Version 4.16, September 28, 2009 Errata file for the FIRST EDITION (2000) of "Building Electro-Optical Systems: Making It All Work" by Philip C. D. Hobbs. Corrections and clarifications listed here will be included in subsequent printings. Anything too big to fit on the page will have to wait for another edition. Besides these errata, there are some corrected figures at ftp://ftp.wiley.com/public/sci_tech_med/electro-optical/fixedfig.pdf Errors are listed in sections corresponding to the LAST printing they appear in, so if you have the first printing, all these errata apply, whereas in the second printing, only those in the second and subsequent sections apply, and so forth. I have examined the first, second, and fourth printings, but a few errors I thought had been corrected in the second printing reappeared in the fourth, so it may be worth checking if you have a second or third printing copy. NB E-books: *All E-book editions are equivalent to the first printing*. I will gladly acknowledge first finders of new errors by name. No error is too small to be worth fixing. Send errata and queries to the author at hobbs@alumni.stanford.org. Updated versions of these files and other supplementary material can be found at ftp://ftp.wiley.com/public/sci_tech_med/electro-optical or http://www.electrooptical.net Thanks for your help. Phil Hobbs ElectroOptical Innovations Consulting 55 Orchard Rd Briarcliff Manor, NY 10510 hobbs@electrooptical.net hobbs@alumni.stanford.org NB: Actual changes to the text and equations are in double quotes--things in square brackets are comments or attributions and are not part of the emendations. Errors in First Printing, May 2000, and All E-Books *************************************************** Section 1.3.3, P. 13 top of third paragraph: "...Gaussian cutoff in u and v, but for small z,...." Equation 1.39, p. 19 delta(f), not delta(x) Section 1.4 p. 30 end of first paragraph: "...instead, and if obliquity is significant, we have to multiply by a cosine." Section 1.6.2, P. 34 bottom of second paragraph: "...forward power is T sub FP/(1-R)*P sub inc." Section 2.5.3, P. 58 Second paragraph, third sentence: "Its spatial coherence is...." (thanks to Greg Magel) Section 2.7 p. 63 second paragraph, third sentence: "Striking voltage depends on pressure, rising at high vacuum and high pressure; the minimum occurs around 1 torr." Section 2.13.6 p. 77 second paragraph, second sentence: "...from the output coupler to change." Example 4.2 p. 149 last sentence: "...equiconvex lens (f/2 for a planoconvex)." Section 5.1.1 p. 160 second paragraph, second sentence: "...between s and p, so keep careful...." [no superscript on p] Figure 5.6, P. 169 Internal and external AL should be interchanged, and p and s should be interchanged (the new figure is in FIXEDFIG.PDF). Figure 5.9, P. 172 Internal and external AL should be interchanged, and p and s should be interchanged (the new figure is in FIXEDFIG.PDF). Section 5.6.12, P. 182 second sentence: "Make the channels narrow enough...." Section 5.6.10 p. 182 third sentence: "...at least three bounces from shiny black surfaces...." Section 6.2.4 p. 188 first sentence: "...orthogonal when their eccentricities are equal, their helicities are opposite,...." Figure 6.4 p. 199 caption: "...(a) simple calcite prism (ne < no) (b) quartz Glan-Taylor (ne > no); (c) Calcite Glan-Thompson." Equation 7.12 p. 220 in second term, should be cap M in numerator Figure 7.6, P. 221 Interchange S and P curves (the new figure is in FIXEDFIG.PDF). Section 7.6.2, p. 221 second sentence: "...along the wires (s-polarized) causes strong...." Section 7.6.2, p. 221 second-to-last sentence: "...across the grooves (p-polarized) causes...." [Thanks to Qin Chen of RPI for spotting this and the previous one] Figure 7.6 p. 221 add at end of caption: "The Richardson book has p and s interchanged." Section 7.6.3, P. 222 top line: "...kgrat is the base of the isosceles triangle...." Section 7.9.1, P. 225 fifth sentence: "...0.05% gain / degree C, with some...." Section 7.9.4, P. 227 first paragraph, second-to-last sentence: "...90 mm across)." Section 7.9.7, P. 228 second sentence of second paragraph: "...where you're sweeping a focused beam back and forth...." Table 7.1 p. 229 caption: "Approximating a Constant Linear Velocity Scan with a Sinusoidal Galvo" Table 8.1, P. 240 multimode radius: 25-900um Section 8.3.3, P. 247 top line: "...