Home › P › Poising the Balance

Poising the Balance

In merely poising a balance for a cheap movement there is no great difficulty, that is, putting it in equipoise sufficient for the reasonably good performance of the movement; but to well and thoroughly poise for a high grade of movement embraces means and methods not necessary in the first mentioned. In a cheap balance a high degree of accuracy is not expected, and so the manipulations are, in the poising, simple, provided all the parts are in condition to admit of poising. The following will be about all the conditions and means used generally: In the outset the balance should be in poise without its staff, and this is approximated before the staff is in by putting into the staff socket in the arm a piece of true wire, sufficiently tight to allow of the balance being held onto it with friction, so that the balance can be trued in the flat by the fingers or with tweezers and remain while poising on the parallel bars.

Fig. 13 illustrates a form of tweezers made especially for balance truing. To here explain the parallel bars and give a few points regarding the essential features will be well, and help to make clear some points that follow in the poising instructions. The parallel bars for the use of watch repairers with the following features, will be suited to all the cases met with: The two bars, if made of steel, for instance, must have only the top edges on which the pivots rest made of this metal, and the less the better. The top edge should not be over 1/100 of an inch thick and the bar 3/4 or I inch long. The bars must have the guides that carry them move them open or shut for different lengths of staffs, and keep the bars parallel during the movements. The bars, after they are in their places and securely fastened to the stand carrying them, must be ground true, straight and parallel, on a flat piece of glass (plate glass is the best), charged with emery of about 140, with oil sufficient to make a paste. The glass can be held and used as a file or the bars can be held down on the glass and moved about with a circular stroke, but if the stand is large and heavy this operation will not be readily performed with good results. The main reason for using the glass referred to, is that it is a ready way of getting a grinding bed comparatively true without labor or preparation. A flat metal surface, marble or stoneware, would answer well, but would not be so readily had. After the emery has ground the surface true, clean off all the emery and use fine oil stone powder or pumice stone; clean, and follow the pumice stone with any polishing powder, or follow the pumice stone with a large and true burnishing file, keeping the surface wet slightly. In making the parallel edges, the object is to give them a perfectly straight surface on the edge and highly polished. These parallels are probably best made of bell metal, as there is then no danger of their being affected or accumulating magnetism. In the construction of a poising tool, to avoid the use of iron or steel in its make-up, will be found the most satisfactory, as then magnetism will not be a disturbing element that it might otherwise be. The whole tool should be heavy and low and stand on the bench firmly, and, if a fine one, have two level vials set in its base to level up the parallels with, before using. With a level bench and a tool made so that the feet are parallel to the top edge of the parallels, there will be little trouble in the balance rolling by gravity while poising. There are a great variety of poising tools, and any that have the parallel bars true and straight and parallel to one another, readily adjusted for distance, and have a firm and heavy stand, will be easily and satisfactorily handled.

Holes for the staff pivots are not good for poising, although jeweled, as the pivots must turn in them with a slipping action, whereas they roll without slip or friction on the parallels. The extreme top edge of the parallels, if of hard substance, can be made as thin as the 1/200 of an inch and be all the better, as will be explained. The plain, straight portion of a conical pivot of a fine staff is frequently not over the 1/100 of an inch long, and this is the part of the pivot that is to be exactly concentric with the center of gravity of the balance after poising is accomplished and is that part of the pivot that rests on the jewel. Now, from this it will be seen that the thickness of the parallels can not be great, not over the 1/200 of an inch, as the conical part of the pivot must not touch the parallel, and the end of the pivot should be outside of the parallel. Fig. 14 will show the situation and give the best idea. After the balance has been trued on the wire, then test on the straight edge, and if the balance rolls freely and gravitates, then lighten it on the down or heavy side. Or in the event that the balance is rather light it may be advisable to weight it on the top or light side. It will take a little practice to poise in this first operation, and there are several points to look at. First, if the balance is a heavy one, then in poising take away weight; second, if a plain (not comp.), remove little bits from the under side of the rim with a graver or drill; if very light, add weight by drilling in the rim and driving in several pins and then filing away till poised. The pins must be put into the rim at such points as are indicated by the circumstances. Soft solder, if used on the under side of a plain balance, is very easily handled, but the risk from the soldering fluid is great and requires great care in cleaning, but when all is well done, it serves a good purpose. As the wire on which the balance is hung is large in diameter, the poising will not be very delicate, but can be made good enough for the end served. In poising a compensating balance, the balance must be hung on a wire with each end pointed, turned to points, so that the wire can be held in the calipers and the balance made true in the round. (See Gauges.) Set the gauge of the calipers so that the rim at the end of one arm shall exactly coincide with it, and then turn the balance slowly under the gauge and see if the rim turns truly under it. If not true, bend in or out with the fingers and try by gauge till the balance will turn true in the round, then put onto the parallels and poise as in case of the plain balance, but alter the weight with the screws. The screws that are at the bottom can be put into a split chuck and a little turned away from the under side of the head, or a washer (See Balance Screw Washers.) can be put under the head of the top screw, and this method pursued till a reasonably fine poising is obtained. In these operations all the points relating should be well considered, and not make moves without method and good reasons. Care is required all through poising in all its branches.

