Difference: UpgradeIT (1 vs. 2)

Revision 22010-03-24 - MariuszWitek

Line: 1 to 1
 
META TOPICPARENT name="WebHome"

Informacje dotyczace upgrade-u IT

Chlodzenie

Changed:
<
<
  • mail od Hansa
>
>
  • maile od Hansa
For the power:
o my previous mail was WRONG. We now expect between 1-1.5 W/FE chip, i.e
  3-4.5 W for a hybrid. Fortunately MUCH less than the 10-18 W I mentioned 
  earlier. Just for the discussion below: 
  1 W assumes a FE like for the Beetle now (600 mW), i.e. with spill-over, and
  a 4.8 Gb/s back-end. If the FE would be mad faster to avoid spill-over,
  this might add (rough guestimate) 0.5W, i.e. 1.5 W for the whole chip.
  For a purely binary FE, with Beetle speed, the channel will be "dead"
  25 ns after a hit. We expect occupancies between 2-5%, i.e. this would 
  imply an in-eff of 2-5%.

Occupancies:
o We would like to compare the simulation performance (occupancy, tracking 
  efficiency) for the follwoing cases:
  - no-spill-over, but 25 ns dead-time after a hit in a channel.
  - spill-over like we have it now. 
  - no-spill-over, no dead-time, i.e.  the minimal layout spill-over killer
    as introduced some time ago.

 
o Power per module:
  The idea is to see how we can cool a module with two sensors and a hybrid.
  The power consumption has two variants:
  1) 3 FE-chips, which consume 3 W/chip, i.e. 9 W total. And we drive
     copper Gb cables over 3 m, before going optical. Total power on hybrid 
     ~10 W. 
  2) same as 1), but we drive optical. These GBT drivers consume 3 
     W/bi-directional link. 1 bi-directional (lets see what the occupancy is, or
     drive out all binary info without 0-supp?). Hence total/hybrid: 18 W.
  My idea is to have a large surface hybrid, and a good thremal contact between
  Si and hybrid. Also  glue the two Si-sensors on top of each other, to not have
  any "hole", like we did in Delphi.
  Paula will calculate the temperature we need at the tip of the sesosrs away 
  from the hybrid for radiation damage control.

o Size of IT. Based on the mail I sent to Alexandr (cc to you yesterday), I 
  propose to go for S_109L7_200L6_200, which means we have to cover with
  the IT a total width of 7 full planks, and one half-plank, i.e.
  7.5x340 mm=2550 mm.
  If I look at IT2 (TDR), then the beam-pipe hole+tolerance is 19.8 mm,
  hence we need (2550-19.8)/75.6(sensor-width-overlap)=2x17 modules in the 
  two side boxes. (Compared to 2x7 now).

Do you agree? Please check the numbers. If we agree, I'll ask Alexandr to "open" 
the OT by this amount.

 

Linki

-- MariuszWitek - 13 Mar 2010

 
This site is powered by the TWiki collaboration platform Powered by Perl This site is powered by the TWiki collaboration platformCopyright © 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback