Status report on Brick: Controller, Mechanics, Power Consumption

1
Status report on Brick Controller, Mechanics,
Power Consumption

• Work done by
• Marco Incagli, Franco Spinella, Marco Piendibene
• Electronics meeting 7/8 feb 2005

2
PART 1 the new HV controller board
3
New HV controller round 1

• New HV controller received in Pisa Oct, 25 2004
• To test it, an interface board (made by CAEN),
  that sends the equivalent of LeCroy commands, is
  used
• First tests on analog part
• HOT part OK
• COLD part not working
• Controller back at CAEN where the problem on COLD
  part was solved
• Shorts due to some residual soldering material

4
New HV controller round 2

• Back in Pisa on Nov 15
• After few tests the controller stops working
• CAEN engineers come to Pisa and find a chip that
  was exploded presumably a wrong alimentation
  problem (our fault)
• Controller back at CAEN where the chip is
  replaced
• At the same time, the maximum possible voltage is
  moved from 1000V to 900V (900/1024 steps)
• Controller in pisa end of November all
  regulators work fine

5
New HV controller round 3

• Controller at CERN 9-10 December tests with A.L.
  V.K. using JINFV2 and tracker backplane
• JINFV2 does not correctly talk with HV controller
• Try to reproduce the problem in Pisa and then
  study the problem together with CAEN engineers
• For this we need the ECAL backplane
• The EBP arrives in Pisa end of december

6
New HV controller round 4
Setting voltages with A.L. program

• The system with EPPCANEBPJINF2EHV is mounted
  and tested
• Software from Lebedev, re-written by M.I.

Keithley readout (20channels)
EPPCAN interface
7
New HV controller round 4
JINF
New EHV controller
10 PMTs
Data cable

• EBP

8
New HV controller round 4

• CAEN engineers come to Pisa and identify the
  problem in the ACTEL code (their fault) ?
  synchronization ? see talk of Spinella
• Controller back to CAEN, ACTEL chips (AB)
  unsoldered
• Only one is mounted back in order to have another
  possible test on the same board for an eventual
  re-programming

9
New HV controller round 5

• Controller with new ACTEL in Pisa Jan, 24
• Communication problem carefully tested
• With correct commands, the Brick always responds
  correctly (intensively tested sending million of
  commands and checking the answer)
• With pathological commands, the controller stops
  working, until a software reset (64 NULL commands
  in a row) is issued ? possible critical point

10
New HV controller current situation

• HV controller is currently winning, but we have a
  few more rounds to go
• Brick tests scheduled for 9/10 feb at CERN with
  A.L.V.K.
• If CAEN has the green light, and no reprogramming
  (and retesting!) is needed, they can put the
  second ACTEL on board in two working days
• Once both ACTELs are onboard, we can make other
  tests, like swithcing on both A and B sections
  and sending commands on both lines, while waiting
  for the mechanics

11
PART 2 power consumption
12
Brick power consumption

• Current drawn with only 1 section operative (A or
  B) and all PMTs at 800 V (at t0 normal voltage
  is 700V, but it may increase with time)
• 3.3V ? 0.09A 5.0V ? 0.09A 28V ? 0.32A
• Ptot 3.30.09 50.09 280.32 9.8 W
• In worst case scenario (no commands arrive at
  EHV) both sections are on and elevators set to
  maximum voltage (reference voltage and transient
  time tbd)
• 3.3V ? 0.15A 5.0V ? 0.2A 28V ? 0.45A
• Ptot 3.30.15 50.2 280.45 14.1 W

13
Power drop at EHV and at PMT level

• We have measured the current drawn by a PMT set
  at 800 V . The total output power (drop at PMT
  level) is
• Pout 800V 0.160mA 55 7W
• Therefore the power drop at EHV level is
• Pdrop Ptot Pout 9.8W 7W 2.8W in
  standard conditions
• Pdrop Ptot Pout 14.1W 7W 7.1W in worst
  case scenario

14
PART 3 mechanics and SQ tests
15
Mechanics and SQ tests

• Mechanics is the same for all bricks
  (TOFRICHECAL)
• IAC institute, in Canarie islands, will build the
  mechanics for EHV (thanks to Berdugo who is
  following the job)
• The first prototype, to be used for EHV QM tests,
  will be done by in Pisa CIEMAT, Madrid
• SQ tests, vibrationEMI(TVT), can be done in
  Terni ? mechanical fixtures must be built.
  Necessary before ordering FM, which take 8 months
  to build!