These washers are very convenient to use in cases where a balance requires a little more weight, and where it is not advisable to change the hairspring or regulator when regulating to time, and in such cases must be put under the heads of the screws at the ends of the arms. All things being equal, in poising a weighted balance, it is better to add a little weight than to take away any, by turning the heads of screws as described, and then the balance is not in any way injured, and if it was all correct when found, although indications led to other conclusions, by removing a washer or two the balance would be left as originally, and much trouble saved in trying to remedy a mistake. Never make any changes in a fine compensated balance, as, in all probability, it was correct when made and some injudicious handling is to blame for any defect. After a balance has been trued in the calipers as described, so that the rim is truly concentric with the hole in the arm, it should, if it has not been injured, be virtually in poise, but if it is not, add washers to the screws on the light side, and by them try to poise it rather than by lessening the weight. Many times, taking a screw from the heavy side and putting it in place of one on the light, and the light in place of the heavy, will tend to an equilibrium, and so far as it does, is so much gain In removing the screws in a compensating balance, care must be used when they are replaced, to see that they are left just tight enough to stay in place, and at the same time not bind the head hard down on the rim. Screws badly handled in this respect may derange the compensation, also the poising. All the screws of a balance, except those at the ends of the arms, and occasionally a pair of the quarters, should be down, heads close to the rim. The others can be turned in and out at pleasure to poise, or for timing, as required. With a balance with a screw at each end of the arm, it is best not to move them in or out in poising, but proceed as described and leave these screws to be moved in timing afterward, if requires, as it helps to make that operation easy. When a balance has four screws they may be moved to do all the poising and afterward any pair opposite, or the whole, may be moved in timing and not disturb the poising. A compensating balance with four screws as described is much the easiest balance to handle, for by these screws the finer adjustments in poising and timing can be easily performed with greater certainty than by the old methods as described.

The balance staff is a very important element in poising and its pivots should be perfect, that is, perfect cylinders, and all that part that touches the hole jewel should be of equal diameter. By referring to cut of staff, it will be seen that the end after leaving the cone is straight, of equal diameter throughout its whole length, and this is the shape of all staff pivots at that point riding in the jewel holes, no matter what curve or shape may be given to the balance of the pivot. When there is a different diameter in the top and bottom pivots they are each true cylinders and their cylindrical diameters are parallel to each other and to the axis of the staff. When pivots are bent, or out of parallel with the axis of the staff, they are then not in condition to make poising possible, as a bent pivot will make a balance gravitate and act as though out of poise in itself, and with a bent pivot, poising can only be approximately attained. Perfectly cylindrical and parallel pivots to a staff are, in poising, a very essential feature, and without which poising cannot be attained.

When a balance has been poised as indicated and a staff made and fitted with perfectly cylindrical and parallel pivots, proceed as follows, and there will be little to do to complete the operation: First put the balance on the staff with a hollow punch and only press it on sufficiently to hold for preliminary tests; then place on the parallels of the poiser and examine; should the balance appear in poise it must not be taken for granted that it is so, but try a very slight jar given the poising tool, like rubbing over the frame an old file, which will impart to it a very slight vibration, and if the balance is actually out, it will roll and then remain with the heavy side down. If a jar, such as a series if taps with a hammer, be given, the balance will rotate and stop for an instant and then rotate again, and finally jar off the bars and the operation will not prove anything. The jar is such that the balance raises up bodily, when made with a file, and then falls down exactly on the same place on the parallels, rather the pivots come to rest always at the same point; and it will be seen by this means that if any point of the rim is in reality heavier from gravity, that it will by the momentum imparted, fall, overcoming the pivot friction, and finally seek a point in a direct line under the staff.

