Adjusting BOOK TWO. Temperature. South Bend WatchCo. South Bend Watch Company. С. T. Higginbotham. Subject. Continued By. C onsulting Superintendent

Transcription

1 BOOK TWO Subject Temperature Adjusting Continued By С. T. Higginbotham C onsulting Superintendent South Bend WatchCo. Compliments of South Bend Watch Company South Bend, Indiana

2 SHORT TALKS to WATCH MAKERS BOOK TWO SUBJECT T E M P E R A T U R Ę A D JU S T IN G ( CONTINUED ) > BY С. T. HiGGINBOTHAM CONSULTING SUPERINTENDENT SO U TH BEND W A T C H C O. COMPLIMENTS OF S o u t h B e n d W a t c h C o m p a n y SOUTH BEND, INDIANA

3 / H E adjustm ent of w atches is com m only Ё classed under three h e ad s: Isochronism, tem p eratu rę and position. S trictly speaking th e first em braces th e other tw o, but as com m only understood it m eans sim ply adjusting th e hairspring and balance to each other so th a t th e la tte r shall perform its long and short arcs in th e sam e tim e while th e w atch is in a horizontal position. T hen pósition adjustm ent steps in to eąualize th e tim e of vibration for all the other positions. F in ally tem peraturę adjustm ent produces a uniform ity in th e tim e of th e vibrations in h eat and cold. This adjustm ent is e n tirely d istinct from th e other two so far as the m eans used to secure it are concerned. I t is of this adjustm ent th a t th e present article treats. B efore attem p tin g it, th e structure, properties and action of the balance should be carefully studied. C H A R L E S T. H IG G IN B O T H A M M r. H igginbotham probably has a m ore thorough and broader ехрегіепсе in th e science of m aking high grade w atches th an any other m an in the U nited States. A lthough sixty-five years of age, he tak es an active p a rt in all m atters th a t concern th e production of a m ovem ent th a t is gaining a reputation as a m ost accurate and durable tim epiece th e South Bend. S T R U C T U R E. T he chronom eter, or more p roperly speaking, com pensation balance is composed of tw o m etals; brass and steel. The arm s are of steel; th e rim of steel and brass lam in ated ; th e inside steel, th e outside brass. The proportional expansion of m etals, as has been explained in book one, is called the coefficient of expansion. Brass, being an alloy, its coefficient of expansion varies according to the proportion of th e different m etals composing the alloy. Zinc has about th ree tim es g re a te r coefficient of expansion th an steel, conseąuen tly, th e larg er th e per cent. of zinc the g re ater the expansion. Brass used in balances has its com ponent m etals so proportioned th a t it is capable of being brought to a very hard eon-

4 dition and still possess about tw ice th e coefficient of expansion of steel. I f we secure to g eth er a strip of brass and a strip of steel, and subm it it to h e at or cold, th e brass expanding or contracting m ore th an the steel, will cause the b ar to deflect as shown by F u j. 1. в P R O P E R T IE S. The less the thickness of the rim th e m ore activ e; which is to say th e g re a te r its deflection; b u t if it is m ade too th in it would be deflected by centrifical force the tendency of m etals to fly from th e center when in circular m otion when the balance is in rapid m otion and th e tim ekeeping ąualities of th e w atch be m ore or less im paired. The jo in t betw een th e m etals m ust be perfect. Blow holes are fa ta l to uniform action. For this reason the old m ethod of soldering the two th e broken lines in Fig. 1. Should eith er m etal be too g re atly in excess of th e other no deflection will tak e place. The m etal having th e g re ater mass will com pletely оѵегсоте the other carry- F i<j. 2. < I i... ~ 1. : ing it forw ard in a direct line ju st as though the b ar was composed of a single m etal. See Fig. 2. I t follows th en th a t in order to secure th e best action in th e rim th e m etals m ust bear a certain proportion to each other in dim ension and also in density and structure. The proper proportions are % steel % brass. This is to say, th e sectional area of th e rim should show these proportions. The com bined thickness of the rim of an 18s balance should be about 127 hundredths of a m illim eter. 4 с m etals to g eth er w ith silver solder has been superceded by m aking a sm elted joint. The m etals m ust be of uniform thickness and tem per th ro u g h o u t; otherw ise the deflection of both sections of th e rim will not be uniform w ith the result th a t deflection from heat or 5

