1 Leakage Current – Summary Cosimo Cantini, Kevin Fusshoeller, Laura Molina Bueno
Reminder 2 • When the CAEN boards for the LEM were installed, the group from Kiev tested them and observed a non- zero current in several LEMs: LEM 3 down, LEM 4 down, LEM 6 down, LEM 8 up, LEM 12 down. • During the HV tests in June and July this current was again observed, both during ongoing tests and when everything was disconnected. More LEMs showed a non-zero current compared to the tests by Kiev group. • For periods with no tests, the current was identified as a non-constant offset of the CAEN power supply. 0 CAEN power supply: From left to right: Board 0, Board 1, Board 2 The LEM are ordered in logical order: first LEM 1 Up, then LEM 1 down, LEM 2 up, LEM 2 down etc. Board 0: LEM 1 up , … , LEM 4 down Board 1: LEM 4 up , … , LEM 8 down Board 2: LEM 9 up , … , LEM 12 down
3 Motivation • To investigate the source of this current, it was monitored over time with and without HV to understand if there is an intrinsic effect of the power supply or it is related to the HV system. • We would also like to understand if the HV tests have had an impact on the evolution of these currents over time.
4 Current in periods of no HV test: Offset Current
5 Observations Affected LEMs: LEM 1,3, 7,8,12 up and down LEM 2, 4, 6 down 1) The current is (anti-)correlated to the hall temperature. (plot for the period of 07.07.2017 – 09.07.2017)
Observations 6 2) The current is higher on some LEM (see table) and the current drops from Board 0 to Board 2. 0 1 2 Fan providing a uniform air flow
Observations 7 LEM 1 up LEM 1 down LEM 2 up LEM 2 down LEM 3 up LEM 3 down LEM 4 up LEM 4 down Board 0 Max. Current 0.3 nA 4.8 nA 0 nA 0.4 nA 1.4 nA 0.4 nA 0 nA 10.9 nA Rate of change 0.50 nA/K 2.15 nA/K / 0.5 nA/K 0.50 nA/K - 0.47 nA/K / -1.10 nA/K LEM 5 up LEM 5 down LEM 6 up LEM 6 down LEM 7 up LEM 7 down LEM 8 up LEM 8 down Board 1 Max. Current 0 nA 0 nA 0 nA (0.3 nA) 0.1 nA 0.3 nA 4.6 nA 3 nA R ate of change / / / / (< 0.2 nA/K) 0.50 nA/K -0.50 nA/K 1.25 nA/K LEM 9 up LEM 9 down LEM 10 up LEM 10 down LEM 11 up LEM 11 down LEM 12 up LEM 12 down Board 2 Max. Current 0 nA 0 nA 0 nA 0 nA 0 nA 0 nA 0.6 nA 2.8 nA Rate of change / / / / / / 0.24 nA/K 0.18 nA/K Since the board on the corner of the CAEN power supply is affected the most, this strongly suggests that the forced air flow is not homogeneous in the crate where the CAEN boards are placed. CAEN people assured us that variations of up to 100 ppm/K are still considered to be good values.
Observations 8 Affected LEMs: LEM 1,3, 7,8,12 up and down LEM 2, 4, 6 down 3) There is a shift of 1-3h in the (anti)correlation. We don‘t think this is a timestamp problem. More investigations are being performed. One idea is that the hall temperature is not the same as the local temperature at the boards. (plot for the period of 20.07.2017 – 23.07.2017) Max Max Min temp. Min temp. Max current Min temp. current current The current seems to predict the change in temperature.
Observations 9 Comment: because of the time shift, the current oscillates by up to 1.5 nA for the same temperature and depends on whether the temperature is rising or falling.
10 Current during HV test:
Observations 11 1) The slope of the current oscillations does not seems to be affected by the HV. However for some LEM the absolute value of the leakage current can change. For example we start seeing a current on LEM 6, which we were not seeing before. This current is anti-correlated to the temperature and behaves like the other leakage currents. • Test 1: Cathode: 20 kV, Grid: 1.0 kV, LEM down: 1.0, LEM up: 0 kV • Test 2: Cathode: 20 kV, Grid: 2.5 kV, LEM down: 3.0 (corner: 2.6 kV), LEM up: 0.2 kV • Test 3: Cathode: 20 kV, Grid: 0.5 kV, LEM down: 1.0, LEM up: 0 kV Slope Test 1 Slope Test 2 Slope Test 3 Slope no HV No obvious difference. LEM 3 down -0.40 nA/K -0.45 nA/K -0.47 nA/K 0.47 nA/K Difference of 0.1 nA. LEM 4 down -1.07 nA/K -1.20 nA/K -1.35 nA/K -1.10 nA/K LEM 6 down -0.14 nA/K -0.14 nA/K -0.13 nA/K / LEM 8 up -0.57 nA/K -0.39 nA/K -0.59 nA/K -0.50 nA/K No obvious difference. Current at hall temperature T = 298K Current Test 1 Current Test 2 Current Test 3 Current no HV No obvious difference. LEM 3 down 1.0 nA 0.9 nA 1.0 nA 1.0 nA Absolute current lower. LEM 4 down 5.25 nA 4.1 nA 5.3 nA 7.0 nA New current. LEM 6 down 0.4 nA 0.3 nA 0.2 nA / LEM 8 up 1.7 nA 1.7 nA 2.45 nA 1.75 nA No obvious difference.
12 Observations 2) The offset current changes suddenly when the HV is turned on. For LEM 12 it disappears completely as long as the HV system is on. For the other LEMs the current drops by at least 50% when the system is turned on. Abrupt change during start and end of HV test. No voltages applied No voltages applied
13 Observations Sudden drop when we turn on the HV system.
14 Observations 3) The leakage current does not seem to be affected by trips of the grid or of the LEMs. Leakage current behaves the same before and after trips.
15 Observations Short drop of the current when the LEM trips, but then it recovers to the value before the trip and behaves normally.
16 Observations 4) Comparison between number of trips on LEM and which LEM show a leakage current. CRP3 CRP2 LEM Up 4 trips 4 trips 4 trips 5 trips 4 trips 4 trips LEM Down 8 trips 12 trips 13 trips 20 trips 12 trips 10 trips N S CRP5 CRP1 8 trips 4 trips 5 trips 7 trips 4 trips 4 trips 19 trips 12 trips 13 trips 23 trips 10 trips 7 trips CRP8 CRP7 CRP6 • The following LEMs always show leakage current during HV tests: LEM 3 down, LEM 4 down, LEM 6 down, LEM 8 up. • The highest leakage current is found on LEM 4 down and LEM 8 up. • The other LEMs only show a leakage current at high hall temperatures (>29°C). • Data not conclusive enough to say if LEMs with leakage current trip more often.
Conclusion 17 • There is a non-constant offset of the power supply (anti-)correlated to the hall temperature. This correlation is there whether the HV system is on or off and stronger on the outside board. This strongly suggests that it is due to a non uniform air flow in the crate. Moreover CAEN people assure us that the variations we see are still considered to be good values. • Turning on the HV influences the offset of the leakage current. For most of the LEM the current drops by at least 50%, but can recover later on. For LEM 12 it disappears completely, while for LEM 6 we see a current we did not see before. • The leakage current behaves the same before and after trips in the grid or the LEMs. Future plans/tests • Disconnect the cables and put some voltage on the board to check if the LEMs/cables are causing problems. • See if we can improve the forced air flow in the crate. • Install parallel resistors to each channel to monitor such variations of current.
18 Backup
19 Position of temperature sensors
20 LEM 12: Current and Temperature Clear correlation between hall temperature and LEM (offset) current. From 03.07.2017 to 17.07.2017
21
22
23
24
25
26
Recommend
More recommend
