US570394A - Hair-spring for watches - Google Patents

Description

(No Model.)

W. W. GRISOOM.

HAIRSPRING FOR WATCHES.

No. 570,394. x Patented 00 12. 27, 1896.

wvewboz Elttmcwup UNITED STATES PATENT OFFICE.

IVILLIAM IVOODNUTT GRISCOM, OF IIAYERFORD, PENNSYLVANIA.

HAIR-SPRING FOR WATCHES.

SPECIFICATION forming part of Letters Patent No. 570,394, dated October 27, 1896.

Application filed October 9, 1895. Serial No. 565,183. (No model.)

To all whom it nut l (,OTLCC/IL.

Be it known that I, WILLIAM 'WooDNUT'r GRISCOM, a citizen of the United States, re siding at Haverford, Montgomery county, State of Pennsylvania, have invented certain new and useful Improvements in atches, of which the following is a specification.

My invention relates to a method of correcting the vertical-position errors of watches and analogous instruments; and it consists in ascertaining the resultant of the various pen dulum and metronome (inverted pendulum) effects of each molecule and compensating them so as to make the net resultant effect in all positions of the watch approximate very closely to zero. I accomplish this compensation by removing a portion of the metal of the balancespring in an accurately-located position. It is very easy to compensate these effects in the balance and those parts of its attachments which are stationary with reference to its pi vots by putting them in equilibrio, but it is very diflicult, in spite of the invaluable researches of M. Phillips and his discovery of the theoretical terminal .curves of balance springs, and in fact it has heretofore been impossible, except by a practically prohibitory expense of time and labor, to accomplish this result in the balance-spring on account of the constantly-Varying relation of each molecule to the axis. During the expansion and contraction of the oscillating spiral it is evident that each molecule vibrates around a central point which is itself constantly shifting with reference to the axis of the balance and that it alternately increases and diminishes its distance from that axis. Moreover, if the balance is normally describing an arc of two hundred and seventy degrees on each side of its position of rest, or a total of five hundred and forty degrees of oscillation, the various molecules of the balance-spring describe arcs varying from live hundred and forty degrees to almost nothing, depending upon whether they are situated near the point of attachment to the pivot or near the point of attachment to the cock.

I-Ieretofore position errors have been corrected by carefully bending the balancespring and by other means requiring long and tedious trials. Theorists have generally said that to correct these errors the center of gravity of the balancespring must be brought within the axis of the balance. This would be true if the spiral were a solid of uniform density and section each molecule of which maintained its position with reference to the pivots during the period of oscillation; but it is not true, as appears evident from the above considerations and as will be proven later, when I show that by my method the position errors may be corrected in any case by removing a portion of the metal of the balancespring from either side of the center, for the center of gravity could be brought within the axis by removing the metal from one side of the axis only, to wit, the heaviest side, whereas one of the methods I describe must inevitably still fu rther increase the distance of the center of gravity from the axis. Thence it follows that the center of gravity per so has nothing whatever to do with the vertical position errors of the balance-spring. It is a question of balancing not weights but forces.

It is true that if the center of gravity were found to be within the axis of the spring in all positions of its oscillation, as with a mathematically-perfect Phillips curve, then it would follow that all the active metronome and pendulum effects of all the molecules would be exactly balancedtheir resultant would be Zero.

It is physically impossible to constitute a spiral or helical spring with a mathematicallyaccurate Phillips curve, but it is practically possible by my method to balance these effects without having the center of gravity accurately within the axis, and my method is well adapted to reduce easily and promptly even the small resultant errors of well-constructed Phillips springs say from seven to ten seconds a day to one or two seconds.

It is evident that all the metronome effects are opposed to all the pendulum effects in any one position, and that the resultant effeet, which I call the resultant pendulum effect or the resultant metronome effect, (depending on the position of the watch,) alone has any influence in advancing or retarding the rate of the watch in vertical positions. However accurately a balance-spring may be conditioned and mounted in the first instance, one or the other of these effects always preponderates. Nevertheless, the watch can be regulated, if otherwise in good order, so as to keep accurate time in any one vertical positionas, for instance, pendant upbut in the other vertical positions-to wit: pendant left, pendant right, and pendant downthe watch will gain or lose. If in one of those other positions the resultant be a pendulum effect, the watch will run faster. If, on the contrary, the resultant isa metronome effect, the watch will run slower.

My method consists in ascertaining the exact daily rate of the watch in the four vertical positions, and thus locating the maximum resultant of the pendulum or metronome effeet and then lightening the balance-spring in that location so as to bring the resultant of gravity to or toward zero.

If the watch were adjusted to mean time or +0.0 seconds pendant up, and should show +5.0 seconds pendant right, 11.0 seconds pendant left, 6. 0 seconds pendant down, it is evident that the maximum resultant metronome effect (causing the watch to run slowest) would be found at the top of the spring when the watch was in a position between pendant left and pendant down, and the outer coils should be treated by removing some of the metal in the part which is uppermost when the watch is in that position. It also shows that the maximum resultan t pendulu m effect (causing the watch to run fastest) is obtained between pendant up and pendant right, but nearer pendant right, 2'. 6., that part of the spring which was uppermost when the watch runs slowest is lowest when the watch runs fastest. The spring might also be treated on the inner coils on the opposite side. In a well-conditioned balance-spring the amount of weight to be removed is a few tenths of a milligram, or a few thousandths of a grain.