5 cold would throw th e balance out of poise. See F ig. 3. In this case th e brass and steel are correctly proportioned a t one side which would of course produce th e proper deflection, while a t th e other side th e steel is g re atly in excess of th e brass for th e g reater portion of its length w ith a result th a t th e deflection of the rim a t th a t side would be slight as com pared w ith the opposite side. A C T IO N. The action of h eat expands the m etals of which th e balance is composed. T he arm elongates, throw ing th a t p a rt of th e rim to w hich it is attach ed out from th e center. The rim also elongates b u t the brass on th e outside expanding m ore th an th e steel on th e inside causes it to curve in w ard ly ; this curv atu re increasing from th e arm to th e cut. F ig. 4 will illu strate ju st how this curvature tak es place. The fuli lines show the norm al condition of th e balance; th e broken lines fhe effect of heat. It should be understood th a t in this draw ing, as in th e preceeding ones, the effects illu strated are som ew hat exagerated, which is necessary in order to show th e action clearły. It will be seen by reference to Fig. 4 th a t under th e influence of h eat th e rim recedes from th e cen ter a t A, approaching th e cen ter to a much g re ater extent a t B, while a t С it m aintains a uniform distance. To fu lly understand th e conditions described above it is well to study this draw ing, F ig. 4, carefully. If all the m etal em braced by the balance were concentrated a t A th e result would be th a t it would recede from the cen ter when heated and approach th e center when cooled, while if it w ere concentrated a t В th e opposite condition would be produced, which is to say, it would approach th e center w hen heated and recede from it when cooled. Now if all the m etal w ere concentrated a t С it would m aintain ex actly the sam e distance from th e center w hether h eated or cooled. This featu re is tak en advantage of to secure adjustm ent to tem peratu rę. This adjustm ent is called C om pensation by which nam e it will be referred to in this article in fu turę. 1 desire here to cali atten tio n to a peculiar condition produced by th e deflection o f a lam in ated b alan ce rim : A C U T COM PEN - S A T IO N B A L A N C E CAN O N L Y B E ABSO- L U L E L Y T R U E IN O N E T E M P E R A T U R Ę. This can be readily understood by reference to Fig. 4. 7

6 Position of screws. G O M PE N SA T IO N balances are provided w ith several m ore holes in th e rim th an are needed for th e screws used. This is done to afford m eans for changing th e position of th e screws in order to effect com pensation. F o r сопѵепіепсе of reference we shall num ber th e holes a t each side of the balance beginning a t th e arm w ith num ber 1. m eantim e screws, which are пеѵег changed to o th er holes. W e shall ta k e tw o w atches w ith balances screw ed up exactły as shown in Figs. 5 and 6. These are tw o regular South Bend W atches w ith th e balances sim ply screw ed up for this particu lar test. W e shall give th e rates and errors w ith the screws in position as shown in Figs. 5 screws in th e rim. H oles 1 to 9 inclusive hołd screws, while 10 to 13 inclusive are vacant. This condition we c a li: A li screws b a c k. In F ig. 6, w hich has th e sam e num ber of holes, 2 to 5 inclusive are v acant, while 1 and 6 to 13 inclusive hołd screws. This is c a lle d : Ali screws forw ard. H oles 1 and 7 гесеіѵе the and 6, and also th e rates and errors shown by certain changes in position of th e screws. L et it be understood th a t when we say, for in stan c e : Move No. 9 to No. 10 it m eans th a t a pair of screws are to be m oved; this is to say Nos. 9 a t opposite side of th e rim. I t will be observed th a t Fig. 6 shows b u t h alf a balance. This m ethod is for econom y of space and will be used in all subsequent drawings. The student having equipped him self w ith tem p eratu rę cabinets should m ake trials of several w atches for th e sake of practice in tim ing before attem p tin g adjustm ent. F or this purpose he should rule a few ra te papers. These are papers conveniently ruled for keeping records of watches under tria l; those illustrated

