The Spectrum of a High Frequency Gas Source at Low Pressure Citation Martin, Francis Crawford (1928) The Spectrum of a High Frequency Gas Source at Low Pressure. Master's thesis, California Institute of Technology. doi:10.7907/3KSA-YA86. https://resolver.caltech.edu/CaltechETD:etd-02252005-093711 Abstract No abstract. Item Type: Thesis (Master's thesis) Subject Keywords: Physics Degree Grantor: California Institute of Technology Division: Physics, Mathematics and Astronomy Major Option: Physics Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 1 January 1928 Record Number: CaltechETD:etd-02252005-093711 Persistent URL: https://resolver.caltech.edu/CaltechETD:etd-02252005-093711 DOI: 10.7907/3KSA-YA86 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 757 Collection: CaltechTHESIS Deposited By: Imported from ETD-db Deposited On: 28 Feb 2005 Last Modified: 03 Oct 2019 23:02 Thesis Files Preview PDF (Martin_fc_1928.pdf) - Final Version See Usage Policy. 360kB Repository Staff Only: item control page

THE SPECTRUM OF A HIGH FREQUENCY GAS SOURCE AT LOW PRESSURE THESIS by FG MARTIN IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE CALIFORNIA INSTITUTE OF TECHNOLOGY PASADENA CALIFORNIA 1928 ELECTRICAL We [BELLING DEPT THE SPECTRUM OF A HIGH FREQUENCY “os... GAS SOURCE AT LOW PRESSURE The high frequency radio oscillator had first to be milt. This coneisted of a Hartley circuit using an "R.C.A." U.X. 852 "short wave" tube which has a plate dissipation of 75 watts. The oscillator worked well with a grid leak resistance of 48,000 ohms and a grid leak condenser of 0.25 frst. The filament was lit by an 11 volt 50 ~y circuit with a rheostat to control ite current to 3.QSameres. The plate voltage was 2000 volts 50, with a rheostet in the 110 vwolt primary side of the transformer to | limit the plate voltage to 2000 volts and the plate current to about 120 .milliamps when the set was oscillating. | Dr.P.S.Epstein drew my attention to the following stand- ardization of radio~frequency chokes through Mr.A.P.Alford. The five radio-frequency chokes that were used were standardised to the required choking wave length from the following table taken from a graph in the paper by Leng in the Ann. der Physik: Vol.43, 1914, p.749 on the choking wave length of close wound cylindrical coils, The dielectric may affect these wave lengths somewhat but these coils should all choke more or less together which is an important point in preventing dissipation of the high frequency energy. Je vn 8 = length of cylindrical coil. D" A@iameter of cylindrical coil. L = length of wire wound on coil. A= wave length of fundamental shoked 8 A 3 A D L D L 0.50 2.95 3.0 1.6 0.67 2.655 4.5 1.4 t.0 2.35 6.0 1.2 1.5 2.0 8.0 1.0 2.0 1.855 The radio frequency choke coils for the plate end grid leads were wound of No. 34 S.W.G. silk covered wire but for the filament leads No. 22 S.W.G wire was used because these coils had to carry a current of. 3.5 amps. By varying the positions of the plate, filament and grid leads on the induct— ance coil and the setting of the 35 » condenser it was possible to obtein 2 maximum energy Ye the discharge tube because these adjustments changed the weve length to that of the radio frecuency chokes. Precautions had to be taken to prevent body contact with the 2090 volt circuit end also to first open the plate current 3 cireuit-——- in the pr@mary of its transformer-- before opening the filament switch so as to protect the tube from the induct— ive current surge from the transformer. The discharge tube source of light consisted of a soft glass tube 2.5 oms. diameter and 87 cms. long and this was connected thréugh a mercury trap to a Lengmir two stage marcury pump end fore pump. The glass cut off light of wave lengths less than 3100 i. The high frequency field from the oscillator was applied at two sheets of thin copper foil wound round the discharge tube and separated by about 5 cms. at their inner edges. Thies alternating field causes the electrons and ions to oscillate synchronously:.back and forth in the tube and the consequent collisions with molecules cause