The MEELECTRIC

JubileeCLOCK

An entirely new design for a one-second pendulum impulse clock,specially suited to amateur construction

By Edgar T. Westbury

ALTHOUGH it is often con-sidered that clockmaking isnot within the legitimate

scope of model engineering, it hasalways been very popular amongME readers.

There are very good reasons forthis ; clocks are among the mostinteresting of mechanisms, providescope for good craftsmanship both inhand and machine tool operation,and, when completed, can be appre-ciated by everyone for their beautyand utility.

Electric clocks, in particular, attractthe amateur constructor, not only onaccount of their novelty, but alsobecause they depend less on specialisedhorological skill for their successthan the conventional type.

Many designs for electric clockshave been published in ME during its60 years of existence-it may berecalled that the first constructionalarticle on the now famous Synchro-nome system was published in thismagazine-and material proof thatfull advantage of these designs hasbeen taken by readers can be foundnot only at most model exhibitionsbut also in innumerable homes.

In presenting yet another designfor an electric clock, my aim is not tosupersede all previous designs or toclaim any very special advantagesover them, but to introduce someoriginal features and interestingmethods of construction, which, Itrust, will appeal to an even widercircle of readers.

Although I have never submittedan electric clock design for publicationbefore, I have always possessed a verykeen interest in the subject and havemade many experiments with an aimto improving or simplifying design.In particular, I have studied theproblems of design from the aspect

DIAMOND. JUBILEE

of the non-specialist constructor, whooften encounters small but irritatingobstacles to success in this branch ofwork.

Among several essays on construc-tion, I may mention that many yearsago I built a clock with only threewheels in it, and this idea has sincebeen exploited in at least two clockswhich have been exhibited at MEExhibitions.

TYPES OF ELECTRIC CLOCKSBefore describing this particular

clock, it may be helpful to review thevarious types of electric clocks thathave been designed in the past, toexplain what literally “makes themtick.”

The term “ impulse clock ” is usedto define any type of clock driven byelectro-magnetic impulses applied tothe pendulum. which. in effect.becomes a precision-governed motor;capable of driving the dial motion(or other time indicator) at a constantrate. In this way the mechanism maybe said to be inverted in comparisonwith weight or spring-driven clocks,in which the pendulum is driven bymotive power applied through theindicating gear train.

From the mechanical point of view,therefore, the electric clock offerssome advantages as it avoids theheavv loading on the gearing. affordsa very robust pendulum action, andenables interference with free pendu-lum action to be reduced in compar-ison with most kinds of escapements;the latter property, however, is notalways exploited to its fullest ad-vantage.

Some kinds of impulse clocks havea balance wheel instead of a pendu-lum, but the same general conditionsapply. Except in cases where it isnecessary for the clock to be readilyportable, however, the pendulum is

569

The completed clock, fitted in a handsome longcase with glassfront and sides and a batterycompartment fitted in the base

generally preferred, and only one ortwo types of balance-wheel electricclocks have achieved any degree ofpopularity, though they are interestingto construct. A notable example wasthe Eureka clock which was fullydescribed in ME some years ago.

The type of electrically-driven clockin which a normal gear train andescapement is employed in conjunctionwith a short spring which is rewoundat fairly frequent intervals by an

MODEL ENGINEER

-

electro-magnet,. might loosely be de-scribed as an impulse clock. But amore accurately descriptive term is“ electrically wound,” or “ remon-toire,” to borrow from the French,who are experts in coining technicalterms.

This system is common in motor-car clocks, and has also been used inelectric supply meters; it offersattractive possibilities for the adapta-tion of existing clocks-from cheapalarms and grandfathers, to auto-matic electrical maintenance.

Prejudice against electricityThe modern a.c. Synchronous clocks

are not in the true sense devices formeasuring time, but only remote-controlled repeaters, following themotion of a master clock which isused to govern the frequency of themains supply. Under normal cir-cumstances, they give accurate andreliable timekeeping. But their totaldependence on external factors maybe a serious disadvantage as some ofus have discovered to our cost in thebad old days of overloaded mainsand power cuts.

For various reasons, traditional andotherwise, practical horologists ofthe orthodox school are not in lovewith electric clocks of any kind. Nodoubt many of them adopt as anarticle of faith the dogmatic pro-nouncement of Lord Grimthorpe, thedesigner of the Westminster clock,who in his book Clocks, Watches andBells, stated : “ These clocks neveranswered in any practical sense; norwould anything but the strongestevidence, independent of the inventor,convince me that any independentpendulum directly maintained byelectricity can succeed in keepinggood time for any considerableperiod.”

In my opinion, however, thisevidence is now forthcoming in thelarge number of electric clocks whichhave been, and still are, really accurateand reliable timekeepers.

A more rational reason for thedislike of electric clocks was oncegiven to me by a working clock-maker, who said that he would nothave a magnet of any kind anywherenear his work bench, where toolsmight become magnetised and delicatebalances deranged. Generally, how-ever, it may be said that humannature dislikes, and often fears, any-thing which is unusual and, even inthe remotest sense, mysterious.

HISTORY OF ELECTRIC CLOCKSThe first inventor who applied

electrical power to drive a clock, sofar as is definitelv known. wasAlexander Bain (1846) who used twosolenoid coils in conjunction with apermanent magnet on the pendulum

rod, to drive the latter in alternatedirections by current controlled by asliding contact switch. A furtherinteresting feature of one of Bain’sclocks was the use of an “ earthbattery,” consisting of carbon andzinc electrodes sunk in moist earth, toprovide the current for driving it.

Although his clocks undoubtedlyworked, the contact mechanism gavea good deal of trouble. and theyviolated basic principles of good clockdesign by undue interference with thenatural swing of the pendulum.

Sir Charles Wheatstone, the pioneerof the electric telegraph, with whomBain was at first associated but latercame into violent altercation, alsoapplied the principles of electro-magnetism to drive clocks and thetransmission of time impulses toremote-controlled dials. But hereagain incomplete understanding ofbasic requirements in clock designlimited his success.

Similar limitations were apparentin the clocks designed by C. Shepherd(who exhibited a clock at the 1851Crystal Palace Exhibition), Sir DavidGill (1879), and Professor Froment.

Major problemsThere were two outstanding

obstacles to complete success of elec-tric clocks, which were never success-fully surmounted in these earlyattempts, namely a that of obtaininga reliable and efficient electric contactwithout interfering with the freeswing of the pendulum, and b regu-lating the electro-motive force of theimpulse to maintain a constant arcof pendulum swing. which is essentialto really accurate isochronism-inother words, good timekeeping.

Inventors have tackled these prob-lems in many ways, and with equallyvaried success; the inventions ofM. Fery (1908), Frank Holden (1909)and M. Fabre-Bulle (the designer ofthe once-popular Bulle clock) deservemention in this respect. And out ofthese experiments, two distinct butequally successful principles in utilisingelectro-magnetic force to drive aclock have emerged. These are: athe use of an impulse directly appliedto the pendulum rod, uncontrolledin strength but governed in frequency,and b indirect action by an impulseof constant frequency which is usedto lift a weight. the fall of whichnrovides a constant imvulse to drivethe pendulum.

