Using Stock Optics. ECE 5616 Curtis

Size: px

Start display at page:

Download "Using Stock Optics. ECE 5616 Curtis"

Oswald Derick Wheeler
6 years ago
Views:

1 Using Stock Optics What shape to use X & Y parameters Please use achromatics Please use camera lens Please use 4F imaging systems Others things Data link

2 Stock Optics Some comments Advantages Time and cost Can still choose custom coatings (typically) If vendor is good can eventually transfer quality measurement to vendor after process is established. ALL optics have to be inspected before being used for quite some time. Disadvantages Limited performance choices in EFL, lens diameter, and function. Most lenses are design for infinite conjugates and visible wavelengths Part can be obsolete even if you are buying it. Must confirm supply if used in production and guarantee transition time if part is to be discontinued. Vendor can go out of business. Vendor quality can vary dramatically with time as they change suppliers. Even the most simple things can be messed up. The vendor s ability for SQE can vary a lot.

3 What shape lens should I use? Aberrations using Thin Lenses Shape Parameter X

4 Aberrations using Thin Lenses

5 Spherical Aberration for a thin lens

6 Thin Lenses approximation

7 Summary Chart from Smith Note direction of flat side of plano convex lens toward the focus. This is the correct way to use it.

8 When ever possible use camera lens Nikon AF Micro-Nikkor 105mm f/2.8 Wavelength range is ~450 to 680nm Use in 4F system for imaging, collimators, or imaging at distance

9 Camera Lenses in 4F Object Image Put film side of lens (BFP) toward image/object and set focus in infinity Alignment back propagate Plane Wave to put focus on Image/Object and set distance between lenses with a PW though both lens and a shear plate. Most camera lenses have FFP INSIDE lens if EFL is less than ~60mm Magnification use different EFL for first and second lens.

10 Stock Lenses Use 4F system for imaging Two achromatic doublets corrected for infinite conjugate Narrow field flatter surface towards image/object Wider fields reverse orientation works better Limited Field with really good performance Ratio of Focal length is magnification Why 4F systems??? - symmetrical systems about the stop (between the lens) then the system is free of coma, distortion, and lateral color. Also 4F is telecentric and is insensitive to image/object position.

11 Stock Lens Field Lens Use a field lens to flatten the field (Eliminates Petzval Field Curvature with little effect on other aberrations). Use a negative lens to flatten field resulting from positive elements.

12 Stock Lenses Splitting elements Reduce SA by factor of 5 by using two lenses together

13 Stock Lenses Placement of Stop

14 Stock Lenses Higher NA alignment of lenses

15 Stock Lenses General considerations Please use achromatic doublets. Much better SA at F#2 can get diffraction limited focus from achromat but not PC lens. Accuracy of EFL is typically better for achromatic doublet lens than singlet. Remember to get correct AR coating. Pick correct material for lens depending on wavelength (BK-7 or fused Si are typical choices) Look at scratch and dig (quality) and surface quality (λ/8 or better for laser apps) needed for your application. Can measure actual lens performance and thickness and re-optimize mechanics Scratch-Dig Cost Low Moderate High Applications Used for low power laser and imaging applications where scattered light is not as critical as cost Excellent for laser and imaging systems with focused beams that can tolerate little scattered light For demanding laser and imaging systems where minimizing scattered light is critical

16 Stock Lenses Spatial Filtering use microscope objective When a positive lens of focal length F focuses a Gaussian beam, the image at the focal plane (the Optical Power Spectrum, or OPS) will be an inverted map of spatial wavelengths present in the beam. Short wavelength noise (d n ) will appear in an annulus of radius Fλ/d n centered on the optic axis. The long spatial wavelength of an ideal Gaussian profile will form an image directly on the optic axis. A pinhole centered on the axis can block the unwanted noise annulus while passing most of the laser s energy. The fraction of power passed by a pinhole of diameter D is: This passes 99.3% of the total beam energy and blocks spatial wavelengths smaller than 2a, the diameter of the initial beam. Since d n is always much smaller than the beam diameter, the filtered beam is very close to the ideal profile. and the minimum noise wavelength transmitted by the pinhole is A pinhole of diameter D opt :

17 Stock Optics Beam splitters or plates Place in collimated beam if possible. Even tilted plate in collimated beam will not introduce astigmatism. If not in collimated space place close to image/object plane (like field lens) to have minimal impact. Remember to align lenses with the BS or plate in the system to adjust spacing.

18 ZEMAX Click on Len to get

19 ZEMAX Click on the down arrow to select a vendor of you choice

20 ZEMAX Vendor Melles Griot EFL between 80 to 100 mm Diameter 15 to 25 mm Lens can only be doublet and have all the type of shapes selected

21 ZEMAX Click on Insert to place doublet lens

22 ZEMAX Need to enter EPD using Gen Button and use the marginal ray solve to find the paraxial focus

23 Data Link Design As an example of the use of some of the basic concepts, lets consider the problem of designing an optical communication link using an optical fiber for transmission of binary digital data over a distance of 5km. The system requirements are 1) Data rate 5x10 8 bits/s 2) Error probability after amplification at the receive must be less than ) Parameters: Loss is 4dB/km, λ=1.04um, total capacitance of diode (junction plus package) is 3x10-12 f, noise figure of amplifier is 6dB, QE of detector is 50%. System is shown below. How much power do we need? Amnon Yariv, Optical Electronics

24 Detecting Pulses Threshold Detection Noiseless data Detected data with threshold Data stream after detection i n < -i s (1-k), 1 turns into 0 i n > i s k, 0 turns into 1

25 Detecting Pulses Threshold Detection Let the noise current be random Gaussian, then σ is the RMS of the noise current, σ 2 = i n2. Setting k=1/2 give probability of error P e of Using Gives P e

26 Detecting Pulses Threshold Detection

27 Data Transmission Power Needed Dominant noise sources are amplifier and Johnson noise in resistor, so Signal current is given by P s is the peak pulse power incident on detector i i s s SNR after amp is We need this to be >11.89 N Peη / hν = (4kT Δν / R ) e L 1/ 2 So we need to know T e, Δν, and R L to find needed P s and then cascade that power back to the diode to find required diode output power.

28 Data Transmission Power Needed T e is obtained from the amplifiers noise figure. T F T e T 1 o e = (4-1)290 = Lecture 21 K Bandwidth of signal is given as 5x10 8 pulses per second (binary data). Thus, conservatively Where τ is 2x10-9 s from data rate Δν = 2/( πτ ) Δν = 3.18 x10 8 Hz Now the R L must not exceed the value below to support this rate 1 R L = 2πΔυC So using the given value of C, R L < 167 ohms

29 Data Transmission Power Needed Plug in these values into equation below and solve for P s P s 10-5 watts Peη s / hν (4kTeΔν / RL ) = 1 / Now loss to detector include 5km of fiber (20dB) and assume another 4dB coupling loss into the fiber for a total of 24dB (factor of 251). Therefore the required laser power must exceed P laser > 10-5 x 251 = 2.51 mw

30 Reading W. Smith Modern Optical Engineering Chapter 21 See Optical Resources on class website.

Optical System Design

Phys 531 Lecture 12 14 October 2004 Optical System Design Last time: Surveyed examples of optical systems Today, discuss system design Lens design = course of its own (not taught by me!) Try to give some