ODU Status, QC of some ODUs, Grease Tests

1
ODU Status, QC of some ODUs, Grease Tests

• Howard Budd
• University of Rochester
• PMG Meeting
• Feb 14 2007

2
ODUs Optical Decoder Units

• ODU Optical Decoder Unit (CMS term)
• Takes light from the cable to the PMT
• The ODU is made like an optical cable with DDK
  connectors on both ends
• Its is 50 long
• But a cable must be made light tight - light
  tight boot
• ODUs are cut 2 by PMT box factory
• Installed by PMT factory in the PMT boxes
• 4 ODUs/PMT box, 8 optical connectors/PMT box

3
ODUs and Cables

• ODUs and optical cables are made using the same
  same techniques same groups
• This was the same philosophy as CMS had
• All the techniques that are applied to making
  ODUs can be applied to making cables
• Polishing technique Eileen Hahn Lab 7
• Lab 7 Lab 8 group have done a lot of work in
  getting excellent polishes
• Lab 7 understands the polishing tolerance
• Run through the same QC apparatus as cables
• This will be shown in this talk

4
Toward ODU Production

• The DDK connectors contain fiber glass
• Fiber glass is very hard on the polishing
  diamonds
• In MSU technique, the fiber glass is not polished
• The front face of the connectors are cut back
  with a carbide bit
• The fibers are glued in with the front face
  pointing down
• The epoxy flows through the holes and forms a
  pool on the front face
• The epoxy should surround all the fibers
• The connectors are polished about .2 mm from the
  fiber glass part of the connector leaving .2 mm
  of epoxy on the front face
• The polished fiber sticks has no epoxy cover
• Saves diamonds and we get a better finish

5
Debris on the Edge of the Hole

• Problem was the facing off of the connectors left
  debris on the front of the hole
• With the glass fiber in the hole, epoxy doesnt
  consistently flow out to the front face
• Some fibers had very little epoxy encapsulating
  them
• Cladding would get ripped in polishing
• Epoxy flowing out was the technique that seemed
  to work the best from earlier RD and we were
  reluctant to give it up
• We needed to remove the material from the hole
• But it is held by fiber glass and doesnt come
  off easily

6
Debris on the Edge of the Hole

• We tried a variety of techniques to remove the
  material
• Dremel 3.2 mm cutting tool bit
• Slightly countersink of fiber holes by hand
• Visually this seem to worked best
• Connectors are blown with dry N2 after reaming
• This is our default technique

7
Transmission QC Device

• LED on Calibration Pin Diode and WLS pigtail
• Calibration Pin Diode just in front of LED
• WLS fiber - jumper- cable-jumper-pigtail-pin
  diode
• Cable Measurement
• Measure without cable
• Direct Measurement
• Transmission (Cable Measurement / Direct
  Measurement)

Pin Diodes
LED
Tested Cable
Jumper
Jumper
WLS Pigtail
8
Measuring Transmission

• Measure transmission through ODU by starting off
  with a measurement of the jumpers connected
  directly together
• We call this the direct measurement
• There is a direct measurement at the beginning
  of the measurement and the end of the measurement
• For this we use the direct measurement at the
  end.
• I discovered during the connector acceptance
  tests that the transmission across a connection
  depends on how you inject light into the fiber
• Therefore, the transmissins shown here might not
  be the transmissions for the ODUs in the
  experiment
• A test using an apparatus which simulates the
  detector would have to be done to determine the
  transmission in the experiment

9
Transmission QC Device

• The pin diodes are connected to a switcher
• A picoammeter readouts the switcher
• Switcher goes one fiber to another
• Lab View controls the readout

10
QC of ODUs

• We insert the cables between the jumpers and
  measure the ratio of the measurement/direct
  measurement
• This is the transmission of the 0.5 m cable
  including one connector transmission
• One would like to know the connector transmission
• However, we have only measured fiber attenuation
  using PMTs not pin diodes and the spectral
  response of pin diodes are different
• For example, the transmission of 0.5 m fiber
  using 8 m attenuation length is 0.94
• The maximal connectors transmission we expect is
  .92
• The numbers shown include the loss from the 0.5 m
  cable and the additional connector connection
• A maximal transmission we might expect is .865

11

• We will show the QC plots from the QC programs
• Transmission
• Individual fibers
• We have used 2 sets of jumpers
• These are all the connectors from the 2d set of
  jumper
• For the 1st set of jumpers there was high rms,
  especially high transmissions, for one set of
  measurements
• Not sure of the physical mechanism

12

• Individual fibers normalized to 1

13

• Previous plots put on one graph
• Lit/conn and cable plot show whether the RMS is
  due to the individual fibers or cables
• Old polishing 5 cables
• Av79.0
• RMS1.5

14

• Top plot gives transmission cable average
• Bottom plot sets total average to 1

15

• Transmission for individual fibers with fiber
  numbers coded by symbol and color

16

• Averaged to 1
• Program prints out outlying fibers
• Bad fiber file
• The CMS ODU cut was nor lit 0.9 for all fibers
  in an ODU
• 10 out 156 ODUs fail this cut
• Most of these at beginning of production
• 3 close to the cut
• The CMS cable cut was 0.88

17

• Earlier set of ODUs
• Transmission using the first set of jumpers
• Showing the fiber glass polishing technique
  (traditional technique) Cables 81-85

18
ODU Cable QC

• For ODUs
• Pass transmission test
• Look at them with 20 magnifying glass
• Excel file that lists QC status
• For Cable
• Pass transmission test
• Look at them with 20 magnifying glass
• Pass light tight test with light leak checker
• Box with PMT in it connected to picoammeter
• Check the fiber order is OK in cable
• Box with multicolored LEDs
• Excel file to list QC status

19
Cable Status

• Cable are made like ODUs, but we need to make
  them light tight
• A polyurethane boot is put at the connectors
• The boot mold and the mold procedure have been
  developed at Rochester
• We have tested the mold at FNAL
• We have made a set of suggestions
• Mold is continuing to be tested at Rochester
• We will retest it at FNAL in the beginning of
  March
• We expect to be able to start producing cables at
  the beginning of March.

20
What about greasing the connection

• No Grease/grease
• Measured with cable QC program
• The increase of light from grease appears to be
  independent of how the light is injected in the
  fiber
• Ave.924
• With old technique .86
• With optical grease we get an increase of light
  of 1 over old technique, consistent with 0
• The optical grease fills in the imperfections due
  to the cutting of the fiber glass

21
Grease Stable over a Month
22
Conclusions

• We have QCed 86 connectors using the nominal
  countersinking technique and they look OK
• Some of these which the epoxy could have flowed
  better
• We have already produced something like 146 ODUs
  which look fine according to the QC program
• Although the front face isnt covered quite as
  well I think they should be but they are probably
  OK ODUs
• An additional 25 wait QC for a total of 171 out
  of a total of 400 needed for tracking prototype
• Grease still looks OK
• With grease the transmission appear to be
  independent of the polishing technique