In view of having the invention the better understood I will now proceed to describe it with reference to the accompanying drawings, in which a balance-spring treated according to my method is shown in three different positions of the balance-axis.

Figure 1 shows the balance-spring in the position of rest or zero. Fig. 2 shows the balance-spring when the balance-axis has been rotated three-quarters of a revolution from right to left from its position of rest shown in Fig. 1. Fig. 3 shows the same balance-spring after the balance-axis has been rotated three-quarters of a revolution from left to right from its position of rest shown in Fig. 1.

111 all the figures the same letters refer to the same parts.

A is the axis of the balance-wheel, and B is the point to which the balance-spring is fixed by any suitable means whatever.

Supposing now the line A X to be the direction of the vertical line when the watch is placed pendant up and A Y to be the direction of-the vertical line when the watch is placed pendant right, and supposing that between those two vertical positions of the watch a daily difference of, say, four seconds has been observed, the watch running, for instance, ten seconds slow with reference to mean time in the position pendant up and six seconds slowin the position pendant right, this will show that the resultants of the metronome and pendulum effects produce an error of four secondsthat is to say, that part of the balance-spring is too heavy and acts either like a pendulum in the one sense or like an inverted pendulum or metronome in the other sense. The rate being, for instance, too slow in the direction A X and faster in the direction A Y in the above-cited example the net results of all the forces produce a metronome effect in that part of the spiral spring situated between the line A X and the line A Y, but somewhat nearer to A X than to A Y. New I may correct this either by suitably reducing some of the outer coils of that portion of the spiral spring, such as w 5U y .2, or by suitably reducing some of the inner coils of the spiral spring, such as w to 3 z, in the opposite portion of the spi ral spring, or by reducing both. I prefer however, to reduce the outer coils, for instance, as shown in a my .2, because those parts of the spring will never reach in their vibration an arc of three hundred and sixty degrees. This reduction may be accomplished, for instance, by means of scraping slightly the edges of those coils by any suitable means whatever. The resultant metronome or pendulum effect of the balancespring will therefore be reduced, and if the spring has been treated at the right spot and to the right amount the observed differences in the rates of the timepiece in the four vertical positions will be so considerably reduced that they will be considered as practically suppressed.

Figs. 2 and 3 show the respective positions which are assumed by the treated portions or a; 1 of the spiral spring when the balance wheel axis A makes three-quarters of a revolution in either sense from its position of rest shown in Fig. 1.

It will be observed that the treated portions of the spring never describe a complete revolution, and this allows us to exactly compensate the spiral spring even for such amplitudes of oscillation as those in which the balance-wheel makes more than one complete revolution.

I am aware that inventors who have been imbued with the idea that the center of gravity must be brought within the axis have naturally endeavored to accomplish this result by attaching suitable weights to spirals, and it is.

true that theoretically such weights might be added according to my method of calculation in such a way as to effect a certain degree of compensation, but the reasons for removing weight from the spring rather than adding weight on the opposite side are obvious. The adding of weight whether by ceinentry, soldering, or electrodeposition cannot be done without seriously affecting the freedom of the springs play, to say nothing of its appearance, nor can it be done so permanently, whereas the weight can be reduced on the un der side, if preferred, so as to make the treatment invisible, innocuous, and permanent.

I find by actual experiment that the quantity required to be removed from the said coils of the balance-spring in treating it according to the above-described method is so slight that it does not visibly affect in any way the correct action of the spring as a whole, nor disturb any of its functions.

Having thus fully described my invention, I claim 1. A balance-spring for a Watch or analogous instrument a portion of whose metal has been removed thereby bringing the resultants of its pendulum and metronome effects to or toward zero.

2. A balance-spring for a watch or analogous instrument a portion of Whose metal has been removed from the exterior spires or coils of the spiral as and for the purpose specified.

A balance-spring for a watch or analogous instrument, aportion of whose metal has been removed from the interior spires or coils of the spiral as and for the purpose specified.

4. A watch or analogousinstrument the balance-spring of which has a portion of its metal removed for the purpose of bringing the resultants of its pendulum and metronome effects to or toward zero, substantially as and for the purpose specified.

5. A watch or analogous instrument the balance-spring of which has had a portion of the outer coils of the spiral removed as and for the purpose specified.

6. A watch or analogous instrument the balance-sprin g of which has had a portion of the inner coils of the spiral removed as and for the purpose specified.

7. A balance-spring for a watch or analogous instrument, a portion of whose metal has been removed from the exterior spires, and another portion from the interior spires substantially as and for the purpose specified.

8. A watch or analogous instrument the balance-spring of which has had a portion of the metal of the exterior spires removed and another portion of the metal of the interior spires removed, substantially as and for the purpose specified.

9. The method, substantially as hereinbefore set forth, of locating the maximum resultant metronome or pendulum effect in an oscillating vertical spiral spring, which consists first in timing the watch in the four vertical positions and then from the results of these timings locating the position where the weight of the spring should be varied.

In testimony whereof I have signed my name to this specification in the presence of two subscribing witnesses.

- WILLIAM WOODNUTT GRISCOM.

IVitnesses:

ALBERT I-I. POTTER, E. IMER SCHNEIDER.

Images