LW lil THESIS Lae PROSE Lege) My Or eh Um MLO TU ae Rome as UTM ib) 2 ie 1922 Bein MEP ST aes 720 DASE ae oe a oy ad . j - * - - - lee oS - e . —_ : - ~ - aren &. Le ee. - «@ Mee.- =. .. wen See ™ -~ 7. ee oo a ~ os + Roo = a °. : ‘~ BY A ae ee a at oe edad Sone ten merienenere ge ania nag nn ene nanni habeas: nie, eer ee ee ena nanan nt ind sealer aa aeemae ieee The Design and Construction of a Synchronous Commutator for High Voltages. Submitted to the Faculty of The Michigan Agricultural College by Walter S.Bersey Paul V.Nelson Candidates for the Degree of Bachelor of Science Jume 1922. THESIS ( C tr. re Table of Contents. an ee ee ee l. Preface 2e Discussion Be The A»paratus (a) Comnutator (b) Synchronous Motor 4. Photogranhs de Tests 6e Results 7. Conolusions 93916 Preface. The object o* this thesis on the construction of @ Synchronous comnutator, was primarily to provide a means of obtaintng a hign potential direct current ,ané secondarily, for the »ur»vose of leaving something more than a typed covy of a summation of ideas. It has been the desire af the authors to produce something tangible which would be useful and which could be left as an addition to the anparatus of the Electrical hmgineering Department. In the design and construction of this apvaratus but few derivations, calculations, and standard formulae could be uted. This is because of the fact that there is but little da@a available on this type of rectifier and the work of this thesis is largely experimental. The working qualities and precision of any apparatus approach the highest desirable standards only after much designigg pbuilding, rebuilding and improvement. This being the first apparatus of its nature to be brought into the department, it is the hope of the authors that it may further the incentive, and furnish the starting point for further experiment and research along the l&ne of synchronous connutat- ors In choosing the methods of procedure and determining the lines along whigh the construction should proceed, the authors are greatly indebted to Professors A.R.Sawyer M.M-Cory, 1.S.Foltz and E.E.Kinney of the Electrical Engineering Department for their supervision, advice and nany timely suggestions. We wish also to thank Mr.A.2. Krentel of the Wood Shop and Mr.G.C.Wright of the Machine Shop for their many accomodations and their kindly assistance. WeSBersey. P. V.Nels ONe Discussion. Electrioso currents are classed under two general headings: alternating and direot. Oftimes both are needed where only one is available and it then becomes necessary to devise some means of conversion whereby one typw of current may be changed to the other. There are many methods of brf&nging about this change, and each method has its advantages and applications. The most important methods of rectification are:Rotary Converter, Motor Generator, Mereury Are Rectifier, . Vacuum Tube, Vibrating Rectifier, Electrolytic Rectifier Point to Plate Discharge Rectifier ,and the Synchronous Commutator. Of these the most common are the rotary converter, motor generator and mereoury are rectifier. The vibrating rectifier ig also commonly used with small battery charging outfits. For the conyersion of large amounts of power the rotary converter ia@ andoubtedly the best adapted, and it can be used for conversion of A.C. to D.C. or D.C. to A.C. The motor generator set is Usually more costly, while the mercury are rectifier is limited in size and annlication. Among those thus far discussed, there are none which are adavtible to high notentials. The »oint dischage rectifier and the synchronous commigfator erc th2 only ones which lend themselves to use with high voltagese Of these, the poimt discharge rectifier ic at once eliminated, since it uses only one half of the A.C. wave and is consequently very inefficient. The synchronous commutator is atype of rectifying apparatus which may be eonstructed to meet nearly any condition of potential or current.It has the one dis- advantage that it does not give satisfactory results when operating inverted,( i.e., converting D.C.to A.C.) as does the rotary converter. On the other hand, the rotary converter or motor generator ean not be be used in handling high votentials on account of insulation troubles. Seeing the neeR