Space requirement and hardware needs in PP2, PP3 and USA

1
Space requirement and hardware needs in PP2, PP3
and USA

• C.Gemme, P.Sicho, S.Kersten

2
(Many) assumptions

• Starting from Ver2.4 of Services Estimate
  (Danilo)
• Main numbers
• 120 modules, 60 optoboards.
• Only 1 LV per module (I2A) but up to 2 HV per
  module.
• Same services as today for the optoboards and
  overall schema.
• USA15 installation preferred
• easier maintenance, access and control.
• Few pixel racks already available.
• Splitting between two Counting Rooms would need
  more time for installation.
• Time installation in cavern should be minimized
  (env activation).

3
Assumptions Ver 2.4 2009/01/29
4
Pixel DCS in the pit
5
DAQ crates

• Requirements
• One output line 160 Mbits per chip of 20mm x 20mm
  size (present B-layer has 2 DATA output lines,
  overall 160Mbits per module of 16.4mm x 60mm
  size).
• 8-ways ribbon in and out for each optoboard (2
  modules) -gt Need 60 DATA output 8-ways ribbons
  and 60 TX input 8-ways ribbons.
• USA15
• As for the actual b-layer, one BOC per 8 input
  channels and receiving 8 data signal.
• Then 60 BOCs are needed.
• As a crate can allocate max 16 BOCs, 4 crates are
  needed, 2 racks (7 spare).
• They have to sit no too far from the TTC crate
  (Y2905). Delay compensation for trigger cables
  propagated between racks is anyhow possible (do
  not know limit). Have also the doubt that the
  current TTC crate should need to be expanded.

6
PP2 crates

• Current services
• The ratio between LV module supplies/LV opto
  supplies is (72)/1.
• Each regulator board can regulates 16 LVs up to
  14 are used for modules (2LVs per module), 2 for
  optoboards (one is redundant).
• A PP2 crate serves up to 12 optoboard and 127
  modules.
• IBL Requirements
• 120 LV modules to regulate, 60 VVDC LV for the
  optoboard.
• Ration between Modules and Opto supplies is now
  2/1.
• Need_at_PP2
• Need to change the input schema from Counting
  room to accommodate the 60 optoboard supplies.
• Just assuming that
• there are 180 LV channels to regulate
• Each regulator board can regulate 16 channels,
• Each crate can accommodate 12 boards
• In principle 2 crates, one per side, should be
  sufficient (7 spare).
• Not considered the NTC opto boxes (just passive
  PP, not cumbersome)

7
PP2 crates room
Sector 9 (detector side ) Easy access, lot of
space!
Sector 13 (bottom of the detector) Room for a
crate, not very confortable
8
LV PP4

• Current services
• In the actual design one LV crate is done which
  one Wiener supply (12ch) that are splitted in
  12x6-7 by a LV PP4. Up to 15 V, 15 A can be
  delivered per output.
• 48 crates are used to supply 2722LV PP0s. 6
  crates are sitting in one rack.
• Requirements
• 120 LV supplies for modules. Higher current
  (2A).
• USA15 or US15
• Just assuming that
• there are 120 LV channels to regulate,
• LV PP4 splitting factor equal to 4
• then 30 Wiener channels are needed -gt 3 crates
  (0.6 rack) are sufficient with 6 spare channels.
• All this is based on existing Wiener crates (a 12
  channel), but there are new crates nearly ready
  which allow a much more dense packing in the rack.

9
PP2 aux PS

• Current services
• Some Wiener channels are also needed for the PP2
  controller board n PP2 crates, n controllers, n
  x 2 Wiener channels.
• Requirements
• In case 4 PP2 crates are needed, 8 Wiener
  channels should be used.
• One extra Wiener crate could be installed where
  other PP2 aux PS are located as there is place
  for an another crate there (Y.15-14.A2).
• No extra space requested!

10
HV PP4

• Current services
• One ISEG module has 16 independent floating
  channels each up to 700 V, 4mA max. output. In
  the beginning of the experiment one readout unit
  (6 or 7 detector modules) is supplied by one HV
  channel. PP4 is needed to do the splitting.
• One rack (four crates) is powering 470 modules.
• Requirements in USA15 or US15
• 2120 HV supplies for modules. 240 channels.
• Assuming no splitting ?240 HV channels -gt15 HV
  modules 2 crates (a 6U).
• Assuming a splitting factor 4 at PP4 ? 60 HV
  channels ? 4 HV modules 1 HV crate (6U) 1
  HV-PP4 (6U)
• In both cases 2 6U crates are needed.

11
SCOLink

• Current services
• SCOLink provides 4 outputs (signals) to the
  optoboard (Vpin, Viset, Opto_Rst, VVDC) and has
  an Interlock input.
• 1 SCOlink crate (3U crate) provides 16x signals
  (16 channels).
• Requirements in USA15 or US15
• Assuming that the same supplies signal are needed
  for 60 optoboards in the IBL, we should assume 4
  SCOLink 3U crates (64 channels). 7 spare.

LU

• Current services
• LUs receive NTC module and opto temperatures and
  provide interlock to the relevant PS.
• One LU crate (3U) has 84 NTC corresponding
  inputs.
• Requirements in USA15 or US15
• To connect 120 module and 60 optoboards, we would
  need 3 LU crates (28 spare).

12
BBIM/BBM/OptoServices _at_ PP3

• PP3 opto services
• crate splits the signals coming from SCOLink
  crate and send NTC opto to BBIM
• we need to have 3 opto services crates 3U high
  (3x2472 channels) for 60 optoboards
• BBIM
• (6U high) you need to measure 12060 NTC signals,
  so you need 180 channels NTC PP2 (10 more??)
• you need 1 or 2 BBIM crates (4 x 56 channels)
  what gives you 224 or 448 channels
• BBM
• BBM crate (3U high) assume 1 crate
• OVERALL WE NEED ONE rack at PP3 (could be
  foreseen in USA??)

13
Optoboards

• Current services
• In the actual design detector, optoboards are
  cooled and then optoheaters may provide some more
  temperature tuning (6 opto granularity).
• Blayer optoboards produce more power so any
  6-opto array might have a different temperature
  profile. Not optimal!
• Not knowing which will be the future optoboards
  performances vs temperature, I think we still
  would assume a temperature regulation will be
  needed.
• Requirements
• 60 identical power consuming optoboards.
• USA15.
• Some more power supplies channels could be needed
  and should be foreseen.
• Dont think it is an issue for the actual
  services requests but should not be forgotten.

14
Summary

• Needed racks (requested to G.Mornacchi et al)
• Control room (not a final assignment!)
• 2 DAQ racks,
• 1 rack for HV/LV
• 1 rack for SCOlink, LU and IDB. The latter two
  could collapse in one new unique I-matrix crate.
• 1 spare
• PP3 1 rack (could be moved in USA15?)
• PP2 2 PP2 crates plus NTC opto boxes. Strongly
  depending on routing out of PP1. No space problem
  in sector 9 or 13.
• Still some space free in Pixel Racks
• USA15 Y1614 and Y1014
• PP3 we have in UX US side Y30.03x4 with only one
  BBIM. Nothing on USA side
• We need 3 more racks in USA15 and one at PP3 USA
  side (There are free racks close to the pixel
  rack it should be perfect)

15
USA level2 - 14
Pixel empty racks
CR rack (3 PC inside)
From the rack wizard, view of USA15 level2
ROD racks