wide ly different candle-power on the same circuit, which is accompl ished by changing the coil to conform wi th the lamp. In this part icular , of course, i t is an improvemen t over the ordinary series system, wi th s h u n t coil regulat ion. The n u m b e r of lamps which it is poss ib le to run in series from a i ,ooo-volt circuit or t rans-

Oct., I896. ] Manufacture o f White Lead. 309

former depends upon the candle-power of the lamps. T h e coils, as now const ructed , will, in this ease, take care of twenty-nine 2,000 candle-power lamps. Mr. Hor ry has a lso devised a swi tch-board to be worked in eonneet ion wi th a special t ransformer in such a way as to g ive va ry ing pri- mary E.M.F.'s, so tha t any number of lamps can be run ir~ series wi th in reasonable limits.

As for the pract ical working of this sys tem, I would say tha t it is in use in several places in thi:s country. As far as I know, it has been very sat isfactory, and there is l i t t le doub t tha t a grea t fu ture awai ts it.

ON A

C H E M I C A L S E C T I O N .

Staled Meetizg, held September I5, z896.

DR. H. F. KELLER, President, in the Chair.

N E W P R O C E S S ~OR THE M A N U F A C T U R E OF W H I T E L E A D .

BY MR. WM. TATHAM, Member of the Institute.

The process I am abou t to descr ibe is the jo in t invent ion of Win. P. Ta tham, of this city, and myself . Before go ing into details, I should like to g ive a very br ief rdsumd of w h a t has been done heretofore.

The oldest me thod of making whi te lead, c o m m o n l y t e rmed the Du tch process, is based upon the fact that , i f metal l ic lead be exposed to a hea ted a tmosphere of w a t e r vapor, carbonic acid, acetic acid vapor and oxygen, i t i s conver ted into a bas ic lead carbonate . These condi t ions are obta ined, as is well known, by placing pieces of lead, of convenien t size, in ear then pots conta in ing a small quan- t i ty of acetic acid, and packing these pots in a bed of s p e n t tan or horse manure . The fe rmenta t ion of the tan, or manure , genera tes carbonic acid and wate r vapor , while the hea t l ibera ted by the process of f~rmenta t ion causes the acetic acid to vaporize, and the lead is t he reby corroded.