(e.g. A' is the principal...." Section 8.3.3, P. 247 footnote: "Their difference in propagation...." Section 8.3.3, P. 247 fourth sentence: "...and A'(0) = (beta sub 1 / sqrt 2, beta sub 2 / sqrt 2) sup T." Section 8.5.3, P. 254 second paragraph, fourth sentence: "...a He-Ne beam through a fiber...." Section 8.5.9, P. 256 second sentence: "as a very high negative power...." Section 9.2.2, P. 288 last sentence: "...source-free region (places where (9.4) predicts wavefronts crossing each other are called caustics. They require special treatment)." [Thanks to Douglas S. Goodman.] Section 9.2.3, P. 289 first and second sentences: "...in the exponent instead,..." [equation] "Since the ray path depends on grad n, we don't know exactly...." Section 9.2.4, P. 290 last line: "...a low dn/dT and a...." Section 9.2.7, P. 292 second sentence: "...and has little or no coma." [Douglas S. Goodman] Section 9.3.6, P. 296 second sentence: "Consider the exit pupil plane of an optical system, having a mildly...." Section 9.3.6, P. 296 fourth sentence: "...is assumed to be small compared to |x - x'|)." Example 9.5 P. 307, third sentence and equation 9.44: The detector's amplitude PSF (before spatial sampling) is rect(x/delta) rect (y/delta), so since u and x/lambda are the conjugate variables, the detector CTF is the product of two sinc functions scaled by delta/lambda. Squaring the rectangular amplitude PSF to get the intensity PSF doesn't change its functional form, so the CTF is equal to the OTF, and we get 2 delta OTF(u,v) = CTF(u,v) = ------- sinc(u delta/lambda) sinc(v delta/lambda) 2 lambda (9.44) Section 9.4.2 P. 310 first sentence: "...1.517), we get the result shown in Figure 9.11." Section 9.5.4 P. 314 second-to-last sentence: "...as equivalent width; if the peak...." Section 9.8.2, P. 319 last sentence: "...illuminated by a cone of light, which is plotted...." Section 10.3.4 P. 329 fourth sentence: "...from Section 1.5.2: in an ac measurement, One coherently...." Section 10.8.6, p. 347 last sentence of first paragraph: "...a Wollaston prism, one polarizer, a sample cell...." Table 11.2, P. 368 second from bottom: "Life proportional to exp(+10500/T)" Figure 12.2, P. 389 Inside and outside focus are reversed (the new figure is in FIXEDFIG.PDF). Section 13.8, P. 432 first line: "...and highpass, which reject everything below fc." Section 13.8.9, P. 439 first sentence of second paragraph: "As shown in Figure 13.11, a narrow filter applied to a pulsed waveform...." Figure 13.11 P. 445 Add to end of caption: "Strobing is done by multiplying by the input pulse shape (Gaussian here)." Section 14.3.10 P. 462 third sentence: "... in series with L); its Q is defined slightly differently, as the ratio of the 3-dB...." Section 14.4.1 P. 465 second footnote: "...so our time dependence is e to the j omega t...." Section 15.3 P. 506 last line: "...the signal power is proportional to ." Section 15.3. P. 507 first line: "...proportional to . Here's...." Equation 15.4 P. 509 "A sub VCL- =..." [subscript is "VCL minus"] Equation 15.5 P. 509 "partial V sub IN- over partial V sub O =..." [numerator subscript is "IN minus"] Section 15.4.4 P. 512 sixth sentence: "...goes as Q dot / (M sub TH f),...." Section 15.4.6 P. 513 first paragraph, add at end: "Rail-to-rail outputs have very much higher R sub O." Section 15.10 P. 535 last line: "...to do this (see Sections 10.8.6 and 18.5.3)." Section 16.4.2, P. 552 third sentence of second paragraph: "ground plane, which is one source of trouble, and the drop between shield and center conductor caused by shorting out the shield loop is the other. Fortunately, it isn't...." Section 17.4.1 P. 599 second-to-last sentence: "...unequally spaced, or have gaps in them,...." Figure 17.13, P. 600 caption: "at f = 0.03 nu and 0.2 V at f = 0.04 nu, with a 100 muV rms noise floor, for two different starting phases and 60 and 122 poles: (a), detail of peaks; (b), full range." Section 18.4 P. 627 just above Eq. 18.10: "...and the voltage- gain