Method a is employed in the Hipptype of clock, which is by far themost popular among amateur con-structors while b is used in the syn-chronome system and also in severalothers which might be described asmodifications and variants of it, ex-tensively employed for multiple controlof secondary clocks.

The Scott intermittent contactgoverning mechanism

570 DIAMOND JUBILEE

A close-up of the dial andmotion work

The term “ escapement ” is generallyused, for purposes of convenience, todescribe the contact-making mecha-nism of impulse clocks, though itscorrectness in this application maybe open to question. Thus in theHipp clock the so-called “ butterflyescapement ” is referred to, while thesynchronome system is said to employa “gravity escapement.”

I do not propose to describe these

.

devices in detail as they have oftenbeen dealt with in ME articles, andanyone who wishes to pursue thesubject can obtain books on electrictimekeeping to which I shall refer later

Some comments on the essentialdifferences between the two systemsmay, however, be helpful to con-structors as they explain some of thereasons for the adoption of essentialfeatures of the present design.

The electro-magnet of the Hippclock is so arranged as to give a briefimpulse or “ kick ” to the pendulumat intervals, which are timed by an

MODEL ENGINEER

DIAL CIRCLE I I Li !llU

SUSPENSION

/COUNT WHEEL

160 T RATCHET _ y-1

/ ’

j CANNON/ 20‘\

MOTION PINION‘,, 16~

“/. ,\(‘MOTlON ,‘--_tMOTION WHEELINAL

‘\

ingenious trigger mechanism in thecontact-making device.

Under normal working conditionswhile the pendulum is swingingover a determined arc-no contact ismade as the trigger swings completelyover the contact block; but as themomentum of the pendulum declines,it eventually reaches a point where thetrigger fails to clear the block so thaton the return swing it drops into anotch in the latter and depresses thecontact spring, closing the contactsmomentarily and thus energising theelectro-magnet to produce an impulsewhich restores the arc of swing.

DIAMOND JUBILEE

/ELECTRO-

, MAGNET

IMPULSE FREQUENCYThe frequency of the impulse may

vary from a few seconds to well overa minute, depending on the efficiencyof the magnet, the voltage or workingcondition of the supply battery-allof which in practice are variablefactors-but the ultimate result is tokeep the mean arc of swing sub-stantially constant whatever the varia-tion of power.

The Synchronome system, on theother hand, does not apply the electro-magnetic impulse to the pendulumdirect, but employs it to the lifting ofa weight which is then dropped at

571

regular intervals (i.e. half a minute)on an impulse pallet attached to thependulum rod, thus imparting aconstant driving force, which also hasthe same effect of maintaining aconstant mean arc of swing.

Both systems, therefore, are immunefrom the interference caused in othertypes of clocks, where electromagneticforce of varying magnitude is appliedto the pendulum at regular intervals(sometimes at each complete or halfcycle) thereby affecting the arc ofswing.

Both the Hipp and Synchronomesystems employ a gathering pallet andratchet wheel in the count mechanism,but whereas the former operates thedial motion mechanically, the latterusually transmits power to one ormore dials electrically, through thecontacts which control the weight-lifting magnet.

Both systems are simple and robust,electrically and mechanically, but ofthe two the Hipp is rather better suitedfor use as a domestic clock, beingquieter in action and readily adaptableto the conversion of an ordinaryweight or spring-driven clock.

The Synchronome is, as it wasoriginally intended, essentially a timetransmitter, suited to the remotecontrol of secondary dials. Althoughit can be used as a single self-containedunit its use for domestic installation isnot popular; the noise of the fallingweight at half-minute intervals isoften found objectionable, and at-tempts to quieten it have not beencompletely successful.

NEW FEATURES OF DESIGNIn the present clock I have incor-

porated some of the features-I hopethe best ones-of both the Hipp andthe Synchronome clocks. Thoughintended primarily as a domesticclock, it can be used to controlsecondary dials if desired.

The governed frequency of impulseis adopted, but the method of apply-ing it is modified, with the objects ofstill further reducing noise (somepeople object to the “ wipe-scrape ”of the Hipp escapement, especially ifit occasionally catches a crab !) andalso of avoiding point-to-point con-tacts.

The latter are often troublesomebecause if they are closed with suffi-cient pressure to be fully effective,they impose load which interfereswith the free swing of the pendulum;while if made very light in action theymay be unreliable, especially if pittedor tarnished.

A further feature which should beattractive to the constructor, is thatthe clock is made up of a number ofself-contained and detachable unitsfor carrying out the various functions,each of which can be removed for

MODEL ENGINEER

overhaul or adjustment. This increasesthe number of individual parts andwould probably not be at all favour-ably regarded in a commercially-produced product, where the prevail-ing tendency is to reduce parts to aminimum-though some of them mayhave to be of complicated form to doso.

However, this unit constructionmethod actually enables the individualparts to be simplified, and the finishedresult looks more complicated thanit is, which may be an advantagefrom some points of view !

The units consist of a the pendulumand suspension unit, incorporating ametal “ chassis ” on which the otherunits are mounted,, with a bracketat the top from which the pendulumis suspended; b the electro-magnetunit; c the contact control unit; andd the clock train and dial unit. In allthese there are optional features ofdesign, to which I shall refer in duecourse.

WORKING PRINCIPLEThe main item of novelty in the

mechanism is the method of control-ling the frequency of the electricalimpulse, which, in common with thatof the Hipp clock, is only appliedwhen the pendulum swing falls belowa predetermined arc. But unlike theHipp, it is practically silent in action,and employs a very lightly loadedwipe contact which is largely self-cleaning and imposes very littleimpediment to the free swing of thependulum.

So far as the basic principle of thisdevice is concerned, I cannot claimcomplete originality as it was firstemployed several years ago by HerbertScott, of Bradford.

It entails the use of a ratchet wheelC of special form, one tooth of whichis gathered at each swing of thependulum A by a paw1 or detentcarried by the rod. A contact-making device D is mounted above thewheel, and in fairly close proximityto it.

Under normal conditions of pendu-lum swing, the paw1 is carried wellclear of the ratchet tooth so that itfalls to the root and does not operatethe contact; but if the arc of swingis reduced below a predeterminedminimum, the paw1 fails to clear thetooth completely and drops into thestep near the tip so that it rides at ahigher level on the return swing,closing the contacts and energisingthe electro-magnet, thus giving animpulse to the pendulum.

In my modification of this devicethe ratchet wheel, which I havetermed the riding wheel, to distinguishit from the count wheel employed inthe clock train, does not necessarilyplay any part in the timekeeping

MODEL ENGINEER

function, its purpose being simply tocontrol the contact.

It may thus be made with anynumber of teeth, unless it is intendedto control secondary dials. But tocope with this eventuality I havedesigned it with 15 teeth so that itrotates once every half-minute so thatit can be equipped with a subsidiarycontact to energise dials on theSynchronome system if desired.

APPLICATION OF IMPULSEThe electro-magnet is arranged so

as to impulse the pendulum rod in thedirection of its travel (near the centreof its swing) when it is moving atmaximum speed. Many impulsemagnet mechanisms are inefficient inthat they apply the power at thewrong instant or in the wrong direc-tion. Hipp clocks and others oftenhave the magnet mounted verticallyunder the pendulum rod, to the end ofwhich an armature is attached, so thatit swings as close to the poles of themagnet as possible.

This generally works quite well, butit does not always apply the power tothe best advantage, and, moreover, itincreases the vertical loading on thesuspension spring, often causing asudden jolt when contact is made.Its worst fault, however, is that ifcontact is not broken at exactly theright instant-as often happens ifadjustment is not quite correct-itmay waste power or even tend toput a brake on the pendulum move-ment.

In some cases a solenoid is usedto attract the armature and this ismore efficient, also less critical inadjustment; a typical example is theBulle clock.

The electro-magnet in the clockillustrated has a pivoted armature,swinging vertically, immediately be-hind the pendulum rod and on thecentre line. At the lower end this isfitted with a roller which rests againstthe rod, and, when the magnet isenergised to attract the armature,pushes it gently but firmly to acceler-ate the swing.

The armature movement is, how-ever, limited when it contacts themagnet poles at the centre of theswing so that it does not prolong theeffect of the impulse while the speedof the pendulum is decreasing. Ifcontact is delayed beyond this pointno retarding effect is produced, thoughelectrical energy would be wasted.

One of the difficulties often ex-perienced by readers who wish toembark on horological work is thatof obtaining the necessary specialmaterials, either finished or semi-finished, which are involved in thisform of construction. Many en-quiries for advice in this matter werereceived following the articles on theME Musical Clock, and considerableattention has, therefore, been devotedto ensuring that the essential suppliesfor construction are available in thepresent case.

The actual clock train could bedriven from the riding wheel bysuitable gearing, but it has been con-sidered more appropriate to use anentirely separate unit, which is oper-ated from the pendulum by means ofa crutch, comparable to that employedin a weight- or spring-driven clock,but acting in the reverse sense. Instead

These will include such items aswheel blanks or flnished gear wheels,pinion wire, pendulum suspensionsprings, engraved or fretted dials andhands, etc., which readers often finddifficulty in making for themselves.At present arrangements for suppliesare not finally completed but anannouncement will be made as soonas they are ready.

l To be continued

572 DIAMOND JUBILEE

of the form of ratchet and paw1commonly used in electric impulseclocks, in which one tooth is gatheredin one direction of pendulum swingonly, the rocking lever operates twopawls in alternate directions of swing. :,

One reason for the adoption of thisform of motion is to enable a secondshand and dial to be fitted if desired;

1

with the usual single-acting ratchet inconjunction with a seconds pendulum,the hand would move two seconds Idistance at each complete doubleswing, and thus would not count

j

odd seconds.In the form shown here the clock

is designed as a timepiece only, butthe possibility of adding striking andchiming gear has been taken intoconsideration, and it is hoped todevelop this at a later date. The clockmotion plates have, therefore, beenmade large enough to allow plenty cfroom for fitting this extra mechanismif desired.

It is possible to fit the completeassembly in either a hanging orstanding case, but the latter is con-sidered most suitable for domesticpurposes, and as pictured the caseis of tasteful design, quite in harmonywith up-to-date ideas in housefurnishing, without being too blatantly“ contemporary.”

A full description of its constructionwill be given, following that of theclock mechanism.

The experimental work in thedevelopment of the clock may bedescribed as a team effort, and I amhappy to acknowledge the valuableassistance rendered by Mr C. B.Reeve, in wheel cutting and producinga very handsome engraved dial, MrJ. Cooper, in experiments with theelectro-magnet, and Mr A. L. Headechin making the case to my roughspecification.

The ME Jubilee ELECTRIC CLOCK

Even those who are not intending to build this timepiece will be interested in the construction of the pendulum, a feature which is vitall to accurate timekeeping. By Edgar T. Westbury

Continued from I May 1958, pages 569 too 572

N the experimental development of this clock, several alternative I methods in the construction of

the various details have been tried out, and the features of design which will be described are those which have been decided upon as most satisfactory from the all- round aspects of constructional simplicity, or the facilitating of machining, fitting and adjustment. There are, however, several in- stances where optional methods or details could be incorporated or minor improvements introduced.

“ Capable of improvement ” is generally regarded as a euphemistic implication that things are not as good as they ought to be, but in its literal sense it surely applies to everything in this imperfect world. Whether I suggest possible modifica- tions or not, many constructors will undoubtedly exercise their prerogative to depart from the blue print; which is as it should be, for this is one of the essential differences between model engineering and mass production. In calling attention to optional features, I can advise readers as to what alternatives are permissible or desir- able, and thereby help them to avoid fallacies or pitfalls in any experi- ments they may venture to make.

NEED OF CHASSIS The object of mounting all parts of

the clock mechanismm on a rigid metal backplate, or chassis,” to use a popular modem term, is to ensure stability in the relative location of the various parts. Many electric clocks are assembled by simply attaching the parts to a wooden backboard. Subject to normal precautions, this is generally satisfactory; but even under the best conditions, timber is sometimes liable to warp or distort under stress, and examples of failure to maintain exact adjustments have occurred from this cause. Most established designs of electric clocks, including the Syn-

15 MAY 1958

chronome, incorporate some form of metal chassis.

Our chassis frame consists of a light alloy casting, which has a three- point attachment to the backboard of the clock case, so that it is not dis- torted in the event of any possible warping of the latter. It has three flat facings, one of which carries the pendulum suspension bracket, while the other two serve as mountings for the electro-magnet and contact control units respectively.

lower facings, may be left till later, as they can be accurately related to holes in the mating parts by jiggingg or spotting; in fact, it has not been considered necessary with the latter to indicate the holes on the detail drawing. Apart from cleaning up any rough edges or surfaces on the casting, no further work on this casting is called for.

BRACKET FROM CASTING Our pendulum suspension bracket is

also made from a light alloy casting, though it could be fabricated if desired. At one stage in development, a built-up brass bracket was actually used, but the work involved in making it, and also the cost of material, was considerably greater than with a casting.

Although the essential work on this casting can, as in the previous instance, be done satisfactorily by filing, it is quite a simple matter to machine the rear bolting face, and also the top surface of the arms, by clamping itt to an angle-plate mounted on the lathe faceplate. The V-notch was milled, in my case, by mounting the casting on the vertical slide, with the top

Facing the end of the pendulum bob with the aid of an improvised wooden steady

No machining of the chassis casting is essentially required, though it may have a skim taken over the rear bosses and front faces if suitable machining facilities, such as a planing machine or a large-capacity lathe, are available. It is, however, quite in order to true these to a reasonable degree of flatness by filing; a non- clogging type of file such as the Dreadnought, Surform or Tresa file will be found most efficient for this purpose. Exact adherence to the thickness shown in the drawings is not highly important.

