Dear collaborators:

  I have been asked to lay out the bonding pads for the SCT Barrel
detector to be prototyped in 1997.  Below is my proposal of the 
bonding pads for the SCT Barrel detector. Any comments?

				Nobu UNNO

===============================================================

Bonding pad layout of the ATLAS SCT Barrel detector (Proposal)


Strip #'s            |XXXXXXX|
 768 **   <>---<>-----------------------<>-----<>
 767 **<>---------<>-----------------------<>-----<>
 766 **   <>---<>-----------------------<>-----<>
 765 **<>---------<>-----------------------<>-----<>

             ...

   4 **   <>---<>-----------------------<>-----<>
   3 **<>---------<>-----------------------<>-----<>
   2 **   <>---<>-----------------------<>-----<>
   1 **<>---------<>-----------------------<>-----<>
pad row: A       B   |XXXXXXX|            C      D
                      Hybrid

(Note: <> : bond pad, ---- : strip, ** : Polysilicon resistor)


Relation to the pads on the pitch-adaptor and the FEE chip
                    |           |
   4 **   <>---<>xxxxxxx<>-<>xxxxxxx<>=
   3 **<>---------<>x<>-------<>x<>====
   2 **   <>---<>xxxxxxx<>-<>xxxxxxx<>=
   1 **<>---------<>x<>-------<>x<>====
         A       B  | Pitch-Ad  |  FEE

  The pad staggering on the pitch-adaptor is made so that the relation
of the pads between the detector and the FEE is the same with or without
the pitch-adaptor.  

  In laying out the detector bond pads, there are several aspects which 
have been paid attention,

1) Rotation(180 deg.) of the detectors:- This is to give optimum freedom
   to use the detector, e.g., placing the hybrids in either detector in
   a module

2) Staggering of the bonding pads in a row:- This is required for 
   wire-bonding.  The pitch of 80 micron is still not enough (for fast, 
   reliable automatic bonding, e.g.) to allow a single row of pads.  

3) Even number of strips:- The number of readout channels is even.  

4) The detector is to be centre-tapped, but leaves freedom for
   end-tapping (e.g., for testing).

  The first aspect was originally requested to allow the rotation of 
detectors relative to each other in a module (e.g., to match/mis-match
the bad strips).  However, combined with the second aspect, it is not
possible to rotate the detector with respect to the other.  (There 
still remains the possibility of rotating the detectors in pair.)

  With the fourth aspect, the staggering of the bonding pads for the 
centre-tapping (Row B in Fig.) will be the same as the staggering of 
the bonding pads at one end (Row D where the end-tapping would be).  
Allowing the 180 deg. rotation (with the detectors in pair) will 
require one more row of the bonding pads (Row C) which is also in the
same staggering orientation as the rows B and D.