the giving out of light. A Ford car ignition coil could be connected to term inals at the ends of thia tube. When the set was oscillating, motion of the hand near the coil altered the plate current, and a screwdriver held by its insulated handle drew sparks at ite tip from either the plate or the grid lead. When a Neon tube,or the wave meter, was brought near the coil it was brilliantly lit. Ali leada of the high frequency oscillating circuits had to be kept as short as possible and as far away as practicable from each other. A slight instability in the light was sometimes not~ iced but this was probably due to the use of alternating cur- rent for the filament and the plate, because if the Ford coil was started while the oscillator was on it stopped the light and this did not appear again 4 until some seconds efter the Pord ceil had been open cir- cuited. The high frequency alternating field gives light at gas pressure of, say, 1079 to lo~+ mim., and these are lower then 1072 to 1072 m.m. at which pressure a Ford car ignition coil ceases to give a discharge, because the moving electrons and fone cen be reversed in motion by the higher frecuency, even at this lower pressure where their mean free path is longer, before they reach the en@®of the tube. After the tube has been runfor some time with mercury vapor it is necessary to run a Ford coil diecharge for about an hour with liquid air on the trap to remove the mercury from the terminel plates and tube. Then to completely suppress the spectral lines of mercury the bottom of the trap mist be immersed in the liquid air and not merely put in the vapor of the liquid air. From experiments carried out by Mr.A.P.Alford it appears that a more intense light source might be obtained by using & mach narrower glass tube, say, only 1 cm. in diameter or lees and by inductively coupling the light circuit to the inductence of the oscillating circuit so as to increase the voltage across the light tube "terminals". 5. Prof.R.¥.Wood of Johns Hopkins University refers to the taking of apectra caused by light from a high frequency electrodeless discharge in a gas tube at a low pressure. See Wood and Loomis: Nature: Vol. 120, Oct.6, 1927, p.510 See also: Kirchner: Ann. der Physik: Vol. 77, Aug. 1925, p.297- J.R.Clarkes Nature: Vol. 120, Nov.19,1927, p.727 Gill and Donaldson:Phil .Mag.: Vol. 2, July 1926, p.129 The following papers describe the High Frequency Low Pressure Lamp: Foulke: S.A.I.E.E.: Vol. 46, Feb. 1927, p.139 J.and W. Taylor: Nature: Vol. 121, Feb. 4,1928, p.171 These papers deal with the problem more or less indirectlys Wood: Phil .Mag. Vol. 42, Aug. 1921, p.729 Wood and Vosat Nature: Vol.121, Mar.17,1928, p.48 Ponte: Nature: Vol.121, Feb. 18,1928, p.243 J.R.Clarkes Nature: Yol.121, Feb.25, 1928, p.282 S.P.MoCallum: Nature! Vol.12%, Mar.10,1928, p.353 Sinelair Smith! Nature: Vol.t21, Jan. 2,1928,p. 91 Emelius® Carmichael :Nature: Vol.121, Jan. 7,1928, p. 14 6 Two spectrographs were used, one having a glass prism for pre~ liminary surveys and the second one having a quartz prism which gives greater dispersion and resolution for the more detailed sur- vey. Each of these was calibrated by identifying the known spectrum of mercury. The usual double "V" slide was used acvoss the slits for the comparison of spectra and by making the edges of these spectra slightly overlap a more sensitive method was available to detect whether any faint line occurs in only one of two spectra being compared because any overlapping will show up as an increased in- tensity, but if there is a space between the edges of spectra it is easier to identify lines when putting one plate on the other. It would be best to make one edge of the central spectrum overlap and the other edge to clear the inner edges of the two outer"half spectra. The results obtained from the mercury spectrum showed that at low pressures the intensity of singleta is relatively stronger than for higher pressures. An appendix will be filed to this report in three weeks giving the spectra of mereury at (1) high frequency at a pressure in the range 10°? to 1077 men. 2 