Three holes are drilled for fhe fixing screws. Other holes, including the tapped holes in the bosses just above the lower facings, and those for the attachment of the units to the

619

edge parallel to the vertical edge of the slide, the base of which was swung round at 45 deg. to the lathe axis. It was then possible to machine the notch with an ordinary side and face cutter, but this arrangement would not have been necessary hadd a 90 deg. angle cutter been available.

A 1/4in. clearance hole is drilled in the centre of the bolting face, and thus coincides with the position of the top fixing hole in the chassis when assembled, but to ensure permanent location, two 4 BA screws are fitted to hold the bracket in place; the exact position of these is not im- portant.

The detail drawing of the suspension bracket also includes the clamps employed to secure the crosshead or

MODEL ENGINEER

CHASSIS

,- -.I ’---k-’ 1’‘_

&LEAR.

suspension bar after the bar has beenallowed to find its own position in theV-notch. These clamps are simplybars of 1/8 in. x 3/8 in. brass, each fittedwith a single 5 BA stud, which passesthrough a vertical hole in the arm ofthe bracket and is pulled down by aknurled nut. To keep the bar hori-zontal a short distance stud isscrewed or riveted into its rear end,having a round or pointed tip to seatin an indentation drilled in the top face

MODEL ENGINEER

1 OFF L.A.

of the arm. To avoid any liability tojam, the stud hole should have ampleclearance and at the lower end itshould be slightly countersunk, theface of the knurled nut being bevelledto correspond, as shown.

With this arrangement the entirependulum assembly may be removedor replaced in a few seconds, byslackening off the two nuts andswinging the clamps round to releasethe suspension bar.

620

PENDULUMIt is well known that one of the

most important factors in accuratetimekeeping is the maintenance ofan exactly constant distance betweenthe suspension point of the pendulumand its effective centre of gravity.Although both the pendulum boband its rod are often made of commonmetals, the thermal expansion of thesemetals will introduce timekeepingerrors if there is any variation oftemperature. To compensate for these,many ingenious devices have beenemployed, including the bimetal or“ gridiron ” pendulum, b u t t h e ’commonest method in modempractice is to make the pendulum rodof an alloy which has practically anegligible coefficient of expansion,known as Invar. This material isrelatively expensive, but is worthgetting when extreme exactness intime recording is necessary; often,however, it does not produce thedesired effect because other errors inthe design or execution of the completeclock may completely nullify anyadvantage which it confers in tempera-ture compensation.

Unless high precision is a definiteend in itself, it is rather doubtful, inthese days of regular time signals,whether extreme accuracy is necessaryor even greatly advantageous in adomestic clock. I know that experthorologists will hold up their hands inhorror at such a suggestion, but theaverage householder is usually satisfiedwith a clock which keeps time withina few seconds a week, sufficient toavoid the risk of losing trains or otherimportant appointments. The goodold grandfather clock, with its plainiron wire pendulum rod, will fulfil thiscondition if carefully adjusted, and sowill h u n d r e d s o f m o d e m mass-produced and not very expensiveclocks. By all means pursue the ulti-mate in precision timekeeping if youwish (you will, I think, find it a full-time hobby !) but do not imaginethat it is just a matter of using acompensated pendulum, as there ismuch more to it than that.

The pendulum rod in this clock ismade of wood, which has beenexploited extensively for this purpose,as it is subject to very little thermalexpansion in the plane of its grainfibres. With a very heavy pendulumhowever, such as is necessary or atleast desirable in an electric impulseclock, it is necessary to exercise somecare in the attachment of the rod to itssuspension and also to the fittingswhich support the bob, as these aresubjected to considerable strain.

The wood used is a length of 1/2 in.dia. hardwood dowel rod, the mostsuitable varieties of wood being thosewith long fibres such as ash, oak orbirch; birch, I believe, being the most

15 MAY 1958

i

readily available, as it is extensivelyused in the furniture trade-mostlyas a substitute for rectangular tenonsfor jointing. The particular specimenwas slightly oversize, even after it hadbeen glasspapered down to a goodfinish in the lathe, but this is not ofthe utmost importance so long asother parts are made to fit.

After cutting to overall length, thetwo ends were reduced to 3/8 in. dia.for a length of 1-3/4 in., holding the rodin a split bush in the self-centringchuck. If the lathe does not have ahollow mandrel, or if its bore is toosmall, this may not be a very easyoperation to carry out, but it isgenerally possible to rig some kindof outboard bearing or steady, andalso a hand-tool rest, for dealing withit. The thread only needs to be arudimentary one, and is best producedby the old-fashioned kind of splitdies, which tend to compress thewood rather than cut it; the solidcircular dies push the chips in frontof them, and may strip the threadsby end pressure. Alternative methods

15 MAY 1958

are screwcutting o r hand chasing.Note that the thread does not extendfull length of the reduced end portions,a plain part being left to ensureproper centring in the socket or ferrule.The rod should be given two or threecoatings of thin shellac varnish orfrench polish to seal the pores, thethreaded parts being liberally treatedor a thicker consistency of varnishused.

The ferrules at each end are madefrom brass ‘rod, which need not bemachined outside if they can beaccurately chucked. Both are drilledand tapped to a depth sufficient toclear the threaded ends of the rod,the mouth of the hole being opened outto the crest diameter of the thread fora depth of about 3/8 in. The top ferruleis flattened on two sides, and slitdown the centre to a width of 1/16 in.and a depth of well over 1/2 in., with a3/32 in. hole across the bottom of thesawcut to improve flexibility, and also,by breaking into the axial bore, toform an air vent when the rod isscrewed in; note that a similar cross

621

hole is drilled in the lower ferrule,though it need not go right throughto the other side.

The lower extension of this ferruleis turned down and screwed 1/4in.tine thread (exact pitch is optional,but fine thread is desirable for nicetyof adjustment) and pointed at the end,the idea being to enable the pendulumto be used as a plumb-bob whensetting up the clock, or to indicatethe arc of swing.

To attach the ferrules, shellacvarnish may be used as a cement,the insides being well coated andallowed to dry, after which they arewarmed up to soften the varnish,and screwed well home to the shoulder.This gives very satisfactory results ifthe threads are a good fit, but if theyare slack it may be found safer touse resin or similar heat cement;alternatively Bostik or other plasticadhesive may be used, but not glue,or any cement which becomes brittlewhen dry.

l To be continued

MODEL ENGINEER

The ME JubileeELECTRIC CLOCK

Continued from 15 May 1958, pages 619 to 621

The construction of the pendulum, its suspension, and theassembly of the detent lever are reached in this article

By Edgar T. Westbury

A MIID steel pendulum b o bwas used in the actualclock, the dimensions being

8 in. long x 3 in. dia., and theweight approximately 15-1/2 lb. Anyother convenient-sized piece ofmetal may be used; some con-structors may prefer to cast alead bob inside a piece of brasstube, which certainly enhancesappearance, though it confers nopractical virtue, and both brassand lead are more expensive metalsin relation to weight. If nothingelse is available, a simple woodpattern may be made and an ironcasting obtained.