of a machine which will produce direct current at high voltages, the authors have des- igned and constructed a synchronous commutator which will convert high voltage A.C.to D.C. at the same voltage fhe design and construction are discussed in detail on the nages following. The Commutator. A synchronous commutator such as is heréin described, is a machine which rectifies alternating current and produces direct current from it, The resulting current is a direct current in that its flow ig in obly one direction, but in.a strict sense it is a pulsating current, the pulsations all being in the same direction. The fundamental primciple of the commutator is that of a reversing switch which reverses contacts when the alternating current reverses direction. The reverskbng of contacts must be periodic and of twice the frequency of the alternating current to he rectified. These conditions are most easily brought aboupf by a commutator which rotates at synchronous speed. Around this commutator are placed brushes to furnish the alternating current and to carry away the direct current. With a 60 eyole current and a synehronous apeed of 1800 repem. there would be two 90 degree segments on the preiphery of the commutator 90 degrees apart. Then, at the instant that an A.C. brash was positive, ene segment would connect between the A.C. brush and the positive D.C. brush. As the A.C. brush becomes negative the segment rotates to a position such that it now connects the same A.C. brush to the negative D.C. brush, and simultaneously the opposite segment connects the other 4.C. brush, which is positive, to the positive D.C. brush. Thus we see that the B.C. brushes maintain the same polarity though the 4.C. brushes are rapidly changing from positive to negative. The commutator herein designed is to be used for high voltages. The high voltage 4.c. is obtained from a special transformer. Thiantransformer was equipped with a magnetic shunt wherby the secondary voltage can be varied from a small potential to about 20,000 volts.Commutating difficulties are often experienced with high preipheral velocities but with such high potentials it is not necessary that the brushes make actual contact with the commt- ator segments, thus this difficulty is eaaily over- come. One of the most important factors in the design of the commutator is the voltage at which it ia going to be used. The next factors of importance are; the speed of rotation, number of poles, and curren’ carrying capacity. One inch separation is commonly considered as sufficient insulation between conductors carry- ing 20,000 volts. But, due to arcing between the brush- es, and the segments, especially at the breaking points; i.e. when the segments leave the brushes, this separation of conductors must be increased considerably. In designing the commufator provision was also made for later using it at higher voltages. Since the brushes do not make actual contact with the commutator, the commutator is not affected by tye peripheral velocity, and the diameter of the commugfator is then limited ohly by its mechanical strength. A commutator diameter of 10 inches was deemed the best under the existing conditions. Excepting the rotary converter and the mercury arc rectifier, most rectifying apparatus is not very efficient. This is because of the fact that only a pert or parts of the alternating current wave are rectified. The commutating rectifier is inherently more or less inefficient due to its mechanical sconstruction. In order that it might rec- tify 100% of th 4.C. wave, if would be necessary that the brushes subtend an arc of 180 electrical degrees. This is obviously impossible for the tips of adjacent brushes would then be in contact and produve a short circuit across the high voltage A.C. source. Thus we see that decreasing the length of the brushes increases the gap between them and increasing the length of the brushes increases the portion of the wave that is rectifind. The most satisfactory length of brush is about 90 electrical degreeg,yhich is equivalent in this case, where there aretwe commutator segments, to 45 mechanical degrees. The potential of the rectified current depends upon that of the part of the A.C. wave which is rect- ified. If the phase position of the