aya (2) Ford coil diecharge at a pressure in range 10 to 107mm. A large number of plates were exposed but only the best and most representative data are tabulated. The writer ie indebted to Dr.R.A.Millikan for his interest and for suggesting this problem and also to Dr.8.S8.Mackeown for con- tinued assistance in the building of the oscillator. Dr.R.D. Benet? and Mr.A.P.Alford also helped with the oscillator. Dr.I.8.Bowen calibrated the speetrographs and identified many of the spectral Lines. Line Spectrum 7 of Mercury. te Excited by a 7.0 Metere High Frequency Oscillator Wave L iO at a Pressure in the Range 10°? to 1074an, ength broad broad #11 ro WN WN Intensities #14 15 7) -_ 2 1 3 2 3 2 2 2 12 12 12 12 2 2 20 20 12 15 1 ‘ 2 47 15 12 10 6 4 8 10 20 12 50. 20 Glasa prism #10 1 2 nm ng ~ VW VW ff Ww ~~ 4 ON ~ ww VW @ Ml A- nw vi Series Designation 7's ~ 8'p 7's - 9'p é'p ~ 6'p é'p — 6D 62,- 778 é'p . a's @P,- 7°75 6’ - 7" é'p - 9's 6P,- 7's Line Svectrum g of Mercury Excited by a 7.0 Meters High Frequency Cscilletor ~- ok at a Pressure in the Range 10°? to 10 ‘nm.n. Wave Length tf 0 3964 5906 3860 3802 3755 * 3704 3663 5655 3650 5993 * 3341 3132 5126 fuzzy 12 12 15 14 wl | Ro nan 12 12 Intensities 15 13 25 Glass prism 6 Glass prian cute off further lines. Series Designation 1 é'p . a'p é'p -10's 6'p — otp 6 Po~ 6D, 3 3 6 Po- 6 D, 3. 6°P,~ ©, " 6'P ~10'D 6 Po- 8S 5) 6p, 670, OP, OD, * Nitrogen bands have approximately the same wave lengths as these. I Band Spectrun of Nitrogen Excited at a Pressure in the Range 1072 to 107*n.m. Wave Length Character of Line Intensities Ford Coil H.F.7 Meters A pte Be 6752 wide fuzzy $ 5 6682 n " + 4 6604 " " 4 I 6524 " n 3 5 6448 " " 2 5 6575 " " 1 u 6048 " " 1 - 5994 n " 1 - 5943 " " 1 - 5891 ' " ; - 5839 , " ; - 5137 " ® x - Agta band } - 4723 " i - 4675 " 5 4 4574 " 3 1 * 4489 " 3 1 AAG " 2 ’ 4276 " k 3 4207 " 4 3 44a n 5 34>. é ‘Oo = ~~ Ol — /é Band Spectrum of Nitrogen Excited at a Pressure in the Renge 107° to 107mm. Wave Length Character of Line Intensities Ford Coil H.F.7 Metera ° A Ae #He 4064 band 6 4 3995 . 8 4 3940 " 4 3 3919 " 3 2 3894 n ' 4 3790 " 15 12 3771 " 15 12 * 3704 " 8 5 3671 " 5 5 36h2 . — 4 3644 " s u 3577 " 20 20 3536 " 15 12 3500 " 8 4 3571 " 20 15 3539 " 2 ; 3309 a 1 $ 3285 " 2 a 3268 " i t 3159 " 6 2 3136 " 1 1 2 * Mereury lines have approximately the sama wave lengths as these. 4 Band Spectrum of Nitrogen Excited by a 7.0 Meters High Frequency Oscillator at a Pressure in the Range 10°? to 107#m.m, Wave Length Character of Line Intensities Glaes Prism A #13 #10 6850 wide fuzzy ? . 6752 " ' 2 3 6682 " n 4 6 6604 " e 6 7 6524 a m 8 7 6448 " " 6 7 6375 " " 6 7 6306 " " 5 6 * 6035 " ' 4 4 6171 " " 2 3 6108 " n - 3 6048 " " m 6 5994 " " 4 «6 2943 " " 4 6 5891 " " 4 6 3839 - " " 4 6 2190 5 z to 6 S737 " " 4 5 5393 bend - 4 5366 " - 4 48t4 " 4 PY 4723 f 4 3 4666 a 5 3 /2 Band Spectrum of Nitrogen | Excited by a 7.0 Metere High Frequency Oscillator at a Pressure in the Range to? to 1 on m Wave Length Character of Line Intensities . Glass Prisn A #13 #10 4648 band. . A 4654 " é - A57A " 8 4 4A89 " 8 5 4A16 " 6 + * 4356 ft 20 4 * AAA i 12 4 * 4267 " 12 5 At4e " 12 ; * 4057 " 12 4 5995 . - 7 3940 " 20 7 * 5919 " 12 - 3894 " 6 4 3857 " - 3 5805 " - 4 3771 " - 20 5755 " - 2 3671 " 12 Further lines 5589 " 4 cut off 5OTT " 12 by the 2200 " 8 Glase 3468 . 1 Prism 3 Band Spectrun of Nitrogen Excited by a 7.0 Meters High Frequency Oscillator at a Pressure in the Range 107 to 107i. Wave Length Character of Line Intensities Glass Prism a #13 #10 3371 band 20 Further 5339 . 6 lines 3309 " 6 cut off 3285 " 6 by the 5268 " 6 Glass Prism There were eight faint unjdentified broad bands of wave lengths from 32453 to 3175 A. 3168 band 4 5159 " 8 , 1 3117 4 * Mercury lines have approximately the same wave lengths as these. (see Lastz pages ) DIAGRAM of CIRCUIT | wid (fem GL E O25 Mf : and | 4.000 w i [1 V. TRANSER 50~n RCA. BLOCKING Light UX. 852 i t 3 2,200 V. TRANSFORMER 507 Switch and Control Resistance in the Prmary . a IS Mr.J.i.Peearson cuggests thet the following circuit would eliminate the need of "chokes".