Date: Fri, 25 Mar 2005 10:21:22 -0500 (EST)
From: Dan Edmunds <edmunds@pa.msu.edu>
Subject: Re: P0 paddle card review

Hello John,

Thank you for the time Thursday afternoon to talk about the design
of the new ADF Crate Transition Paddle Card.  This note is just to
confirm some of the points that we discussed.

1. The P2 BLS stuff is all understood.  Nothing was changed in
   the schematic.  Just the foot print of the P2 ERNI connector
   was flipped end for end,  e.g. what was row 1 is now row 32.
   You are getting a couple of pcb's made so that this part of
   the new ADF Crate Transition Paddle Card can be layed out from
   a known good design.

2. Using the sample connectors that you have, we verified and
   agree that the AMP LVDS cable connectors can not directly
   plug into the 7 column "P0" backplane connectors at the back
   of the Wiener crate.   Thus, even though it was not part of the
   original design, a paddle card for the Channel Link signals has
   been needed all along  just to allow the LVDS cable connectors
   to properly plug into the ADF Crate.

3. As shown in your mechanical drawing of the new paddle card,
   the AMP 1-106014-1  connector that receives the LVDS cable
   connectors is one long connector that receives all 4 LVDS
   cable connectors, packed right next to each other, justified
   towards the top of the connector.  Your were going to think
   about:

       - Is there an advantage to using 4 short separate connectors
         to receive the 4 LVDS cables.  The possible advantage is
         that we could space out where the LVDS cables plug into
         the paddle card  and that might make it easier to change
         a bad cable in the fully assembled system.

       - If the design stays with one long AMP 1-106014-1 connector
         then is there an advantage to putting a row of GND pins
         between each of the 4 LVDS cables.  We thought that might
         allow the Latches on the AMP 1-106014-1 to better hold
         just one of the LVDS cable connectors each.   The row of
         GND pins may provide some isolations between the various
         cable signals.  A one row wide plastic connector housing
         could be put over these rows or GND pins to make it clear
         where the LVDS cable connectors plug in.

4. As drawn, the schematic of the Channel Link section of the new
   ADF Crate Transition card shows all the active connections to
   the AMP 1-106014-1 connector justified towards its top.  If rows
   of GND pins are used between the LVDS cables then the schematic
   will need to be changed.

5. The schematic shows more GND connectors to the "C" column of
   the AMP 1-106014-1 than there are signal connections to the
   "A", "B", "D", and "E" columns.  Did you mean to draw it that
   way ?

6. Has there been a decision about splitting the ADF Crate Transition
   Card GND planes between the  "P0" Channel Link digital area  and
   the "P2" BLS analog signal area ?

7. I hope that a test of data transfer from ADF card to TAB card can
   take place well before layout of the Channel Link section of this
   new ADF Crate Transition card starts.  If there has been any
   mistake in the pinout of ADF or TAB then this could be corrected
   in the wiring of the ADF Crate Transition Card.  I have not
   studied the pin to pin connectivity between ADF and TAB.  I layed
   out the ADF-2 putting the Channel Link signals where the
   specification said to put them, but we need to test this.

Thanks again for getting things drawn up so that people have something
specific to think about.

          Dan