bandwidth of the resulting...." Section 18.4.11 P. 644 first paragraph, second-to-last sentence: "...goes as I sub C but that...." Example 18.3 P. 650 last sentence: Add footnote to "generality": "H. W. Bode, Network Analysis and Feedback Amplifier Design, Van Nostrand, New York, 1945, sec.16.3." Section 18.5.3 P. 653 last paragraph, second to last sentence: "For a fixed value of Delta V sub BE,...." Errors In Second Printing, August 2001 ************************************** (The second printing errata didn't get corrected in the third printing, unfortunately.) Errors In Third Printing, February 2004, And Previous Printings *************************************************************** Section 1.1, P. 1, footnote: "...ftp://ftp.wiley.com/public/sci_tech_med/electro-optical." Equation 1.24 P. 12 first term inside square brackets at right should be -1/w(z)**2 (change sign) Figure 1.5 P. 17 For "marginal ray" read "rim ray". [Thanks to Dan Marshall.] Equation 1.76 p. 35 iNshot = sqrt(2 eta e^2 (Nref+Nsig)) [e squared not just e. Thanks to Daniel Cote of the Ontario Cancer Institute] Figure 1.9(b) p. 40 Image is upside-down. Equation 4.9 P. 146 "delta = -2 arctan(cos theta sub i sqrt[sin^2 theta i -(n2/n1)^2]/sin^2 theta i)" [move the factor of 2 in the numerator outside the arctan] Section 7.8 P. 224 third through fifth sentences sentences and Equation 7.16: "...pretty good retroreflector if its index is close to 2.0: for a single interface, a ray parallel to the axis at height h has sin theta sub i = h/R. Retroreflection requires the light to be focused on the opposite surface, so sin theta sub r = h/ 2R, and by Snell's law, n sub 2 = n sub 1 (h / R)/(h/ 2R) = 2 n sub 1 (7.16) If n sub 1 = 1, then n sub 2 must be 2, which can be done in glass. To prevent most of the light...." Table 8.1 P. 240 "Normalized Frequency V = a*n_1*k_0*sqrt(2 Delta) 1.8-2.4 (single mode)" [Thanks to Peter Sellin of Montana State University] Section 8.4.2 P. 249 First sentence: "...on the order of 200 for a step-index multimode fibre and 100 for a graded-index fibre. For a given value of V, graded-index fibre will have half as many modes as step-index. (Why?)" Figure 10.2 P. 330 Reverse direction of analyzer rotation at bottom, replace alphas with proportional signs (updated figure is in FIXEDFIG.PDF). Equation 10.6, p. 350 derivative in second sum should be d sup j P by d i sup j Equation 13.14, p. 418 "...approximately Pout (ln 10)/(10 Pin)" [delete "1-" in approximation] Section 13.5.3, p. 418 sentence containing Eq. 13.14: "Below P sub 1dB, where no serious waveform distortion occurs, the compression error delta is [equation] and the compression factors (1-delta) of cascaded devices multiply." Section 14.3.10 P. 462 third sentence: "... in series with L); its Q is defined slightly differently, as the ratio of the 3-dB...." Section 15.7.6 p. 528 Rule 1 should read, "1. Choose a lowpass filter with the same full bandwidth you need." Table 18.4 p. 643 The first two devices, BFG25A/X and BFG505X, have had their parameters shifted right by one column, beginning in column 4 (specified I_c for f_T). Type BFG25A/X has an f_T of 5 GHz at I_c=1 mA, beta of 50 at I_c=0.5 mA, and no specified value for R_E'. Type BFG505X has an f_T of 9 GHz at I_c=5 mA, beta of 60 at I_c=5 mA, and no specified value for R_E'. (Thanks to Hansjuerg Schmutz of ETH for this one and the next four). Section 18.4.14 p. 647 Read L for L_1 and R for R_L throughout. Equation 18.26 p. 647 First denominator term should be (1-omega^2/omega_0^2). Section 18.4.14 p. 648 Fifth paragraph: reference to (455) should be (18.30). Example 18.2 p. 649 Fourth line: "...perhaps using a 2:1 transformer (4:1 in impedance, see Section 18.6.3) between the..." and halfway down the paragraph: "...or about 20 MHz, the same bandwidth as the equivalent lowpass...thus I_d R_in only needs to be 13 mV for the shot noise...." Section 18.4.15 p. 649 just below Eq. 18.31: "...becomes a bandpass of 40 MHz full width...." Errors In Fourth Printing, 2005 And All Previous And Subsequent Printings ************************************************************************* (Some of these are old ones that didn't get included in earlier