Spectrophotometer 01 Design Description Document University of Rochester

Optical Design 1

Overview:

A single diffraction grating of specified lines/mm will be illuminated with a ray angle such that there is a specific illuminated area. This light from the 1^st order will then be redirected and focused into the camera. Additionally, the light from the 0^th order will be partially allowed to hit the camera system, allowing the target to be capture as well.

Slit

Cellphone

Camera
Figure a – Optical Design 1

Cellphone

Camera

Slit

1^st order image

0^th order image
Figure –Optical Design 1b- Shows another optical design with order 0 being visible as well as using a prism.



Figure – Shows the result on the detector with 2 visible orders. For order 0 the object of interest (OOI) can be visible as well as its spectrum. This will be convenient in making sure that the OOI is being measured.

From the iPhone5 specification¹ F/2.4 | Focal length= 4.10mm

According to the equation for imaging system



=3.427mm

According to the equation for grating angle

We take m=1 since we use the first order of diffraction

In the experiment we conducted, the wavelength range for iPhone 5 is 402.4nm to 688.54 nm

So

So

Solve for =0.5966rad;

Resolution Validation: CCD for iPhone camera=3264x2448pixel with 1.5um/pixel

Product: http://www.thorlabs.com/thorproduct.cfm?partnumber=GH13-12V

With 1200 groove density.

Characteristic	Expression	Ex1
Lines/mm grating		1200/mm
Area of grating filled		1.16mm
Distance from grating to lens system	?	?
Percent of detector filed	80%	80%
Number of pixels used for the d-line		4.5

Based on this design, it is possible to resolve the d-line with 4.5 pixels of resolution, and by using a standard grating of 1200 lines/mm. The Optical Design 1a and 1b are similar in setup, but 1b requires more complexity since it is able to show the 0^th order for using purposes.

Optical Design 2

Using a pinhole, diffracted light will enter the system. It will then go through two or more diffraction gratings where the first order of each of them will be focused onto the cellphone’s camera lens, while insuring that both spectrums are imaged separately. The result may look something like this on the image:

Based on the calculations for design 1, it is possible to image the d-line with just one diffraction grating, simplifying the system. Therefore, this design is no longer being pursued.

Note: If after verification mistakes were overlooked in design 1, this design will be used as reference.

Mechanical housing for optics

Mechanical attachment

1. 
   Permanently adhere the device to a case
   
2. 
   Build a case such that the device can rigidly slip in and out (on rails)
   
3. 
   Assemble the device to a commercial cellphone car mount, such that the same device can mount to any smartphone within a size range
   

1. 
   Device attachment needs to be rigid
   
2. 
   Device needs to be rigid in order not to change spacing from sample to sample
   
3. 
   Smartphone compatibility of plug and play is desired
   

Software Interface

There are a couple possibilities that still need to be investigated. The first is using a webapp where the photo is uploaded through the smartphone’s browser to a site where the diffracted system is measured against a calibration. The second option is to build an app for android and iOS to calculate and calibrate the whole solution. The third option, is to mix said options together by building an app to assist with the capturing and uploaded process to the webapp.

The following site was found and will be attempted to be used to upload captured samples and receive computed results. http://spectralworkbench.org/

The basic process for using the site is as follows: A spectral image is taken of a known source, such as a specified LED, arc lamp, or CFL source, and then is uploaded to the site through the upload GUI. A calibration is then created by the software matching up peak bands. Next, a sample is taken, and uploaded. The sample is analyzed and the results are returned, relative to the calibration.

Test Plan / Validation

The test plan consists of multiple parts:

1. 
   Test the proof of concept (POC)
   

Testing the POC will involve a stable setup where the optical design is validated. A CFL bulb will be provided as the calibration, and d-line doublet will be used as the test subject. After a spectrum of the CFL is produced, its peaks will be identified manually. Calibration of the optics and alignment will be done manually (or with some aid from computational means). This will establish a baseline for the doublet test subject. The input light will be replaced and a spectrum will be produced and analyzed with the CFL reference.

If the doublet is resolvable after 5 iterations, and the accuracy is to the specification in the PRD, then this concludes the first phase of validating.

2. 
   Test the prototype
   

Similar to above, the prototype will be tested with the d-line doublet. The prototype will be subject to the next test criteria:

1. 
   drop test
   

A drop from 2 ft above an optical table

2. 
   calibration routine
   

The device will be calibrated with a known spectrum

3. 
   shake test
   

After the device is attached to the smartphone, (and calibrated) some light shaking will be applied. The doublet needs to be able to be resolved again without recalibrating

4. 
   switch to a different cellphone model
   

The smartphone will be replaced with another cellphone model, and the above repeated (excluding the drop test).

3. 
   Test the manufactured product
   

The final product also needs to be tested. A similar test will be performed as with the prototype, but on a random distribution of samples from the manufactured products. Depending on the advice from the manufacturer, and further research, an adequate sample size will need to be selected to have a 99% reliability over all the products. There should be no Dead Upon Opening (DUO) products.

Risk Assessment

The largest amount of risk is optical alignment. The product might work under certain highly controlled circumstances, but then fail to meet the resolution and accuracy specifications. With this in mind, the product will be constructed so that it is rigid, and its mounting is rigid. Calibration will be at minimum easy to do. Ideally, the system will be able to calibrate itself, taking small random adjustments into consideration when matching up the sample to the calibration.

Auto calibration might be achieved through a reticle, which the software can measure the deviation from the known reticle position in the calibration routine. This may provide (less than calibration accuracy) but still help in mitigating the error from misalignment.

The second largest risk is creating a 3D printed product where the optical components can be slotted in, and where the alignment, and spacing is correct.

The next largest risk is profile size. Ideally this device will be small enough to clip onto the cellphone and be easy enough to swap in and out to carry.

1^ http://www.anandtech.com/show/7329/some-thoughts-about-the-iphone-5s-camera-improvementsv

01 Rev A Page |

Directory: workgroups
workgroups -> Atlas simulation: Status, Performance, and Future Plans
workgroups -> Family Health History Taskforce Meeting
workgroups -> Toronto Area Safety Coalition

Download 27.08 Kb.

Share with your friends: