Power supply planning for upgrade

OT Scintillating Fibre

Wilco VinkOn behalf of the SciFi group

19 May 2014

OT Wiener Maraton systems◦72 outputs * 300W = 21.6kW◦6 power boxes,12 channels per power box

Wiener output max 8v/50A◦But: 1v margin needed for cable voltage

drop (25m+25m)◦Currently configured Vout = 7V, voltage at

front end +/- 6VDetector power for 432 front-end boxes

+ controls: 11kW

LV Power supplies

Wilco Vink 219 May 2014

Wilco Vink 319 May 2014

LV distr box1

LV distr box2

LV distr box2

LV distr box1

FE Boxes

6 rectifiers and controllers

6 power bins in bunker below trackers3 C-side3 A-side

ShieldingWall

FE Boxes

FE BoxesFE Boxes

Low voltage cabling

TO each C-frame12 35 mm2

Thick cables to detector distribution box◦25m*2 35mm2 for 27A

For upgrade up to 45A

Thin cables from distribution box to FE-modules/ROB◦8m*2 2.5mm2◦9 FE-modules 6 ROB

Same supply cabling for upgrade

Wilco Vink 419 May 2014

Wiener1: +6V/45AWiener2: +6V/45AWiener3: +6V/45AWiener4: +6V/45A

Wiener5: +6V/45AWiener6: +6V/45AWiener7: +6V/45AWiener7: +6V/45A

12 Cframes

LV distribution boxes

FEASTMP DC-DC convertersPower requirements not yet all

known, expected to have more details from test results Q3/Q4 2014◦ Test of PACIFIC2 with analog and

digital circuits second half of this year

◦ Microsemi FPGA, proto boards planned for Q4 2014.

Upgrade FE consumption

Wilco Vink 519 May 2014

Cern DC-DC converter◦ Specifications

Vin=5-12V Vout=0.6-5V Iout<=4A frequency=1.8 MHz

Power supplies needed for front-end modules◦ 1.2v:

GBT-SCA Microsemi FPGA core voltage

◦ 1.5v: GBT 1.5V Microsemi FPGA PACIFIC

◦ 2.5v: LD VTTX 2V / 0.15 A TIA VTRx 2V / 0.1 A Microsemi FPGA 2.5V / 0.5A ,

4*= 2A

Per ROB worst case, not all front-end boxes will be equipped with all data GBT’s transmitters◦ PACIFIC/Microsemi FPGA unknown (??W)◦ 16 GBTx 1.5V/728mA (~17.5W)◦ 16+1 LD 2.5V/0.15A (~6.4W)◦ 1 TIA 2.5V/0.1A (0.25W)◦ 2+8 SCA 1.2V 0.1A (0.12W)◦ 2 Master GBT chipset : 2W (4W)

Total per ROB: (28W + unknown)DC/DC FEASTMP converters efficiency ~75%

Upgrade FE consumption

Wilco Vink 619 May 2014

Reuse the OT Maratons Additional Wiener

power boxes in tunnel◦ Extra Maraton systems◦ Extra thick supply wires

from power box in bunker to LV-distribution box on detector

◦ New distribution box (no of connections and only positive voltages)

Upgrade of the LV

Wilco Vink 719 May 2014

CEAN experiment multichannel system SY1527LC

HV supply boards: A1833BPLC, double width, using 28 HV channels with positive polarity

Cabling:◦ 48 long-distance multi-wire

cables◦ 12 Cframes x 4

cables/Cframe◦ from Patch panels in D3 to

the C-frames are connected to the HV power supplies by means of two junction boxes, one per detector side(located: D3-D03)

OT HV power supplies and distribution

Wilco Vink 819 May 2014

Most of this not usable for SciFi!

100v Max. required◦ 50mV resolution

30V for KETEK SiPM 65V for Hamamatsu

Total number of HV channels:◦ 288 Read Out Boxes * 16 SiPM’s◦ Share a HV line per 4 SiPM’s (1152 HV channels)

Assuming: sufficient current available good uniformity among SiPM array Vbd good temperature uniformity

Ordered for SiPM test @ EPFL◦ Caen 32 10mA channels, per module◦ 8 modules per crate 10mA per output

Planned for the final system 15mA / channel

HV for upgrade

Wilco Vink 919 May 2014

Guido Haefeli (EPFL) looking into HV system for SiPMs

LV supplies◦We can reuse the Wiener Maraton power supplies

(assuming maintenance contract ok!) Re-configure the Negative output to positive

◦We need additional power supplies and cabling◦Voltage drop over thick cables will increase with

higher currents less margin◦Q3/Q4 2014 will give us a better view of total power

HV supplies◦OT HV modules not usable for upgrade (mainframe?)◦ (re)use of cabling needs further study

Conclusion

Wilco Vink 1019 May 2014