Minutes of AFE meeting
    April 27, 2001
++++++++++++++++++++++++++++++++++++++++++++++++

F. Borcherding, A. Bross, G. Ginther, P. Grannis, M. Johnson,
T. Nunnemann, P. Rubinov,  Weerts, J. Womersley

1.  Sanmina has all the change drawings.   They said they are
short of LF347D opamps (we can provide if need be, but believe
we shipped sufficient before).  They also claim to be short
of MCM's; again we shipped sufficient, and they have indicated
that all the shipments that contained MCMs were in fact received.

2.  Paul reported on the tests with AFE board modified caps on 
DVDD and AVDD.  These connections are provided on the existing
AFE8 Rev1C layout.  The differential nonlinearity is much
improved.  Visually, some (expected) odd-even DNL persists, but
clearly this is a big improvement.

3.  Fred reported on the testing plan.  There are 7 separate test
    steps:
    1)  add parts to be installed here, inspect for parts present in 
        correct orientation
    2)  DVM/ohmmeter tests without power
    3)  some LV power on; program the PLDs and microcode; check
        basic power levels
    4)  full power on boards, with SASEQ, 1553, VME, Bit3 interface.
        Scope probe for correct signals;  PC supported tests of board
        basic functionality
    5)  DAB3 test stand test of full board properties, including
        charge inputs, bias supply, cryo controls, characterization of
        operating values, noise scans, gains, disc. settings,
        pedestals, etc.   This to be done with shifters from the
        collaboration, under direction  of engineers (Mike M.?)
    6)  Crate of 16 AFE boards with VRB, VRBC, Sequencer in DAB3
        using Jadzia's SDAQ software.  This is to test full crate
        operation as a '10% test' prior to installation on the
        platform.
    7)  Some operation at Lab3 to verify that boards that pass DAB3
        tests work satisfactorily while cold and with light inputs to
        VLPCs.
Steps 1)-4) will be done at SiDet with 5 techs (2 full time and 3 at
least half-time).   We expect to start installing test fixtures at
SiDet by May 7.  Step 5) requires shifters from the collaboration;
we expect to complete the documentation of procedures by May 14, and
training could begin then.   We imagine that we would ask for 4 hour 
shifts, initially for day and evening shifts, but could expand to
around the clock if this step is a bottleneck.  For 28 half shifts
per week, we see the need for about 10 people to be involved.

At present, development of the software for this test is a critical
path item.  For Bob Angstadt (low level software) and Paul Rubinov
(higher level software and algorithms), this software is the highest
priority.

After the Step 6) tests we would install these ~16 boards on the
platform.  Marvin advocates doing so at the beginning of a 2-3 day
shutdown so that iteration of tests/results can occur.

The extent of Lab3 tests needs to be evaluated after the first set
of boards is tested there.

4.  We will modify one MCM to have lower capacitance inputs.  Each of
the four SIFTs will get different values, so we can determine the 
charge input vs. capacitance function, and thus get a measurement of
the cable and stray capacitance.   Values of 10, 22, 44, and 100 pF
were suggested.   Donna will oversee the modification of this board;
when available, we will ask Paul R. to do the Lab 3 tests.  
We would like to use the highest gain cassette available (thought to
be a 40K gain cassette.)  We need to determine the cap value before
assembling the CPS AFE8s.   (The bare board order is in the works
within FNAL purchasing.  Johnny Green is assembling the needed parts
orders.)    Fred B. was asked to verify the calculation of charge
inputs for the CPS to double check the calculation of Paul G.

5.  George Ginther reported that we are presently out of LHe at Lab3, 
but will get new supplies next week.  This supply will last about
4-6 weeks.   There are 97 characterized cassettes.   93 - 94 are ready
to install.  We need 99 on the platform, and there are about 110 in
all to test.   Progress is limited by flex circuit deliveries;  we
need 68 more and recently the delivery rate has been low.  The last
shipment had 7 mil (not 5 mil) coverlay and is not useable or
salvageable.   George guesses it will be 2 weeks to having 99
cassettes for installation.  Getting Volker back would be great!

6.  We discussed the timetable on which we should decide on AFE8'
vs. AFE12 for FPS.  The schedule is now set by the availability of
engineering to do layout work -- and with the priority given to test
of production AFE8, this will not become available for perhaps a
month.    Marvin believes we can accomplish the AFE12;  Fred noted
that the AFE8' is simpler and can be delivered faster (x2??).   George
noted that MCM availability is an issue, and would favor AFE8'.
We expect to take up the decision on AFE8' for FPS in about 3 weeks.

Next meeting will be 4PM on Tuesday, May 8.