brushes ia such that the highest portion of the wawe is rectified then the potential of the direct current will be the game as the maximum A.C. potential. The D.C. potential may be varied from a maximum to nearly zero by varying the phase position of the brushes. The brushes were constructed of #16 gage brass strips 5 inches long, covering an arc of 45 degrees. The four brushes were held by lock-nuts upon 1/8 inch iron pipes Z3inches long. The brushes and holders were supported by four cross arms 90 degrees apart. see Fig.l. To the center of the cross-arms was attached a 4 inch rod which was free fo rotate in ite support, but he&d in any position by friction. See Pig.2. with this arrangement the phase position of the brushes could be set or varied at will, and would remain in the position set. Figures 1,2,5 and 4 show the sonstruotion and arrangement of the brushes and supyorts. The Synchronous Motor. Since the commutator whose construction was contemplated must revolve at synchronous speed, the problem resolved itgelf into that of produc- ing a synchronous motor of sufficient power to drive the rectifying equipment. A motor of the type used on the oscillografh was considered but it was thought that such a motor, if built in the size necessary, would be extremely difficult to start, since the atart would have to be made by hand. Accordingly, the aim became: to produce a self-starting synchronous motor. A four pole machine was selected as being probably the best in respect to speed, for the commutating eqhipment and subsequent develipments showed that a single phase machine would probably be the simplest to construct. Having available an old Westinghouse fan motor of the single phase, four pole, shading coil type; it was assumed that the power from such a ma- chine would suffice, and arrangements were made to produce a motor which would revolve at synchronous speed The stator of the motor, shown as Fig. 5, is of the salient pole type, and it was planned to build a rotor having small clearance, with four salient poles and a superimposed squirrel cage winding. It was thought that such a machine would start as an induction motor, due to the shift-~ ing flux produced by the shading coils, and when near syneohronous speed would pull imto synchronism. At synehronous speed it was expected to run ina manner similar to the oscillograph motor. However, after building the motor as indicated, if failed to start, although, if started and brought up to synchronous speedit ran satisfactorily. Since our aim was a self-starting synchronous machine, we were not satisfied, and decided to go on and produce such a machine even if we had to junk all our previous work and start over again. On analysis of the situation, it developed that having salient poles on both rotor and stator, the rotor was in a pesition of least magnetic rel- uctance when the rotor and stator poles were opposite each other and the motor was unable to develop sufficient starting torque to pull the poles apart. a It seems that it might be posgible to accomplish this by having a heavy enough shading winding. However, we were hampered in making any changes along these lines by lack of reom on the stator. The rotor as uged is shown in Fige’7. The solution of the diffieulty meant the con- struction of another stator with a distributed wind- ing se built that the motor could start as a split phase induction motor and run as a single phase motor similar to the osoillograph mator,. It oan readily be see n that the stator would provide a uniform magnetic reluctance at all times during a revolution. Rather than to oonstruct a stator to fit the previously construsted rotor, we decided to pro- cure an induction motor and modify the seoontiary structure so as to be polar. fhe most accessible motor was found to be a 1/8 H.