﻿144 
  Dr. 
  Drysdale 
  on 
  the 
  Use 
  of 
  Shunts 
  and 
  Transformers 
  

  

  This 
  implies 
  that 
  the 
  secondary 
  current 
  should 
  be 
  in 
  

   quadrature 
  with 
  the 
  no 
  load 
  current, 
  as 
  is 
  geometrically 
  

   obvious. 
  

  

  To 
  test 
  the 
  constancy 
  of 
  transformation 
  we 
  have 
  

  

  AR 
  c 
  =R,|A(^cos^+A^^sin^|, 
  . 
  (15) 
  

  

  which 
  shows 
  that 
  for 
  constant 
  ratio 
  of 
  transformation 
  both 
  

   A 
  c 
  and 
  A 
  m 
  must 
  be 
  proportional 
  to 
  A 
  s 
  or 
  that 
  

  

  tan<fc= 
  )f^ 
  (10) 
  

  

  For 
  constant 
  secondary 
  impedance 
  A 
  s 
  is 
  proportional 
  to 
  

   the 
  core 
  E.M.F. 
  and 
  hence 
  to 
  the 
  induction 
  density 
  in 
  the 
  

   core, 
  B. 
  Hence 
  the 
  current 
  vector 
  should 
  be 
  perpendicular 
  

  

  to 
  the 
  curve 
  of 
  which 
  the 
  ordinate 
  is 
  -~-, 
  representing 
  the 
  

  

  reluctance 
  of 
  the 
  circuit, 
  and 
  the 
  abscissa 
  is 
  -~ 
  or 
  the 
  ratio 
  of 
  

  

  A 
  s 
  

  

  the 
  core-loss 
  current 
  to 
  B. 
  It 
  is 
  obvious 
  that 
  this 
  is 
  when 
  

  

  the 
  secondary 
  load 
  is 
  a 
  leading 
  one, 
  and 
  this 
  result 
  is 
  contrary 
  

  

  to 
  the 
  frequently 
  quoted 
  experiments 
  of 
  ]\Jr. 
  Campbell 
  *, 
  who 
  

  

  is 
  claimed 
  to 
  have 
  found 
  experimentally 
  that 
  better 
  constancy 
  

  

  of 
  ratio 
  is 
  obtained 
  with 
  an 
  inductive 
  instrument. 
  In 
  the 
  

  

  case, 
  however, 
  where 
  there 
  is 
  an 
  air-gap 
  in 
  the 
  transformer, 
  

  

  A 
  A 
  

  

  the 
  curve 
  connecting 
  —■ 
  with 
  -^ 
  may 
  be 
  practically 
  hori- 
  

  

  zontul, 
  and 
  in 
  this 
  case 
  the 
  regulation 
  will 
  be 
  worst 
  on 
  a 
  non- 
  

   inductive 
  instrument, 
  and 
  better 
  the 
  greater 
  is 
  either 
  the 
  lag 
  

   or 
  lead 
  of 
  the 
  secondary 
  current, 
  In 
  estimating 
  the 
  value 
  of 
  

   Mr. 
  Campbell's 
  experiments 
  it 
  must 
  not 
  be 
  forgotten 
  that 
  the 
  

   two 
  Kelvin 
  balances 
  employed 
  by 
  him 
  must 
  have 
  been 
  of 
  

   very 
  different 
  impedance 
  and 
  consequently 
  tin 
  4 
  core 
  inductions 
  

   would 
  have 
  been 
  decidedly 
  different. 
  This 
  of 
  itself 
  would 
  

   probably 
  have 
  been 
  sufficient 
  to 
  produce 
  the 
  effect 
  found. 
  

   Experimental 
  evidence 
  will 
  be 
  found 
  below 
  on 
  this 
  point. 
  

  

  Effect 
  of 
  variation 
  of 
  frequency, 
  — 
  Increase 
  of 
  frequency 
  

   increases 
  the 
  impedance 
  of 
  the 
  instrument 
  it' 
  inductive, 
  but 
  

   at 
  the 
  same 
  time 
  increases 
  the 
  E.M.F. 
  for 
  the 
  same 
  core 
  

  

  induction. 
  With 
  a 
  fully 
  inductive 
  instrument, 
  therefore, 
  and 
  

  

  if 
  eddy 
  current- 
  are 
  absent, 
  variation- 
  in 
  frequency 
  should 
  

  

  * 
  Phil. 
  Blag. 
  vol. 
  \lii. 
  ]>. 
  271 
  : 
  Journal 
  I. 
  E. 
  E. 
  vol. 
  xxxiii. 
  \\ 
  054. 
  

  

  