Repeated movements of a balance while on the parallels are necessary, together with great cleanliness of pivots and parallels, to thoroughly ascertain the true poised condition of the balance. When it is ascertained that a balance is out of poise or has a heavy side, punch out the staff and put the balance again on it only turned just one-half way around, and repeat as above. In this way a staff can be put into a balance to the best advantage and such little items all tend to save time and make easy the whole handling. When the best position is found for the staff, stake it, and true in the flat, and test again on the bars, and if necessary make further changes as above to affect a poise. When a balance is in poise and a staff perfectly true as has been described, and well staked on, it will in the most cases be found poised and nothing further to do. After putting on the roller it is advisable to test again for position, but it is generally unnecessary, as this will not disturb the poising only in exceptional cases. By staking on the roller too tight the staff may be bent and may destroy the poise.

Care is necessary in handling a balance for any purpose, not to bend the rim, soil or corrode the metal and finish, and in making slight alterations in the curve of the rim, not to bend it at the holes and so destroy its true circle and injure the strength of the metal and change its adjustment. Any one, after poising a balance and testing the movement carefully in different positions, will in many cases be aware of quite a change of rate in the changes of position, and this, at the first thought, would seem to rather reflect on the accuracy of the poising; but it will be found to occur at times with the most carefully poised balance, and that the operation of poising by the parallels does not comprehend the whole, nor the very nicer requirements. In any case the most careful mechanical poising must be attended to first, before any operations of a more delicate nature are attempted. In short, the parallels are to be used in the most delicate methods, but precede the others. When a movement is placed in its case and hung up, after poising on the parallels, its rate should be carefully noted for a given time, then it should be just reversed and set up with the pendent down, when it will be found, as a rule, that after a trial of same duration as the first, that the rate will not be the same. Now, when this occurs in a fine movement, it will be advisable to investigate all the parts which in any way relate to this action. Both hole jewels must be examined, for finish, thickness and truth of the bore; the roller jewel and the lever-fork examined; guard pin and its action with the table; the hairspring and all its relations and connections; the balance must be removed and then the lever, and the lever placed on the parallels by its staff pivots, as in the balance, and tested for poising. The lever should, when placed on the parallels, lay horizontal, like the beam of pan scales, and not swing or hang either end down; the weight should be removed from the heavy end, in such an event, until the lever will lie as indicated. Levers can be, and are made, that will stand in any position, like a poised balance, but it will, in most cases, be difficult to poise a lever for any position other than horizontal. Next, the escape wheel must be poised so that it will perform as a poised balance, when on the parallels; lever and escape jewels examined, as in those of the balance staff.

After all has been so far attended to, and the parts in place again, the balance must stand, when the mainspring is entirely run down, with its arms either perpendicular or horizontal; with a movement, whose balance is near the center, the arms can stand pointing to 6 and 12, or 3 and 9, as the most convenient. In requiring balance arms to stand in some fixed relation to prominent points of the movement, the manipulations are greatly facilitated, though any position the arms may chance to have will not interfere with the result, but a more expert hand will be required to get along with ease and certainty.

When all the foregoing operations are attended to, hang up the watch and take its rate for 12 hours, with main spring fully wound up; then reverse its position, with main spring wound up, and test for another 12 hours. On examination, if there should be any considerable variation in the rates in the two positions, say 10 to 15 seconds, then proceed by changing the screws as follows: in a case where the watch loses when hanging, it indicates that a screw of the balance nearest to 12 or 6, when the movement is entirely run down, must be moved a very little in or out. In this case, it is fair to suppose that the balance is too heavy on the side nearest 6--that this side gravitates, and, to an extent, acts like a pendulum. Assuming this to be the case, turn the lower screw in and the upper one out, where there are four timing screws, and, where not, washers may be added to the top screws, and the two trails repeated. After trial, if the result is improved, then the lower screw may be made a little lighter, but not at the first trial. In the first trials the balance should not be altered in weight, as indications in these manipulations are changed or modified by conditions not yet mentioned. We will assume that the balance has four screws, and when one is turned in and the other out, as indicated, and the end attained, then the watch is to be placed with the 3 or 9 up, and two trials made, as in the first, and the same method used, if indications are similar.

When the handling of the balance has been correctly done, the poising will be found to equalize the rates of the different positions, and the total performance improved. There are, of course, many chances for mistakes, but, with caution, they will do no harm, for if the balance is not changed other than a change in distribution of its weights, the act of restoring will be merely setting all the screws back to the position they were when poised by the parallels, and then proceed again on a new method, reversing the first; and then gradually it will be made clear to the most inexperienced, remembering that what held good in one case may not in another; and that various cases are only compassed by trial; and that the indications in the one may be just reversed in the other.

