Fiber testing procedures:

Design:

The WLS fiber was tested for attenuation by a device designed specifically for the fiber that will be used in the MINOS experiment.  The main part of the device is an aluminum cylinder that has a double spiral groove machined into the outside surface.  The groove is wide enough to accommodate a 1.2mm fiber with the expected tolerance of the fiber.  The fiber is completely embedded in the groove during the measurement.  There are 21 blue LED ports machined at two flat sections on the cylinder.  The fiber is exposed to these sequentially switched LEDs at these positions.  The circuit diagram can be found attenckt.ps as postscript, attenckt.pdf as pdf, or attenckt_page_1.jpg and attenckt_page_2.jpg as jpeg.  One end of the fiber is mated with a PMT tube for detection of the light trapped in the fiber when the blue light is converted to green by the WLS fiber.
 

Procedure:

1) A 10m length of WLS fiber is prepared by polishing the end with lapping paper in sequence:  10 micron, 3 micron, .3 micron.  Finally polishing with .05micron aluminum oxide powder in a water suspension on a chamois cloth.  This reduces overall amplitude variation from fiber to fiber to approximately 2%.  The far end of the fiber is simply cut at 45 degrees which has been shown to be sufficient to inhibit reflections.

2) The fiber is wrapped into the spiral groove by hand, being careful to start at a fixed distance from the PMT.  This assures the correct measurement of attenuation length from fiber to fiber.  This is particularly important at short distances where the attenuation is a strong function of distance from the PMT.

3) Screw black-backed cover plates over the LED ports.

4) The polished end of the fiber is cleaned with alcohol and inserted into the cookie in front of the PMT and the cover placed over the device.

5) The high voltage for the PMT is switched on.

6) After a few seconds, the output stabilizes, and the data acquisition program can begin.

7) The program automatically records the voltages, due to the current at the base resistor.  It automatically finds the beginning of the flashing sequence and records from that point on.  These 22 raw numbers are saved as the output data (one for the pedestal value).

An example of a graph of these data is shown in this figure:  (.eps) (.jpg)  (.ps)

The amplitudes at each position are a combination of the attenuation length and the amplitude of the LED at the particular port.  The driving voltage for the LEDs is kept constant, but even so, different LEDs have different light output.  This must be accounted for by calibrating the output of each LED, and will be different for each attenuation device produced, since it depends on the particular LEDs, and, to a lesser extent, how they are mounted.

The calibration values for this device look this:
 (.eps) (.jpg)  (.ps)

Most of  the LEDs are reasonably consistent, within 10%, but since we want a measurement precision of about 2%, they must be accounted for.  Particularly the LEDs with larger variation.   After accounting for these variations, simply by normalizing, we get an attenuation curve that looks like this: (.eps) (.jpg)  (.ps)

This can then be used to compare the absolute light output and attenuation length between fibers of different batches, or even within a batch.  Four parameters can then be extracted from these data by fitting with a double exponential function.
 
 

Specifications: (boring details)

  • Cylinder diameter:    30.49 cm
  • LED:     Panasonic LNG901CFBW
  • Wavelength:    470+-30nm
  • Phototube:    Six stage Hamamatsu R1666 bi-alkili photocathode
  • Voltage:    800V

    •  

    Minos Fiber attenuation length device pictures.


    First, an overall view of the spiral grooves on the cylinder, the LED ports are seen in the background, along with the  phototube with a WLS fiber entering the cookie. (.eps) (.jpg)  (.ps)


     

    Closeup of the fiber winding on the cylinder and entering the phototube, along with the LED cover. (.eps) (.jpg)  (.ps)

    A view beneath the LED cover, showing the cavity with the fiber passing through.
    LEDS installed, but water clear when not on. (.eps) (.jpg)  (.ps)


     

    Finally, after the LEDS and the assembly were mounted to the base for the light-tight assembly. (.eps) (.jpg)  (.ps)


     
     

    4PPT fiber Results:
    The following figure shows the relative amplitude as a function of distance for all twelve batches of WLS fiber acquired for the 4PPT.  They all show good light output except for one, S35-E, which was not able to be polished to a good finish, thus reducing its amplitude. (.eps) (.jpg)  (.ps)