It is not essential to machine theoutside of the bob except on the scoreof appearance, but the hole should beas truly as possible in the centre, andthe lower end face, at least, shouldbe square with its axis to provide atrue seating for the rating nut. Thecentre may be marked out on the twoends of the bob by the use of any ofthe well-known centring devices, orby surface gauge and V-blocks, thendrilled with a centre drill, after whichit is possible to, drill the bob fromeither end, by holding it against theback centre and running fhe drill inthe lathe chuck.

It is, however, much more com-fortable and also more accurate, tochuck the bob and drill it from thetailstock in the approved manner.Unless a very large lathe is available,the overhang of the heavy bob willnecessitate fitting a steady on theouter end, but the standard three-point steady for 3-1/2 in. lathes will nottake so large a diameter as 3 in.

This difficulty was, however, sur-mounted in my case by making animprovised wooden bush steady, bybolting a piece of hardwood to a foot

MODEL ENGINEER

made of angle iron, clamping thelatter to the lathe bed, and boring ahole to fit the bob by means of acutter bar held in the chuck. It wasthen a simple matter to face andcentre the bob, and drill it from eachend, as shown in the photograph;this also enables the hole to bereamed or otherwise opened out, ifit should be necessary, to fit the rod.It should slide easily on the latter,but not with excessive clearance, asa sloppy fit may cause it to rock fromside to side,. and interfere with theproper working of the clock.

Pendulum suspensionThe method of suspending the

pendulum on a flexible spring followsthe principles which have been uni-versally employed’ in. clocks forhundreds of years, and are verydifficult to improve upon. Variousother devices have been tried, in-cluding ball bearings, cone pivots,and knife edges, but so far fromoffering any advantages, they havegenerally been found less durable andefficient.

One merit of the suspension springwhich is not always realised is that itprovides some slight correction forthe “ circular error ” which occursin a pendulum swinging in a true arc.

A well-tempered flexible steel springproduces very little friction, becausepractically all the work done indeforming it is given back on thereturn swing by its natural elasticity.It is, however, very important thatgood quality spring steel should beused, as fatigue must be considered,and low-grade material will becomebrittle in a relatively short time.Shim steel is sometimes recommended,but this material usually has a lowcarbon content? and such elasticityas it possesses i s mainly due to coldrolling; I should hesitate to guarantee

718

it for this purpose. Fortunately, thecorrect grade of spring steel, speciallymade for clock springs, is not difficultto obtain, or very expensive.

The two leaf springs used in thesuspension of this clock are sand-wiched between brass cheeks, or“ chops ” as they are called in thetrade, and secured by small rivets ateach end. Tempered spring steel isdifficult to drill, and it will be foundeasier to punch the holes with asquare-ended blanking punch, thoughlocating them properly is not so easy.Centre-punching is not recommended,as it is liable to distort the spring.

The method I have used is ratherunorthodox, but produces satisfactoryresults; I make the cheeks by foldingover strips of sheet brass as shown,and drilling the rivet holes in themfirst. Before closing them right down,the leaf springs are inserted, and alsoa slip of foil between them, the samethickness as the springs, to take upspace. The assembly is then clampedtightly in the vice, so that the springsare held in place, but can be adjustedif necessary, after which the punch islocated by the holes in the cheeks,and driven through with a smart blow.

If drilling is preferred, a glass-hard spearpoint drill, or a dentalburr, will generally penetrate thetempered steel if lubricated withturps and run at low speed. The rivetsshould fit the holes closely and shouldnot project too far, so that they canbe closed up with a few light blows;brutal bashing is more likely to distortthem and put the assembly out oftruth. After one cheek has beenriveted, the location of the springsin the other cheek should be checked,so that the assembly is square and thesprings parallel to each other. Theseprecautions are highly essential, asan untrue suspension spring causesthe pendulum to roll, and sound

5 JUNE 1958

assembly is necessary to carry the which it should fit closely, and pendulum rating nut, which is quitevery heavy weight of the bob. clamped by a 4 BA steel screw. When a straightforward job, and can beThe suspension bar is made in two the bar is placed in the V-notches of

parts, which screw together and the suspension bracket, it will take upturned, drilled and tapped at oneoperation. Care should be taken to

clamp the top cheek of the suspension its proper alignment, and can finally produce a good thread in view of theassembly, the lower cheek being be secured by the clamp bars. load it carries; the tine thread em-inserted in the slotted end of the rod, Little need be said about the ployed here, and on the end of the

52

InrrLY c-rn.Tc ,.I-.,.

FOLDED OVER

SUSPENSlON BAR BRASS) &(I OFF 2-OFF

-1BALANCE

I 1 \ WEIGHT

DETENT LEVER ASSEMBLY

1

SPLIT CLAMP I OFF BRASS

1 OFF NICKEL ALLOY

I I 11 .‘!’ Ii --

2 HARDENED STEEL

5 JUNE 1958 719

P I V O T P L A T E _ _ I O F F B R A S S

MODEL ENGINEER

rod, is conducive to precise adjust-ment, but may be varied if desired.Some constructors may prefer toserrate the edge of the nut, withequal-spaced notches to indicate somedefmite increment of axial movement.

In describing the detent leverassembly and other details of thecontact-makine mechanism. it is neces-sary to point out that some modifica-tions have been introduced sincethe general arrangement drawingshown on page 571 of the Jubileeissue was made. The latter should,therefore, be taken as explaining thegeneral principle and order of as-sembly only, in which respects it isquite correct, but details have beenaltered.

Lever assemblyThe lever assembly is carried on a

split clamp mounted on the pendulumrod, and is thus adjustable for heightand alignment. Brass is specified forthe clamp, but it may with advantagebe made of lighter material. such asaluminium alloy or laminated bakelite.It may be machined in one piece andsplit afterwards, or two pieces may beused, by drilling and tapping the screwholes and fixing them together with aslip of packing between for the boringoperation; in either case the holeshould be bored to fit easily on therod, so as not to jam on the “ horns,”when the halves are fitted to the rod.The top and bottom faces of the clampare trued up, and it is advisable tostamp or otherwise mark them forrelative location.

To round off the ends of the clamp,it can be secured by its own fixingscrews to a plate or block which canbe held on the faceplate or in the four-jaw chuck. An alternative method isto drill and tap a hole in the centre ofthe rear face so that it can be screwedon to a stud and held in the three-jaw chuck. The shape of the ends, ofcourse, has no vital significance, andthey may be filed or machined to anydesired contour, so long as they areneatly finished.