Instead of changing the lower screw as previously suggested, another trial may be made with 12 down, and the rate taken for the same period as for 6 down, and the two compared. Now if the watch maintains its former record it is pretty good evidence that the two rates will be its rates for these two positions, and then the alterations may be made. Now, while hanging in this case the watch lost, 6 down, and relatively gained with 12 up, and a very natural conclusion would be, if losing with 6 down, that the lower side of the balance would be the heaviest. Such is not the case, but the indications are that the upper side is the heaviest, and that the screw there should be turned in, and that the lower one may or may not be changed. Change the top screw first, in this case, and then make another trial and compare with the first. In all average cases, after changing the screw, the two rates should be found to be closer than in the first trial, and this will give a pretty good index of how to proceed. The philosophy of the action is the same as that of the action of the musical measuring instrument used to beat music measures, called a metronome. It has a short pendulum with the rod prolonged above the shaft that it swings on, and on the upper end of this rod is a small weight that slides up and down and so regulates the beats. The position of this weight, being above the center of motion, has a very great control of the vibrations and controls them for a wide range. For instance, the whole pendulum of one of these instruments is not over 2 or 2 1/2 inches long, but with the little counter weight it can be made to beat seconds and slower measures, which could not be accomplished with anything short of a 39 inch pendulum and over. Then move the screw as already indicated, keeping in mind the compared pendulum action and its philosophy.

Gravitating on the principle of the simple pendulum is not the whole problem in moving the screws of the balance, but they embrace the philosophy of the instrument described, and this must be kept in mind in the handling. In experiments it will be found that a screw moved at at the top of a balance, will make twice as much changing in the rate as the same movement of a screw at the bottom. Hang up a watch and turn out the lower screw one-half a turn, and the rate will be, for instance, ten seconds slow in six hours. Now put up just reversed, and for the next six hours the watch will be found twenty seconds slow or more. Now, if we proceed in this case on the simple pendulum philosophy, we should make a mistake in moving the screws. In practice it is not necessary to make only tests for 3 and 9, assuming we have an open face watch. First regulate on full spring for 6 or eight hours hanging, and when well regulated place the watch 3 up and then 9 for the same period on full spring, and if any material change in rate is found in the two last, then move the screw as already indicated, keeping in mind the compound action and its philosophy.

The handling of the screws in poising on the parallels and in the running watch are for some indications just reversed, and this is due to the action of lever and hairspring on the balance, with gravity in one case and to gravity alone in the other. In experimenting with the running watch always wind fully up for each trial, and periods of six to eight hours will be found the most convenient. The upper coils of a mainspring are much the most equal in power, and consequently give best results; that is, the fourth, fifth and sixth turns of a spring are much nearer each other in strength than are the second, third and fourth. If a balance is perfectly poised mechanically, and the whole train in perfect mechanical poise and condition, then the running watch should not give any very considerable difference in rate in four positions, but as this is not the case generally there will be a change of rate in the positions and the balance can be then manipulated to correct the error, although it in itself may not be at fault. The reason for not testing a watch for the whole range of four positions, is that in the pocket, a watch is not supposed to get into a position with the stem down, three and nine are apt to be up and down, and so with twelve are the three positions used. The isochronal condition of the hairspring is apt to make trouble in these experiments, and this is another reason for using full spring invariably. The extent of motion of the balance is another element in the matter, and any movement when in perfect poise for a balance motion of 3/4 of a revolution, each side of the center or dead point (1 1/2 revolution) would not be found in as accurate poise for 3/4 of a revolution. A balance making one and a half revolutions, to a certain extent, is self-correcting, as will be seen, and is to be preferred to any other movement, for if any point of the rim is out of poise then the fault is brought just opposite in each excursion, an so does not relatively gravitate. Owing to the fusee, an English lever with a balance making one and one-half revolutions, is the highest form of movement for accurate adjustments of any kind and so is the easiest to realize perfect poising. The American watch is so uniformly well and evenly made by machinery that poising is in it quite easy, and much more so than in foreign makes. A Waltham movement that I tested, just as it left the factory, only changed its rate about three seconds for the four positions. This could not be realized in any medium grade of foreign watch, and I presume this is not a single case, but probably rather a type. The American movement is made mechanically so near perfection that the watchmaker will find poising a balance comparatively easy, and that what he finds to hold good in one case will be pretty sure in another, due to this mechanical perfection. J. L. F.