7 7c)>%p Э 0 CDbcl t ( ) Dutę/Tnn.e 7)i ł' f/a > 7un в, / + 00 Z n cl. /7 0 0 p ) ± J t ± / t In d. t t 00 / 4 У V - 7 J lu d. / 7 / 0 0 i ± f l + // t /n d / > г /7 7>W. / 7 +-_.7_ ( o Г ±9/71 +/7.3 1 n d. /7 -//7 Y tt.fr + 7 +/8.7 + /7? W /7 f) + / ( ) / + /S + /3 l a d ł t &ou.th S en.cl N o. $ 4 М И Т 'а п р 9 O a n d fy~0- ł Tr6. /9U) t a l Ł l/a r. /Г г таг Z ± 1 3 ± 37 Z /a d 19 л t\0 i i / \ ł t 70 //.../л J 9 - J t U m Mur 7L_ 5 W J T )SL ± Ж Е / U t e / / 15' Я * * S J zt ЗЯ 70 _/9 /0.za Ж і / я ŚĄp/ T U L Я0 9 0 Z Щ 75, /6 9 / ± 3 3 / / t o oó : o ó ; / oo - + j5l <7/7/4 0O_ - 7 9, '1 A - O t + 7 Ж& -JŁ / ± I N 00 ± 3 /1 o /*> ± ło -$~tv J 3 L N r ± Z, t - «.ST-. -/./ ± 6 U n, d ± łjrlć C lijacl *H) 3 У 17a d 3 0~ 9 4-0e & T 7 ru d 7 ±. / <F O r c i _ Fig. 7Al. Fig. 7В 11

8 in Fig. 7A and 7B will be found convenient for this purpose. T he first colum n to the left is for th e d a te ; th e next for th e tim e when com parison was m ad e; th e next for th e difference in tim e betw een the w atch and the clock or other standard of com parison; th e next for th e gain or loss as com pared w ith th e previous com parison; th e colum n a t th e rig h t for th e tem p eratu rę error. T he tem p eratu rę a t which the w atch was run or other m atter m ay also be inserted in this colum n if desired. W e also enclose in the booklet a sam ple ra te p aper as used in the adjusting d epartm ent of th e South Bend W atch C om pany. This paper is very elastic in ІЦ possibilities. I t can be used for isochronal, tem p eratu rę and position adjustm ents as well as or m any other kinds of rates. Rules to be observed in trials. ^ " y E E th a t the w atch is in perfect order; clean and freshly oiled. If the oil is * a t all thick or viscid the cold tem peratu rę will m ake it m ore glutenous and retard the m otion of the balance. Be sure th a t th e balance has a sufficient am ount of endshake. T he p late and cock, being com posed of brass or nickel, contract m ore in th e cold th an th e staff and as a consequence m ay retard th e m otion of th e balance if th e endshake is too close. A lw ays ta k e th e tim e when th e second hand of th e w atch is a t 60 regardless of w here the second hand of th e clock m ay be. This for two reasons: F irst, should there be any erro r in the 12 spacing or in th e centerin g of th e d ial it would thus be elim inated so far as any dilference when th e com parison was ta k e n. Second, it lessens th e liability of m istakes in tim ing as will become ap p arent late r on. In tim ing, carry th e tic k of th e clock or chronom eter m en tally, keeping th e gaze steadily fixed on th e w atch. A m arinę chronom eter is excellent to use as a tim er for th e reason th a t it beats h alf seconds, but a little practice will enable anyone to estim ate h a lf seconds very closely from a clock beatin g seconds. P ractice counting first. B eginning when the second hand of th e clock is a t 60, say, h alf o n e h alf two h alf th ree h a lf four h a lf five and so on. In tim ing w atches begin to count when the hand of th e w atch is approaching the 60, and about 10 to 15 seconds from it. L e t us assume th a t the hand of th e w atch being tim ed is approaching th e 60 and th e hand of the clock approaching th e 30 w hen th e count is begun. W e b e g in : Five h a lf six h alf seven h alf eig h t h alf nine h alf th irty h a lf one half tw o h a lf. A t this m om ent th e second hand of th e w atch reaches th e 60. It follows th en th a t th e w atch is eith er 32k> seconds slow or th e difference betw een th a t and 60 which is 2 7 ^, fast. A glance a t th e m inutę hands will determ ine w hether it is fast or slow ; it will also determ ine th e num ber of m inutes it is out in case it should be one or m ore. R un th e w atch both in high and Iow tem - peratures in a horizontal position w ith the dial up. T here are tw o reasons for th is : I t is easily com pared w ithout d isturbing; b u t m ore im- 13

9 p o rtan t still it avoids th e introduction of any position errors th a t m ight exist in the w atch. The w atch m ay be run in its own case or it m ay be run in any close shutting receptacle which while not necessarily air tig h t shoidd a t least prevent the free ingress of air. R un th e w atch for 24 hours in each tria l, tim ing it as n early a t same hour as possible. A few m inutes earlier or late r will m ake no difference b u t should th ere be a g re ater am ount it m ight, in which case it is b etter to repeat the trial. Run th e w atch in the high tem p eratu rę first; th en, a fter tim ing, wind it but do not set it. It is best not to run th e risk of disturbing th e rate by setting betw een the trials. The difference in tim e will show the variation. A fter rem oving from the cold cab in et do not open the receptacle or case containing it for a t least an hour. This gives it tim e to a tta in the tem p eratu rę of the air in th e room which is necessary as a p reventative from rust. If the w atch should be opened im m ediately upon being rem oved from th e cold cabinet it would sw eat as it is c a lle d ; th a t is, th e m oisture from th e air would condense upon the steel parts and possibly produce rust. Figuring rates and making alterations. I N m aking out rates we gen erally use signs and abbreviations in order to economize space and to save tim e. The plus sign is used to signify when th e w atch is fast of the clock and the minus sign to indicate when it is 14 slow. The plus sign m ay also be used to indicate w hat is term ed over com pensation and the m inus sign to indicate under compensation-. A w atch w ith th e balance m ade of a single m etal and having a steel hairspring will, if regulated in m ean tem p eratu rę, g ain in th e cold a c d lose in the heat. This we cali under com pensation. W hen a w atch gains in th e h eat and loses in th e cold we say it is over com pensation. A n im portant rule to be observed in figuring rates is : W hen th e signs indicating th e cold and th e h eat rate are alike subtract th e lesser from the g re a te r; the rem ainder is the error. W hen th e signs are opposite add the tw o to g eth er and th e sum will be th e error. A w atch running in both term peratures on a fast ra te but faster in the cold is of course under com pensation because if th a t w atch, having gained, say, 15 seconds in th e cold and 5 seconds in th e heat, had been regulated so as to g ain 5 seconds in th e cold it would, of course, have lost 5 in the h eat. A w atch th erefo re gaining m ore in th e cold th an in th e h eat is under com pensation; gaining less in th e cold th an in the h eat is over com pensation; losing m ore in th e h eat th an in th e cold is under com pensation; losing m ore in the cold than in the h eat is over com pensation. Moving the screws. ц 'tę IIE N a w atch is under com pensation ^ / one or m ore of the screws should be m oved forw ard as indicated on Fig. 8 w here th e dotted line under No. 9 shows th a t th e screw has been m oved forw ard to No

10 Rate papers W hen the w atch is over com pensation one or m ore screws should be m oved back as shown on F ig. 9. M oving a screw forw ard or back on a rim a certain distance does not alw ays produce th e sam e result. T here are m any causes for yariation, one of which I desire to cali particu lar th e rim as illu strated in F ig. 4. It will be read ily seen th a t a lim ited m oyem ent forw ard or back a t С would produce little or no change in com pensation as com pared w ith th e sam e am ount of m oyem ent a t В for th e reason th a t th e p a rt of th e rim a t С changes its distance from th e center o f th e balance very slig h tly,'if a t all. 16 ł T G. 7A is a ra te p aper giying 8 suppo- I -I sititious trials of w atches in which the r rates w idely differ although the errors are alike. L et th e student study this paper carefully and m ake him self fam iliar w ith the m ethod used. W hen he has com pletely m astered it he will be able to m ake out an y ra te th a t m ay be presented. T he trials are supposed to be of 24 hours duratio n each. The first being in high tem p eratu rę and th e second in Iow, w hich we will refer to as h e a t and cold. T he w atch was wound but not set betw een the trials. In the first tria l th e w atch was set exact and was placed in th e heat. W hen tak e n out it was 14 seconds slow. I t was placed in th e cold w ithout resetting and w hen tak e n out was still 14 seconds slow, not haying yaried during the trial. R eferring to colam n 4 we haye carried out for the h eat tria l, m inus 14, for th e cold tria l, 00. T he loss being in th e h e at we have m ark ed it, U nd. 14. In th e second tria l it did not v ary in the h e at but g ained 14 seconds in th e cold which we find m arked, plus 14, in th e 4th colum n. This w atch was also, U nd. 14. In th e th ird tria l the w atch lost 7 seconds in th e h e at and when tak e n from th e cold was found to be correct, it having gained th e 7 seconds, which it had lost preyiously, conseąuently th e 4th colum n is carried out, m inus 7 and plus 7 m aking 14. In th e 4th tria l th e w atch lost 3 seconds in th e h eat and was found to be 8 fast when tak e n from th e cold, it haying g ained, not only the 8 seconds, but also the 3 seconds it has lost 17