The pivot plate has been madeadjustable in the horizontal plane,and placed at the front of the clampfor ready accessibility. It is madefrom 1/16 in. sheet brass, with the pivotbush spigoted and lightly riveted in.For fixings of this nature, includingwheel collets, it is desirable to makethe spigot a light press fit, and of alength to project about l/32 in. bevondthe plate, which in this case is 3/32 in.

The centre hole of the bush shouldbe left slightly undersize for finishingafterwards-for a 1/16 in. pivot, a No 5drill may be used; and the spigot endshould be bevelled inwards to ashallow angle, so that it can beriveted over without burring up the

MODEL ENGINEER

hole. Support the other end of thebush on a soft metal or fibre pad forriveting, to avoid damaging themachined face. and take care to keevit quite square with the plate. Ifdesired, the fixing screws of the clampmay be modified so that the plate canbe adjusted without loosening theclamp on the pendulum rod.

For the detent lever 1/16 in. sheetbrass is used, cut to the outline shownand drilled and tapped in three places.The 10 BA hole for the pivot should betapped only partially through with ataper tan. so that the pivot will screwin-tightly: As a matter of fact, 1/16 in.rod will not allow of producing a full10 BA thread-it is nearer 11 BA,which may be beyond the scope of theequipment in many workshops. I wasable to obtain silver steel rod approxi-mately 0.067 in. dia., which i s justright for screwing 10 BA, and thepivot. bush was, therefore, opened outto suit.

Another minor alteration of detailin this assembly is seen in the designof the contact wiper, which wasoriginally made in the form of a discmounted eccentrically for the purposeof adjustment. It has now been foundmoreconvenient to provide the adjust-ment on the contact blade, whichenables the wiper to be made simplerand lighter, namely in the form of asmall bush, which is secured to thelever by an 8 BA screw and washer.The best material for this is silver.but nickel alloy or “ German silver ”appears to give quite good results.

Hardening the detentThe actual detent which engages

with the riding wheel is made from1/8 in. silver steel, turned down at theend to 0.086 in. and screwed 8 BA.After locating its position when it isscrewed tightly i n t o the lever, theouter end should be marked acrossthe diameter, square with the lever,and filed half away as shown in theplan view. This part of the detentshould be hardened and polished, butthe screwed end should be “letdown” to deep blue, by holding thecutaway part in the jaws of a handvice or pliers and heating the end ina bunsen or spirit flame. The workingsurfaces, both flat and rounded, shouldfinally be polished.

It may be observed that while manyliberties can be taken with the designof this assembly, the lever should bekept very light so that it does notproduce unnecessary friction when incontact with the riding wheel. Theobject of the balanceweight at theright-hand end is to take some weightoff the detent, and it may be foundworth while to experiment by adjust-ing the size of this or adding metal toit, but a point will inevitably be

reached where the effective weight ofthe detent is not sufficient to ensurereliable engagement.

In order to fit the pivot to its bushin such a way that it works quitesmoothly and freely, but withoutslackness, I recommend making areamer or D-bit from the same pieceof silver steel rod as that used for thepivot itself. This is made semicircularin the usual way, but instead of theflat being parallel to the externalsurface, it is made with a long taper,starting well below the half-diameterand running right out to full diameter.

Finishing the pivot holeAfter honing the flat face,. it is

hardened right out in thin oil. Itshould not be expected to take outmore than about 0.001 in., but itschief merit is that in the final stagesit burnishes the bore of the hole andleaves a high, accurate finish; if thepivot is found to be slightly on thetight side when fitted, a little lappingon a spare bit of rod, using platepowder or oilstone sludge, will enableit to work freelv. This simple tool isapplicable to the finishing of allpivot holes; and can be made in afew minutes in any size required.

The outer end of the pivot studmust be provided with some formof retaining device, and although it isnot orthodox horological practice, theeasiest thing is to screw it and fit anut. This should screw on just farenough to allow the pivot to workwith slight but not excessive end play.A mere suspicion of recommendedclock oil such as Chronax or Ragosine,made by Rocol Ltd, Swillington, nearLeeds, should be applied to the pivoton final assembly.

l To be continued

Plug Cleaners

THE Birmingham Manufacturers andTraders company, of Masterpiece

Works, Aston, Birmingham, ask us topoint out that sparking plug cleanerswhich were described in the article“ Blast Those Plugs ” (24 April 1958)are similar to those manufactured bythem, still in production, and on saleat motor accessory dealers.

The author of the article concernedhas assured us that he wrote thedescription in good faith and with nointention of encroaching on the rightsof the manufacturers. We are gladto publish this assurance a n d atthe same time draw the attention ofreaders to the plug cleaners which aremanufactured by Birmingham Manu-facturers and Traders Ltd.

They are operated from motorists’foot pumps and in a few secondsthoroughly clean sparking plugs. Theretail price is 16s. 6d.

5 JUNE 1958

ME Jubilee electric clock

Making thecontact unit ’

EDGAR T. WESTBURY deals in this articlewith one of the most vital parts of the mechanism

Continued from 5 ]une 1958, pages 7I8 to 720

As mentioned in the previousarticle, several detail modi-fications have been made

in the clock mechanism since thegeneral arrangement drawings weremade, the object being to simplify

construction and facilitate adjust-ment. In the case of the contactunit now to be described, theseinclude increasing the diameter ofthe riding wheel, and altering thefittings which control the timingand period of contact.

The entire unit is mounted on apanel,. which is specified as being ofPaxolm, a proprietary form of lami-nated phenol resin bonded plastic;other very similar materials areTufnol and Delaron. No doubt thereare many other kinds of plasticswhich would serve the same purpose,

and it is also possible to substitutewood or metal, but whatever materialis used, it is most important that itshould be stable, and not liable towarp or distort m any way.

While cutting out the panel, it maybe noted that the external dimen-sions are the same as that of the

=?F

Complete contact unit mounted on.its panel (backstop disengaged)

MODEL ENGINEER

panel which carries the electro-magnet,and advantage may be taken of thisfact to cut them out, or at least finishthe edges, at one operation, by firstmarking out and drilling the fourcorner fixing holes, and bolting thetwo pieces together. The roundedends conform to the shape of theseatings on the chassis, and chamfer-ing of the edges is specified, partlybecause it is difficult to ensure per-fectly clean corners when filing ormachining plastic materials of thiskind. In my case, I produced a hollowchamfer or flute all round the edgesby a simple milling process, using around nose cutter in the lathe chuck,and a flat saw table on the cross slidewith a 1/8 in. strip attached to act asa fence, limiting the width of thechamfer. The work was appliedby hand, feeding from the back ofthe lathe (if the cutter is run in the

normal direction) and the process isthe same as employed in woodworking,using a spindle moulder. Obviouslyit would be just as easy to produce amore elaborate moulding, by the useof an appropriate cutter.

The contact block is made fromsimilar material to the panel, the sidesand edges being finished square andtrue, and the cross groove producedpreferably by milling. It is attachedto the panel by two 4 BA countersunkscrews from the back, and care mustbe taken to set them flush or slightlybelow the panel surface, so that theydo not interfere with its proper seatingon the chassis. This applies also tothe other two screws, which hold theplate from the back.

