Wednesday, June 11, Notes

Present: A. Byon-Wagner, R. DeMaat, R. Downing, J. Patrick, T. Shaw,
         K. Schuh

   Terri Shaw was present in order for the PCBCP to conduct the review
of the TRACER module. Terri provided the committee with the following
documentation:
	TRACER specification
	Schematic diagrams
	PCB dimensions
	Front panel design
	Board stiffener design
	FPGA designs
	PLD designs
	Checkout procedures
	Checkout module descriptions

   The TRACER uses approximately 5 amps @ 5 volts for a power consumption
of 25 watts.  Fuses and tranzorbs are installed on the 5-volt plane and
on the 5-volt precharge circuit.  Each of the two fuses are rated at 10
amps.

   The TRACER implements a modified version of a VMEbus slave interface.
Only 32-bit aligned data transfers will be supported. These may be either
single word transfers or block transfers.  Only extended (32-bit) addressing
modes are supported.  All boards will be assigned a unique geographical
address through the use of backplane pins on the CDF Custom J2 backplane.
TRACER modules will respond to address modifier codes 09 and 0B.  The module
ID is implemented in a fashion developed prior to the VIPA standard.

   Terri pointed out that electrically isolating the production version of
the front panel from the circuit board electronics will be tricky due to the 
need to minimize the leakage of cooling air between connectors and their 
associated holes in the front panel.  Several of the connectors have
metallic cases which easily short to the front panel if they come into contact
with it.  Keith pointed out that it was learned in the FASTBUS days that
leakage through the front panel must be kept to a minimum or cooling becomes
ineffective.
   Solutions discussed included anodized front panels (but scratches can
overcome the insulating property), hardcoating (a special process available
only in a brownish color and before/after thickness would need to be taken
into account), and insulating bushings.  Terri will consider these options
when designing the production version of the front panel.

   Terri pointed out that portions of the injector/ejector handles that
are made of plastic and come into contact with metal parts of the subrack
are wearing out very quickly.  Bob Downing reported that some of the
earlier crates had two thin metal layers for the handles to grab into
and this may be contributing to the failures that Terri's group has
experienced.  This needs to be looked into.  One alignment pin has broken. 

   As a result of the upgrade electronics effort not yet having developed
a system for assigning drawing numbers, the drawings for the TRACER use a
system based on dates.  An action item resulted from this situation.

   Once the module goes into production, fabrication information will be
put on file at B0.
  
   Action items that resulted from this meeting:
   Bob DeMaat will see to it that we assign key codes to each module and
subrack slot.
   Bob DeMaat will talk to John Rauch about assigning a block of drawing
numbers to the upgrade electronics.  These numbers may be applied directly
to drawings created at Fermilab and they can be mapped to those created by
outside organizations with the Fermi number added on a sticker.
   Terri Shaw will see to it that explicit revision dates are added to the
drawings rather than embedding dates in the drawing numbers.