11 previously, m aking a gain of 11 seconds. In th e 4th colum n it is carried out, m inus 3 and plus 11, m aking 14. In th e 5th tria! th e w atch gained 3 seconds in th e h eat and was fast 20 in th e cold having gained th e dilference, 17 seconds. The 4th colum n is carried out, plus 3 and plus Fig. 7 В is a ra te paper m ade from actual tests of South Bend fuli p late w atch No In 24 hour trials, th e tem p eratu res being 90 and 40, degrees F., the balance being screwed up as, shown in Fig. 5. In th e first tria l it was set correct. W hen tak en from th e h e at it was slow 9 seconds. W hen tak e n from th e cold it was fast 28 seconds. A dding 9 and 28 gives 37, the am ount it gained in th e cold. A dding 9 and 37 we have its erro r, 46 seconds and as it lost in 17, subtracting 3 from 17 leaves 14 and as the g re atest g ain was in th e cold it is under 14. In the 6th tria l the w atch was.5 (five-tenths') seconds fast when placed in th e h e a t; when tak e n out it was 4 seconds fast, having gained 3.5. W hen tak en from the cold it was 21.5; a gain of 17.5; th e difference being U nd. 14. In th e 7th tria l th e w atch was slow 1.5 when placed in th e h eat and fast 1.5 when tak e n out which was carried out plus 3 in colum n 4. I t was 18.5 when tak e n from the cold, the dilference, plus 17 being carried out in colum n 4. These rates subtracted Іеаѵе U nd. 14. In tria l 8 th e w atch was fast 1 second when placed in th e h e a t and correct when tak e n out, haying lost 1 second, carried out m inus 1. I t was 13 seconds fast when tak e n from th e cold. These two added m ake U nd. 14. th e h e at and gained in the cold it was under com pensation. The 9th screw was m oved to num ber 10 as shown in Fig. 8, this reduced the erro r to 43.5, am ounting to 2.5 seconds. It will n o t be necessary to ехріаіп in d etail all th e trials which follow but it is desirable to cali atten tio n to th e varying effect produced b y m oving the screws on different parts of th e rim. Moving from 9 to 10, as has been seen m ade a dilference of 2.5 seconds; 10 to 11, of 3.5 seconds; 11 to 12, of 3 seconds; 12 to 13, of 3 seconds. In m aking changes it is well to avoid jum ping a screw оѵег 1 or m ore other screws for th e reason th a t it is liable to m ake the balance m ore diffi- 18

12 cult to poise from th e change of w eight. F o r this reason it is desirable to гаоѵе th e screws in seąuence as shown in the alteratio n betw een the 5th and 6th trials, w here 8 is m oved to 12, 6 to 10, 5 to 9 and 4 to 6. This a lteratio n m ade a difference of 29.5 seconds reducing th e erro r to U nd M oving 3 to 4 m ade a difference of.5 only. This slight change was oceasioned by th e slight curvature of th e rim as illu strated in Fig. 4. M oving from 4 to 5 and from 2 to 3 m ade a difference of 4.5 seconds, bringing the w atch from Und. 4 to over.5. This left the screws in the position shown in Fig. 11. Variation in the position of screws. E S P IT E all the care th a t it is possible to Д J use in the m anufacture of balances or th e springing of w atches no en tire lot will come out th e sam e, a fte r com pensation, so fa r as th e position of th e screws in th e balance are concerned. Fig. 6 shows a balance in w atch num ber w ith all screws forw ard. It showed a com pensation erro r F frrck^^seco n d s. M oving 6 to 4 reduced it 3 seconds. M oving.4 to 3 as shown in Fig. 9 m ade no change. F in ally, when com pensation was secured th e screws occupied the position shown in Fig

13 Biblioteka UJK Kielce UJK : r--; 4-.