printings.) Inside front cover 3rd line from bottom: "...3*10**-5 to 10**-3 (step-index MM)" (Thanks to Gary Woods of Credence Systems Corp.) Figure 1.2 p. 10 Revised figure is in FIXEDFIG.PDF. Caption should read, "Scalar addition is a reasonable approximation to vector addition except near high-NA foci and caustics. Near focus, there are longitudinal field components and a reduced phase derivative with respect to z (|Delta phi| < |k Delta z|)." Section 1.3.11 p. 28 Equation 1.62 and following: "E = n**2 A Omega prime where A is the clear area, n is the refractive index of the medium in which the projected solid angle Omega prime is measured." Footnote: (Solid angle is the integral(0,theta){sin u du}, while projected solid angle takes account of the cosine obliquity factor [see Section 9.3.4]). Thus projected solid angle has a sin**2 in the integral instead.) (Thanks to Ray Delcher of Teledyne Scientific.) Example 1.8 p. 28 fourth line, "...= 5*10^-9 cm^2*sr. If we...." (Thanks to Gary Woods of Credence Systems Corp.) Section 1.3.11 p. 28 last sentence of first paragraph: "The *coherence area* of an optical field gives an idea of how far apart the slits can be and still produce fringes." Equation 1.73 P. 34 Denominator: "1+[4R(1-R)**2] sin**2(...)" (move the square outside the parentheses) [Thanks to Bruce Tiemann.] Equations 1.87 & 1.89 P. 45 Lambdas should be in the numerator, not the denominator. (Thanks to Jesse Shaver of Vanderbilt University.) Section 2.8.2 P. 64 Last sentence of first paragraph: "... (about $10**-23 each),...." [Thanks to Dan Marshall.] Example 4.2 p. 149 last sentence: "...equiconvex lens (f/2 for a planoconvex)." Section 7.8 P. 224 third through fifth sentences sentences and Equation 7.16: "...pretty good retroreflector if its index is close to 2.0: for a single interface, a ray parallel to the axis at height h has sin theta sub i = h/R. Retroreflection requires the light to be focused on the opposite surface, so sin theta sub r = h/ 2R, and by Snell's law, n sub 2 = n sub 1 (h / R)/(h/ 2R) = 2 n sub 1 (7.16) If n sub 1 = 1, then n sub 2 must be 2, which can be done in glass. To prevent most of the light...." Section 7.9.1, P. 225 fifth sentence: "...0.05% gain / degree C, with some...." Section 7.9.4, P. 227 first paragraph, second-to-last sentence: "...90 mm across)." Section 11.5, P. 367 first sentence of third paragraph: "...where rho=(u^2+v^2)^1/2,...." Figure 12.5, P. 397 Op amp should be LT1013 (any low offset, low bias current op amp will work fine). Choose galvanometers and series resistors such that the full-scale op amp output is about 105% of full scale on the meters. (Thanks to Guenter Hoffman of TUE.) Equation 13.14, p. 418 "...delta = 1 - P_out/(GP_in) approximately Pout (ln 10)/(10 Pin)" [correction done incorrectly in 4th printing] Section 13.5.3, p. 418 sentence containing Eq. 13.14: "Below P sub 1dB, where no serious waveform distortion occurs, the compression error delta is [equation] and the compression factors (1-delta) of cascaded devices multiply." [correction not quite right in 4th printing] Figure 14.4 P. 462 Caption should read, "In the high-Q limit (Q > 5), near resonance, an L-network multiplies R_s by Q^2+1 to get equivalent R_p. The three forms have very different skirt selectivity." Equation 15.5 P. 509 "partial V sub IN- over partial V sub O =..." [numerator subscript is "IN minus"] Section 15.4.3 p. 511 fifth sentence: "... the zeros of the transfer function are at zeros of Z_F and poles of Z_I, while its poles are at poles of Z_F and zeros of Z_I." (Thanks to Jorge Alcoz.) Section 15.7.8 Delete 2 pi from equation (Thanks to George Herold for this one and the next.) Section 16.3.4 and Skin depth of copper at 60 Hz is 0.85 cm, not Section 16.5.3 0.85 m. Section 18.4 P. 627 just above Eq. 18.10: "...and the voltage-gain bandwidth of the resulting...." Figure 18.14 P. 639 Caption should read, "Final circuit: cascode Q_1 plus bootstrap Q_2 cope with the obese 100 pF diode; Q_3 corrects drift due to V_BE of Q_1 (300 mu V/ degree C at the output). Section 18.7 p. 665 Item 9, second line: "...multipliers in Example 14.1; ...." (Thanks to John Cole of Draeger PMLS Ltd.) ******** END ********