Minutes June 26th AFE Meeting

 

Present: Daniel Mihalcea, George Ginther, Fred Borcherding, Paul Rubinov, Rick Jesik, Jerry Blazey, Harry Weerts, Satish, Kotcher, Bross, Marvin, Stefan, Thomas Nunnemann.

 

Agenda:

 

Sanmina production and AFE 54 production status (John A.)

Status of AFE8-FPS layout (John A.)

Tests on modified AFE RevA board for CPS charge split (Paul R.)

DAB3 test stand and training (Rick J. and Daniel M.)

 

Minutes:

 

Production: Seven boards have arrived from Sanmina.  Initial impressions of the first board were quite good as visual inspections were successful, voltage and current levels were acceptable, and the JTAG and 1553 tests were successful on the first try.  Information on the boards can be found at:

 

http://d0server1.fnal.gov/users/janderson/afe_status/default.htm .

 

However half a dozen water drops were observed seeping out of the MCMs, and three of the eight MCMs had water.  Likely the water was introduced during a production cleaning cycle.

Meltronics has indicated that the epoxy seal on the MCM lids should be watertight, but examination shows gaps on the three wet MCMs.   One of the MCMs with water had voltages lightly off nominal.  Since the board was powered less than an hour there should be no metal migration between traces.   The cards are now in a 150F oven to be baked overnight.  Observation of the oven temperature levels will be added to the control room checklist.  After drying, the MCMs will be inspected for damage and testing will resume.  The other six boards will also be examined.   [Note added in proof:  The AFE and an MCM have been examined.  There are significant evaporative deposits.  Sanmina has been notified and will not mount MCMs on the remaining un-built boards.  More studies are underway with the seven AFEs at Fermilab to determine the extent of the problem and possible remedies.]

The 73 cards representing the first batch are all in rework at Sanmina and likely already washed.  John Anderson will inform Sanmina of the water problem.   The rework also includes correction of some reversed capacitors.  Given the upcoming holiday week, the second batch of cards (73 LHBs) will likely be shipped mid-July.

The AFE54 bid specifications have been rewritten and a requisition put into the system.

 

AFE-FPS Layout: Jim Franzen will be available this week to start working on the FPS layout. John Anderson will go through the schematic one more time and add a feature which permits AFE-FPSs to be instrumented as an original AFE. The board manufacturing bid has been prepared and will be released soon.  Forty cards will be ordered since there are sufficient parts and to ensure adequate FPS spares (25%).  This greater number will require modest MCM recovery from the production check out, likely tens of MCMs.  Mike Matulik and Pat Sheahan will start recovery immediately so the parts will be available for stuffing.

 

Capacitance Tests:  Paul showed a plot of the SVX normalized gain vs. the coupling capacitance for a single channel.  The data was taken on the test bench. The gain was measured from the slope of the SVX response versus input charge.  The normalization was relative to a prediction assuming 40 pf for the parasitic capacitance (observed-predicted)/predicted). The plot also included normalized predictions for other levels of parasitic capacitance.  The coupling capacitors ranged from 10 to 470 pf. For both low and high gain, the data is well described by the curve, which assumes 40+/-5 pf for the parasitic capacitance.  This may be reduced to 35 +/-5 pf due to real resistance on the cables.  

Tonight Paul will start measuring the response on a cold cryostat.  Data will be collected for thresholds over the entire dynamic range needed for preshower.  The channel-to-channel gain will also be studied since they vary by a factor of two. This is in conflict with early results which showed much less variation.  The variations have been confirmed with Thomas’s fitting routines.

 

DAB test stand:  The DAB3 test stands are now assembled. All the cables in place but a LHB has not yet been tested. Charge injection test have been exercised for all channels of an AFE.  Pat Sheahan will brace the backplane for the AFE.  The software still requires modification.

Five people are available to man the test stand.  Two understand the software and two more are learning. Mike has written a well document skeleton for non-experts, which should be done within a few weeks.  The students will then insert individual test modules into the skeleton.  Tina Hebert will help with off-line analysis and will supervise testing while Rick Jesik is gone next week. There has been some progress on database development.

 

            SIDET preparation: Two test stands are now functioning.   Some software is missing, in particular the final version of code that interfaces with the AFE cryo-control.  Stefan Gruenendahl, John Anderson, and Bob Angstadt (in that order) need to develop and install various pieces code. Once Stefan is done we will evaluate John’s task list before proceeding.

Staffing is adequate but replacements will need to be found for three students departing in August.  Daniel will talk with Marcel about additional technicians.  The new personnel will need to be trained before the students depart.

 

            MCM Replacement: Work has begun on the schematic, which should be done by the end of August. A commercial firm will layout the hybrid.  The current topic of concern is temperature control and there is some effort to understand power usage.  A design rule is that there should be no or minimal modification of the AFE. At present there is some thought an AFE regulator may need to be replaced but beyond that the replacement will be pin-to-pin compatible.   

            Die submission is planned for November and turnaround is expected eight weeks beyond that.  The first prototype should be available in January.  Fifty to 75% of the functionality should be tested before prototype receipt since commercial parts will be incorporated.  If a second submission is required testing at Fermilab will pace the schedule.  The second submission would probably be available late spring or early summer.