Wheel mounting frameThe riding wheel is mounted in a

little motion frame of its own, con-sisting of a square plate, two turnedpillars, and a front bar. Both theplate and the bar are cut from 3/32 in.sheet brass, and must, of course, bequite flat and true. As the centredistance between the two pillars istaken on a diagonal, there may bea slight discrepancy between thestated dimensions on the plate andthose on the bar, but it is quiteobvious that these must be made tocoincide, and this can easily be en-sured by using the part first drilled asa jig for the other.

This applies equally to the centrehole for the pivot, which should hedrilled undersize and finished byreaming, as described for the detentlever bush; the same tool should besuitable, as the pivot can be madeto fit. Note that pivot holes should

19 JUNE 1958

alV

Liz:the

%

Emaup

1wh

Zin 1to1betSFUI

SPil713.

Rid7

Pasowitheobt,bycut1

I

is iskg,uslyteause

romidesandIced:hedunklust

?ltL$ting1 tothe

n aon-nedthei n .be

itrei sbe

themdliteto

en-! as

ltre

:;entbe

ldeuld

958

always be countersunk on the outside,to form an oil sink ; a very slightbreaking of the corner is also desirableon the other side, to eliminate theburr, or the possibility of riding onthe fillet of the pivot. These operationscan best be carried out by means of aflat drill, ground to 90 deg., with nottoo fierce a cutting angle, and twirledbetween the finger and thumb. Finallymark the two parts, to show which wayup and which way round.

The pillars are a plain turning job,which may be carried out betweencentres, after first deep centring anddrilling (but not tapping) the holesin the ends. It is, of course, importantto ensure that the length of the pillars,between shoulders, should be thesame within a few thou, and that thespigoted ends should fit closely in the7/32 in. holes in the plate and bar.

Riding wheelThe production of this wheel may

possibly worry some constructors,owing to its specialised design, andthe improbability of being able toobtain one ready-made. It is, however,by no means a difficult job, and thecutting of the teeth can be carried out

without very elaborate equipment.Some readers are no doubt quiteequal to the task of dividing out andfiling the teeth by hand, as the con-tours are simple, and the standardof accuracy much less exacting thanthose of a gearwheel. But for thoseless highly skilled with hand tools,milling with the aid of a simpleindexing gear is recommended.

A pierced wheel blank, if obtainable,may be used with advantage, but other-wise it may be cut from 1/16 in. hardbrass sheet, drilled and reamed fin.dia., and mounted for external turningon a shouldered and nutted mandrel,which serves also to hold it whilecutting the teeth. Here, it may beobserved that it is well worth whileto make two or three wheels at once,there is not only the possibility thatone might be scrapped in the course offurther operations, but a pack ofwheel blanks will be much morerigid to withstand side pressure duringthe cutting of the teeth. If a singlewheel is to be cut, it will be desirableto provide check washers of at least1-1/2in. dia. to support the blank, Iomitted these in my first attempts,and nearly came to grief as a result.

19 JUNE 1958

Cutting main teeth of ex-perimental ratchet wheelswith a home-made cutter

The cutter employed in my casewas a home-made one, turned from5/8 in. dia. silver steel, with an integral1/4 in. shank to hold it in the colletchuck of my milling spindle. At thebusiness end, it was turned to areverse cone with an included angleof 50 deg., and the end face was well

MODEL ENGINEER

undercut to give clearance. Threeaxial grooves were then milled toform teeth, with deliberately in-accurate spacing; these were, backedoff, the cutter hardened and tempered,and finally honed.

Cutters of this type can be made ina matter of minutes, and cut brassvery efficiently if kept sharp (as theyeasily can be) and run at high speed.Alternatively, a flycutter, or a multi-tooth angle cutter of suitable shape,can be used. Incidentally, I had tomake several cutters in the course ofexperimental work with teeth ofvarying depth and rake, some of themhad up to five teeth, but did not appearto work any better, and were moredifficult to hone by hand.

For the cutting operation, themandrel with the pack of blanks inposition was held in the chuck, infact, at the same setting as for turning,thereby avoiding possible inaccuracy.The milling spindle was mountedhorizontally on the vertical slide, atright angles to the lathe axis, anddriven by a plastic belt at a speedof about 3,500 r.p.m. Indexing wascarried out by means of a 60 t changewheel mounted directly on the lathemandrel by means of an expandingbolt, and locked by a strongly spring-loaded plunger (no trains of wheelsor bits of clock spring for me,thanks !)

Alternative methodsFor those who do not find it con-

venient to fit up a rotating spindletype of milling attachment, the alter-native arrangement of running thecutter in the lathe chuck, and mount-ing an indexing attachment on the

vertical slide, is equally practicable.Full details of these and other millingoperations, can be found, in the MEhandbook Milling in the Lathe.

In setting up the gear for thisoperation, it is important that thecross position of the cutter relative tothe blank should be properly located,so as to produce the specified rakeof the teeth. To do this, the cuttershould first be adjusted, by puttinga point centre in the mandrel socketand lowering the spindle to its levelby the vertical feed, then moving thecross slide till the front edge of thecutter is exactly in line with thecentre point.

If the wheel was cut with the cutterat this cross setting, the teeth would beexactly radial, or in other words,would have zero rake, it is, therefore,necessary to make a further adjust-ment of the cross slide, beyond thecentre position, to produce the re-quired positive rake. As the radiusof the blank is 1 in. and 10 deg. isequivalent to approximately 2.100 in.in 1 ft it follows that the amount ofextra movement required is 0.175 in.

MODEL ENGINEER

Cutting the stepsof the tee thSecond operation

The main portion of the tooth maybe cut at one pass, and the cutterdescribed is quite equal to the job,but it will usually be found moreconvenient to take more than onecut all round at a time, or at leasta roughing cut and a finishing cut.No exact depth of tooth is specified,as it is not critical, but a trial cut onone or two teeth, will give a goodindication of final depth, whichshould be such as to leave a l/32 in.land at the tips. When once set todepth, the vertical slide should belocked by tightening the gib strips,likewise the cross slide to avoid riskof indexing errors.

Mounting on the colletAfter finishing the main teeth, the

cutter should be re-set by moving thecross slide a further l/32 in. and thenick, or step, cut in the tips of theteeth. It will probably be found thatburrs will be produced, and possiblysome roughness of the milled surfaces,but these can easily be removed bythe judicious use of a dead smoothpivot file.

The wheel is mounted on a hubor collet, which should be turned,drilled and reamed at one setting toensure concentric truth, the end of

the spigot being bevelled inwardly sothat it can be riveted over withoutburring up the hole. A spindle orarbor is made from 1/8 in. silver steelby chucking it truly and turningdown at each end to form the pivots,the accuracy and finish of whichshould be as high as possible to ensurefrictionless running. I hope to beable to give some information onpivot polishing when dealing withthe motion work of the clock, butthe subject has already been dealtwith by much more competent clock-makers than myself.

