Stiff%20tail%20troubles

1
Stiff tail troubles

• The new stiffer hybrid tail flexes the
  interconnect card on the rod by about 1 to 1.5 mm
• Some tilt in the connector as well
• Force needed to push board flat 250g
• Rod tested fine at room temperature, but not in
  the cold (could be open connectors)

2
Old, softer tail in the same position

• Some flexing of the PCB board
• No tilt in the connector
• Force needed to push board flat 80g

3
New bends to the tail

• Pos. 3 on the rod (smallest gap between module
  and interconnect card)
• Tail bent in situ on the rod
• Bends are close to c-fiber than before
• Some improvement, but still deflection of the PCB

4
New bends to the tail

• Same bend moved to pos 5 (larger gap)
• No deflection of the PCB
• But Highest point of tail is 1.5mm above
  stay-clear, not acceptable

5
New bends to the tail

• Tail bent like original (soft tail)
• Added kinks with tweezers, while module was on
  the rod

• Less deflection of the PCB ( 100g)
• Also works in pos. 5
• The 2 modules used for these tests showed no
  opens in the tail after extensive bending

6
The tail tale continues

• The goal is to bend the tails only on the vacuum
  fixture, not on the rod, to avoid putting stress
  on the connector
• Try out new bending tools (the tweezers used
  previously crease the tail dramatically)
• non-tapered rounded tweezers
• Modified sur-clip pliers

7
Bending with the new tools

• Was not able to repeat the tail bend shape
  previously developed (see slide 5) by bending on
  the vacuum fixture with either tool
• Try different shapes
• Tweezers work better than surclip pliers

8
Best bend so far

• Small radius on top, large radius on bottom
• Use new tweezers to make the small radius bend at
  the connector end, and bend tail back over the
  module frame by hand to make bottom part more
  rounded
• Board deflection 80g
• This seems to be as good as it can get without
  bending the tail while the module is on the rod
• A SS6 rod was assembled using this bend on all
  modules, still needs to be electrically tested in
  the cold