® General Blectric, single phase induction motor running at 1775 repeMe The motor is built with the winding on the retor and the secondary or squirrel cage winding on the stator. This, of course, meant that the stator would have to be made polar. The stator had 37 slots and, since it was desired to »roduee four salient poles, two stator bars were :enoved at each of four points ag nearly 90 degrees anvart as possible. The ineluded tooth was then remove’ leaving four polar me 2 OS WEage oe of Fo opr ork tor asoea oT *. : | * Oe nN rae . Ur gers 7 BE PIT. ~ Fort * o. a bt SLE ca PrP VBS 1 £ rE PON At - 4. Y iS Sys sae gt fovvecce orluep oy. Gy ee! erotic oi pe moe M4 Jaret ee eorkl eo cods eV ek ud gorirea 2k beer rs —MOe StF Sagat Stuer ef ot sity feu el —oaiw botedivtsts g@ sdtwoweu- ou: * Spe bE Teserd & CEfieosg MB es Prope Sfreeo raver enc oA. Tod O@ unk eae fe aFeate Bee tusr fro Mfoo; Ta. £ ace eT Gr Seg OP ts, crlinta wetes EPings woeteu. aw J, VEL Sot meo TL moped oh, he = oifo. cv. gerae foo gp eokyore encticfisver 8s 3 tech an der ef veadksate 8 tomrnrter ny ge nso vet tel jt OF bShtesh ow ,tctor Feiy co eb uy ieascivere wersorcose oft wrEbom Sirs re PRTo 8 boas otese crea ec rtesusecos Faom sd -T£I6O eet ote wu. ofnrourve oust sotrvochd Lstsnon ee) Ct US Cow ome ret CYNE Sp en kre soc rE eR ey git ca cantiiacty ott Ativ oF fis st wetter et 2 ta Rtulth euso Partkeps cl con enc s end Son retary wove a gay Fely MTree poaupee Sec geist? euetetes ott ree eyoafe Val aod tied g oN ete Top owe Gd ovat ofsray, Steoblen woh fegpleotke ¢. fo "Sf a Uf fSonmke , brie wegh FQ Tosa “ye oe poor yn OR eo hor gill eOnfdtusar aa Von G coset oa GO che por ap Borkor pg FQ curret Ce ype oo grat oo aw ¢ittood Be front projeotions witha superimnosed squirrel cage winding. ‘The motor was then assembled and run in this condition end was found to be almost but not quite synchronous in speed. Another tooth was then removed from each stator pole and the motor was found to revolve at synchronous speed wheh the applied voltage was 240. The rated voltage was 220. | Themaghine could probably be made synchronous at lower voltages by removing another tooth from each pole, but, since this would increase the exciting current it was deemed inadvisable.The stator as completed with its polar projections is shown in Fig.8. The speed of the motor could not be deter- mined accurately enough with a tachometer so to determine whether of not the motor was running synchronously the following was resorted to. The motor shaft and generator shaft were fitted with projections on their ends which would touch when they passed each other, thus making an electrical contact. When running synckron- wusly their relative positions were the same since the generafor used was also a four pole machine. By connecting a telegraph sounder circuit to the m= sro -b yes , TY we » * : ~ e e Soh ° ish B89 ot. “3 bs -D s+ 4 al “UO CVO via oe. so @ . 4 NPT rT ee i Srl \ 4 e SOSTO LD na on 32 8 oe ~”~ ¢ 5 hot ey : an 2 97. FS I ww. wom wh ce t w"F re ned us wei. rN eV 8 = 4 Se al al oEVv swo@oel ef =” o- mye f ) 4 Y , 4 - = woe abs wv *: ws ; bu UW “ 3 Wwe oO -— wel * +} rat . ‘ eg kde coe ‘a - v5 0 Ay DOSE * as J i 4 we wy m, wae ot! ~ ’ he MwA —T9 O re a. ’ {: —e9 ve . Jn _ 4 Mira ‘ we generator and motor an indication of the difference in speed was obtained. A periodic clieking of the telegranh sounder indicated a difference in speed. Another method of Judgine whether on not the motor is running synchronously is to listen for "beats" or "hunting". If beats are perceptible it is reasonable to assume that the motor is not running synehronously. . as e 4 2c + e ® co 1‘ t 7” - . r ot. ty) - 77 of . + a 1”? - . e- - oe we ~. d , 2 . ola. et Pe, } +e AL > - * Ne ". 8 alt ~s . - e sas e . *% ao 9. ho wet 7 " _™ + - > -- e, . : - + oy -y se os f*. ey o mye “4 4 r. aso ete ; @:. sia Be UO Tooke ina 4 wee et ee da dee w.. 41) c . e e ic - 6 r ’ f° ' , ° rn e - e . - oor \ > "¢: em ~ - ee a ”, . ° 7 - ‘ ~ OS, a) ~2 48D: { ya ewe . ~ 7 4 ryt POL re rc o a 71 ° Fr mo? > Tee Tyr fe ey oe. ° wae i go at. Jeeabu arth. ft ) a nd YY cer . “Fok ( r mMNy? t- ., C. s ”. os cr, ~ f i. . of eo. 2 Ney ewe 0 J "lL: actodat oO .7 A le ee ui SO, Woe] tba dy - s . ' e o- . r . . te . .f “Y * t . t ff ee vs , Oo 1c a? + oy fe . r 45 my: *y * ¢ ) é id web oe , ‘4 teve ‘ e bos, tiuel, . al eo. eet tow oe -°* ws od . ‘es te « ¢ - r . . 2 . No. ey pen enue ry swe. ae e i Mt SOL 5 cae addr ae peas Fige be . Stater of eriginal motores os ees Pee | ee ee t™S Roter of original motore 4p | Pige 7 « Phe wound rotore | } * - oe a F * >» / a « Pige 8e The stator as modified. Sésts and Results Having eompleted the construetien of the synehroneus motor ami the rectifier, the twe elements of the apparatus were agsembied en a suitable base together with a transformer whieh was for the production ef the high voltage alter- Rating eurrent. The assembly is shown in Figs.3,4, and 9, The trausformer is shown in FPige 10-¢ Since thea apparatus was constructed, it was deemed necessary toe tegt it te see whether or net it performed the work fer whioh it was designed. The first test was made by connecting a theroughly insulated D.C. milliamnetor in the D.C. eirouit which was sherted. The machine was set in operation end a small variable deflection was noted. This infieation inoreased as the voltage was inereaséé. The eurrent was of the order of 3c milliamperes, It goon became apparent that the machine was not suitable for any appreciable eurrent, for with mugh gurrent there wes considerabke arcing and consequent burning of the scommutator. Not feeling that the ammeter iest was conclusive we then turned to the oscillograph to inéleate more clearly ani vividly what was taking place. —— if = 7 . . ‘ 7 4 eo. . ' ~ > ' . . 4 ' ‘ a . . .. - ws ~N 7 ~ . a . ‘ aa ‘ 1 x . ° . . ' e - ‘ - . t : - . - eo. - . . -~ o . . : . . . . . ‘ ‘ - - ’ . e : * . ~ On account of the small current hancled gveas di??iculty was experienced in operating the oscillogrash, and no oscillograms are shown because they were not satis2aotery fot photogranhing. The osaillogranh however, did indicate that a rectification was taking »lacee The direct ourrent produced was not constant in value and had superimposed upon it @ high frequency A&C. of the order of 1500 cycles. This we attribute to the fact that there is a Svark gay in the circuit. The oscillogravh element war finally connected directly in tho ahorted high voltage D.C. circuit, although a} first a ahunt circuit was used to pre- vent injuring the oscillograph. Great difficulty was experienced in operating the oscillograph due to the extremely high voltages handled. Rotation of the brushes changes the wave form and polarity of tye D.C. The best method of obtaining the brush setting for maximum output is to rotate them until a maximum spark length is obtained in a gap connected in the D.C. circuit. ~ ey) a . - = we af . ” \ Ne Nee a o.: ry ““ - wr & Noe . Tf ~ . , ro ¢ r 8 . * * ‘ wt L ‘ “TL Roel ol oO J ees - ‘ . - ~ e -—-— @« 7) rite ¥ : ee Fr - C\ - fr. : c rd ' a ( ° - ae eed - ~a - ‘ - - eo the @ ’ we ° ‘ Ms LT .? 7 4. 1 . i'@o- -¢ ” . ' . + ° o - oe ee ww. \y iG wee . ik fae rose . ot, om e e e » Cd ar a aya do BO coe , ‘ . . a? ~. a oy a fe Fm e ‘ ~ - - 6 , ; ’ . “c . ' . ~erer — a o- c mo . e . i} , - > - ee ik - e aces . ! ' . ~ ‘ “ # « ™ ter - - ° Y o c gsur ~ . ; - : ‘ ! . _ os ake Y. A 4 . i! : tee ved, wa’ J — wea et ») - ~~. AUN ~* nN op Ff . a , ? . rf Ngee Yow J ' we, CG " { ‘ { j e tee te Ww ~i Xe oe = artt oe a i e e\e.t ° s , te. a r - a s a ° a hen + boy ~ nw or , "ON Ue r cele sD w aw , ‘ . ae _ a — ~ s s oi! e@ ‘ ag -. i - «-« t r “rr n ’ x") “7 ee . Hoan 7 a 3 - act °) ‘ - - . - ‘os. mL ~ « ~ yw t ‘4 Cd x 74 1 t oO e @ ue \ we io ‘ ‘ *". oN fy c ae NS 4 rf { oe j re e) J. “1 Bs qf om Cis ‘a Shea et reo¢ be Srypes om fbn « oe mn wo ade onche Tang os PF ory chr cai yacfho o« to FE Ove Tren es aoftfosge. eg ante ta peagtelinisgner eff t sradg wo13 Pay -vissad ed tect: FE .vtotam sroanmorconce £fse2 Bowe + rr PAR + yerer rome sect -} r arm att or c rf 3 .OUTCENGS OF SYnrgrr ear Ae S9toanag sor Ss.d8 r “4 iy * v er r 7 ’ . - - 6 -r Ors: @Q. Le oe. -h eee “3 edis ta TOWOL WG 4 af I woe BG i an a. OF 1d us ature S 1p, ooracesk tot ,rbarecuo wrotdsd «tf bean es dora ‘ =e 4 e og 2 es nt rr wy > . "? e - ry T°, rm PP CD rr sy a) re > Fae c “~~ ta A eOO1i 2 EL " rey VOSS . 3 ZOLTBS YHTO OTIS ste A SiLT COTS ta? oe Roo! USE ONLY. | . . . ? ‘ t a Cooder tee Ree . ear errs soe 5 aes - aa Tr ie ; a — : 7 on . Pon eee hd a nee Sa Je A Pee SP . ee te thee eee . rt OD oe ee Pa Porerwrs = = er enye wera — ° Te mm a Le mee res : a i a woe ee eT ee eee are ee ae ee ae 3S ae er eee: ep et pe : - bel Hevatbtaier dp eaietnatenae ail 7 - 4 ene en ah nats ¢ - . . —_ _—? . ae rere ean eiats «= Too o Tbe ee TE 8 ee pe 74 : see footee a + ows Se LID OD. Ses 7 * rn Shae geet ne eo an Sat crake duane ae oe ee rte Pee eS mee ef . -- 3 Al 03057 ty