When in position in the motionframe, the arbor should have a slightamount of end play, and the colletshould be a light press fit on it, asthis is much the simplest and mostsatisfactory method of securing it inposition, and if necessary, adjusting itendwise to coincide with the plane ofthe pendulum swing. As alreadyexplained, room is allowed on thearbor for the fitting of extra contactgear to energise secondary clocks ifdesired, but this matter need not bediscussed further at present.

It is necessary with any form ofratchet operated feed device to providesome means of preventing the wheel

l Continued on page 804

19 JUNE 1958

Pan0pro]strabutmer

Thesholandplatthewarapp

A

19 JZ

CASEY JONES POEMSI R,-As a boy in 1913 I used to

watch the trains on the LBSCR lineand I well remember two Pacific tankengines named Abergavenny andCharles C. Macrae. In all his LobbyChats, I cannot recollect LBSCmentioning these engines. What istheir history ?

LBSC sometimes gives us a hintof an apparently well known railwaypoem all about a driver namedCasey Jones, and how he used todrive his engines till they groaned.Neither I, nor any of my fellow clubmembers have come across the fullr;Fusould he dig it out and give

Grpingtdn, Kent. HAYDN. D. SMITH.

GARDEN TRACTORSir,-Would any readers be in-

terested in model common-or-gardenengineering ? Here is a photographof a garden tractor just slappedtogether from what was lying around.The engine was borrowed from anothertask. Even then, bolts,. chain, andshaft ran to $50. Everythmg costs toomuch these days. The engine is2-1/2in. x 2-3/4in. (3-5 h.p. AmericanWisconsin), and it needs another5 h.p. of muscle to hang on to it.Nothing stops the engine even if aplough catches. Cleats or spuds justdig down and the whole business willbe through to China in no time.

As a lad in Hertfordshire in theearly years of the century I was ableto see the steam ploughs in operation.I always thought the fella on theplough was due for decoration (aboveand beyond the call of duty).

I considered some winch arrange-ment to tear up my cabbage patch(75 ft x 200 ft) but I should needsteam to do it’right. I like to seethe pictures of those large ploughingengines. Incidentally, it was under-stood that the crew made big moneyin those days; £ 15 a week beingmentioned at times. Probably owneroperators.Vancouver, 14 BC. R. C. FAJRALL.

METALLURGICAL BOOKSSir,-In modification and ampli-

fication of the reply to S.L.H.‘sinquiry concerning early springs, mayI suggest that reference is made tothe following books: Metallurgy, Ironand Steel, by Dr John D. Percy.Published 1864 by John Murray ofAlbemarle Street, London. -Metal-lurgy of Iron, by H. Banerman, F.G.S.Published 1868, 1871 and 1874 byLockwood and Co., London. Ironin All Ages, by James M. Swank.

Published 1884 (1st ed.) and 1892 byAmerican Iron and Steel Association,Philadelphia. Manufacture of Iron, byFrederick Oerman. Published 1854

MODEL ENGINEER

(3rd ed.), published at Philadelphiaand by Trubner and Co., 12 Pater-noster Row, London. Metallurgy ofIron, by Thomas Turner. Published1895-1918 (5th ed.), by CharlesGriffin and Co., London.

All these contain excellent, andaccurate, information on the materialsS.L.H. IS interested in, but un-fortunately without going into greatdetail, the answers given in the lowerhalf of column three page 469April 10, 1958 issue are, in manyphases, inaccurate and misleading.All the references cited in turnfurnish other references, and theinteresting story of the hows and whysof early iron and steel manufacturemay be readily tracked down.

Incidentally, as Hexham appears tobe reasonably close to Newcastle upon

The garden tractor which Mr Fairallbuilt from a collection of oddments

Tyne, perhaps S.L.H. could contactthe metallurgical people at the Univer-sity of Durham and obtain moreguidance in his sleuthing !Birmingham, R. V. HUTCHINSON.Michigan.

SUNDIALSSIR,-I am very interested in sundials,

especially the more elaborate types,and when in Britain recently thor-oughly enjoyed looking through thecollection in the Science Museum.

I also saw a very tine one in theBodnant Gardens (a National Trustproperty near Llanrwst in Wales) butthis was unfortunately incomplete.1 wonder if anyone could give medetails of this, or any other accuratetype of dial so that I might be able tosatisfy my ambition to make one ofthese interesting and ornamental time-keepers. Perhaps you could refer meto a good book on the subject.

I look forward very much to thearrival of my copies of ME, whichusually come three or four together !Being at present away from my home-land (Australia) and my workshop, Ican only look forward with pleasur-able -anticipation to the day when Ican remove all the rust-protectivecoatings from my lathe and tools andget to work again.Christchurch, K. G. ANDERSON.N e w Z e a l a n d .

ME Jubilee electric clockContinued from page 784

from (running back on the reversestroke of the pawl, or detent. In thisparticular case the backstop, as it iscalled, presented a special problem,as an ordinary click, or spring detent,would not necessarily stop the wheelinvariably in the right place, due tothe presence of the step, into which thedetent might engage mstead of theroot of the tooth. In early experi-ments, a form of silent or frictionratchet was tried out, in which agrooved wheel on the arbor wasengaged by an eccentrically pivoteddisc if the wheel tried to run back.This was certainly silent. and provideda remarkably powerful one way grip,but still the indexing was not positiveenough for the purpose, and othermethods had to be found.

The backstop gear finally employedmay possibly look like an after-thought, but its position on the bottomend of the pivot bar is ideal foraccessibility and adjustment, so I havenot attempted to alter it very muchfrom its experimental form. It willbe seen that instead of a detent, thisemploys a roller, which cannot engagein the step, but must necessarily go

804

fully into the root angle of the tooth.The lever is made with a counter-

poise at its tail end, so that it engagesby gravity and it may be foundnecessary to adjust the weight toensure that it acts quite positively butwithout imposing more friction on thewheel than is absolutely necessary. Idid this by -fitting an 8 BA screw inthe tail and adding one or morewashers, but a neater method wouldbe to solder a little extra metal onthe back, should it be required.

The construction of the lever andits bracket calls for little explanation,as both parts are quite straightforward.Both the pivots should be turned fromsilver steel and highly finished, theends being threaded only just farenough to enable them to screw stifflyinto the plates; lock nuts could befitted, but should not really be neces-sary. The roller should spin freely onits pivot, and after fitting should be hardened and polished. By means ofthethe

slotted bracket, it is easy to adjust

theposition of the backstop to arrestwheel dead, with no run back

whatever. A washer may be desirableunder the screw which holds thebracket, but I have used special largeheaded screws which serve the purpose quite satisfactorily.

l To be continued

19 JUNB 1958

03modelstrengand Iavoidi

E:;smallpleasescatemWatfo

st!%lnew r5 todimenard; theadsand SCWhist4tered,andfir

AIlCwhichengineof ga1back \:fied b]so aslevel.metersselecte

19 JIJb