Problems of the Front-end Electronics
Low Voltage Power Supply in CMS
SUBDETECTORS:
·
TRACKER (TR);
·
ELECTROMAGNETIC
CALORIMETER (ECAL);
·
HADRON CALORIMETER
(HCAL);
·
MUON CHAMBERS (MU).
·
EMU – 483830;
·
ECAL – 236830;
·
CMS > 106.
POWER CONSUMING FRONT-END ELECTRONICS
Table 1
SUBDETECTOR |
DIGITAL I, kA |
ANALOG I, kA |
TOTAL POWER P, kW |
ECAL |
13,7 |
7,2 |
92,5 |
HCAL |
2,35 |
1,7 |
21,6 |
EMU* |
3,7 |
2,6 |
43 |
*EMU – Endcap Muon
Chambers
CLASSICAL SOLUTION – POWER SUPPLY CLOSE TO THE LOAD
·
MAGNETIC FIELD
·
RADIATION
4 VOLTAGES IN ONE CHANNEL: +5, -2 – analog
+5, +2 – digital
HUGE POWER LOSS!
ULOAD, V |
ULVPS, V |
DU, V |
I, kA |
DP, kW |
+5 |
+11,9 |
5,9 |
11,9 |
70,2 |
-2 |
-5 |
3 |
1,9 |
5,7 |
+5 |
+8,3 |
3,3 |
2,3 |
7,6 |
+2 |
+5,8 |
3,8 |
5,0 |
19,0 |
Table 2
Total
102,5
102,5 kW > 92,5 kW
IF 120 m AWG 2 – 169,5 mm2 instead
of 8 mm2 CABLE
2,2 kW instead of
47,6 kW
CABLE WEIGHT 176 900 KG
CABLE CROSS-SECTION AREA 6320 cm2 (80X80 cm)
CABLE PRICE 1 726 000 $
THREE STAGES LOW VOLTAGE SYSTEM
VICOR DC-DC CONVERTERS
Fig. 3
FOR ONE ENDCAP
30 kW input
power at 300 V (70% efficiency) -
100 A current
trough the 100 m long cables
AWG – 2 -
VOLTAGE DROP – 1,5 V
- POWER LOSS - 144 W
MAGNETIC FIELD
SHIELDING FOR DC-DC CONVERTORS
FERMILAB’S PROTOTYPE - 1,25 cm SOFT IRON
PLATES
WITH WATER COOLING
Fig.
4
RADIATION
PROBLEMS
TESTS WITH LOW
ENERGY NEUTRONS:
OPTO COUPLER
REPLACING BY HEWLETT-PACKARD’S
DEVICE
NEUTRON ENERGY – 60 – 200 MEV – SINGLE EVENT EFFECTS
TEST – 60 MeV
PROTON BEAM IN Louven-la-Neuve, Belgium:
V300B12C250AL,
VIN=300V, VOUT=
12V, IOUT=21A
TESTED AT, VIN=200V, VOUT=7,5V, IOUT=20 A
PASSED 2x1011
protons/cm2 WITHOUT PROBLEMS.
CONCLUSION
VICOR DC-DC CONVERTERS ROBUST AT REDUCED
INPUT VOLTAGE – 200 V AND REDUCED OUTPUT POWER
PRICE OF EMU
LVS SECOND STAGE:
·
DC-DC CONVERTERS – 150 k$
(with filters and PCB)
·
MAGNETIC SHIELDING –70
k$
TOTAL 220
k$
TRANSFORMER
SOLUTION FOR SECOND STAGE:
400 Hz AC
FULL WAVE
RECTIFIER
PROVISORY COST
ESTIMATION:
·
100 kVA 400 Hz
MOTOR-GENERATOR - 30 k$
·
TRANSFORMERS - 15 k$
·
RECTIFIERS AND FILTERS - 8
k$
LINEAR VOLTAGE
REGULATORS
TESTS:
·
SPURIOUS CURRENT IN
LOW VOLTAGE CONDITION;
·
NOT OPERATING INHIBIT
FUNCTION
·
INSUFICIENT OUTPUT
CURRENT - IOUT £
2,6 A (INOM=3 A)
ALL
CONTROL FUNCTIONS IN FIRST STAGE
TEST – ON PSI
OPTIS BEAM with 1,25x109 protons/cm2s
of 64 MeV
GENERAL CONCLUSIONS
1.
The comparisons between three described variants of LVS
show clearly, that the best solution is the three-stage system with
transformers in second stage. Only reason to abandon it could be safety or
noise considerations.
2.
Further investigations and efforts needed for
application of this system in ECAL and TRACKER.
3.
Improvement of LHC4913 and further investigations of
commercial low voltage regulators are necessary for finding and optimal solution.