Peter Barnes, LLNL
August 29, 2000
TPC High Voltage Tests
Abstract
While the cathode plane was installed at Fermilab during 25 July – 3 August 2000 we ramped the cathode and anode voltages to their operating points. This note documents the test procedure and results.
Contents
Introduction
Gas Flow
Ground Connections
Cathode Plane High Voltage Tests
Anode Wires High Voltage Tests
Conclusion
Contents / Introduction / Gas FlowIntroduction
During the period 25 July – 3 August 2000, after we installed the cathode plane, we applied high voltage to the cathode plane and anode wires, and measured the current. This note documents the high voltage test procedure and results.
Contents / Introduction / Gas Flow / Cathode PlaneGas Flow
While we were installing the cathode plane (documented elsewhere), Fermilab technicians established a flow of dry nitrogen through the TPC. Flow was initiated at ~4 PM on 1 August 2000. The flow path is illustrated below.
Dry nitrogen gas from a bottle flowed at ~1 psig through a flow meter into the TPC. The vent from the TPC ran to a mineral oil bubbler. The TPC operates at ambient pressure, that is to say, the pressure inside the cleanroom. To accommodate the cleanroom overpressure with respect to the outside air, the bubbler was filled (initially) with ~1.2 cm of oil above the exhaust tube.
The connections to the TPC can be seen in the following picture. The gas inlet tube is the copper one which is being connected with the wrenches. The exhaust is the black plastic tube, which is connected to the two exhaust ports, one of which is visible in the upper corner of the TPC.
Initially, the regulator was set so the flow meter indicated 1.5 cfh (cubic feet per hour) = 0.04 m3/h. The bubbler tube extended ~12 mm into the mineral oil. Despite the continuous flow, we saw no bubbles in the bubbler, suggesting that the gas was venting upstream of the bubbler. The liquid level in the entry tube was below the free liquid level, indicating a slight over pressure, as expected, but evidently not enough to bubble. We also heard periodic "pings" from the TPC, though we could see no indications of dynamic activity, by watching for flex in the side panels, for example.
At this point, because of a shortage of screws of the correct length, only ~1/3 of the lid screws were installed. The lid is approximately 1 x 2 m2. With as little as 1 psig pressure difference, the force is ~3200 pounds. Guessing that this could be enough to open the O-ring seal, we filled in the remaining screw holes in the lid with long screws with washers. This caused the liquid level in the bubbler tube to drop, but still not enough to bubble.
We left the TPC and gas system in this state for several hours while we went to dinner. When we returned ~10 PM, we decreased the quantity of mineral oil in the bubbler. When the level reached ~8 mm, we observed bubbling at a rate of 2-3 Hz. This did not change the flow rate indicated on the flow meter. We left the system in this state over night.
By the time we turned on the high voltage on the morning of 2 August 2000, nitrogen had been flowing for 18 hours. The TPC volume is approximately 1 x 1 x 2 m3, so at 1.5 cfh, it requires ~47 h per volume change. Therefore, we had change ~1/3 of the volume, i.e., the gas was likely 14% O2, 84% N2, and 60% relative humidity (assuming that the ambient air was 90% RH—it was very humid that week).
Contents / Gas Flow / Ground Connection / Cathode TestsGround Connection
We connected a 1" wide ground braid from one of the copper ground lugs on the TPC to the High Voltage Rack.
Contents / Ground Connection / Cathode Tests / Anode TestsCathode Plane High Voltage Tests
We used the Bertan "Field Cage Supply" to apply high voltage to the field cage input of the TPC.
Both the supply and the TPC connector mate to a Reynolds 167-3516 bayonet connector. We used a single RG 213/U cable. We set the power supply high voltage limit to 10,100 V and the current limit to 80 A.
Using the direct control mode, Gulshan started ramping the voltage. He used the control knob to increase the voltage, while monitoring the current on the meter. We observed a steady increase in the current with voltage. During puases in the ramp we saw no evidence of current noise or drift. At ~5000 V the current limit was exceeded and the supply tripped off. He tried again and reached ~8000 V. Because the Bertan dial is rate sensitive, it is easy to twitch it while turning and introduce a large voltage step, leading to a momentary large current pulse as the TPC capacitance charges up. We believed this was the cause of the over-current trips, though we did not see large current excursions on the display. We increased the current limit to 120 A, but this did not seem to help. Finally, by very carefully turning the knob for 10 minutes, Raja was able to reach 10,008 V.
The following plot shows the current and resistance as a function of voltage for all measurements we recorded during all the ramps. Ignoring the first point, the average of all the resistance measurements is 124.2 ± 0.3 M.
Contents / Cathode Tests / Anode Tests / ConclusionAnode Wires High Voltage Tests
To power the anode sections we used two dual channel (four channels total) NIM high voltage power supplies, Bertan Model 362 (0–2000V). These have SHV outputs, matching the SHV anode connections on the TPC. We fed each output through a channel of the "Anode Current Monitor," then on to the TPC.
It appeared throughout the testing that several channels of the current monitor were (or became) unresponsive, so we moved the cabling around several times, especially for the first set. The calibration of the current monitors is unknown.
We connected four anode sections at a time, ramping the voltage to 1300 V, but not exceeding ~500 A indicated on any section. Most anode sections showed a slow decay of the current when we left the voltage at a fixed setting. Since the time constant appeared to be many minutes, we did not wait for the currents to reach steady state values. Because of some initial confusion about where to read the current and what the maximum voltage should be, anode sections 0–3 may have been energized to as high as 2000 V for a few minutes.
The tables below show the applied voltage and indicated current for each anode section, as well as the power supply and meter channel used.
Anode Section / 0 / 1 / 2 / 3HV Channel / 0 / 1 / 2 / 3
Current Monitor / (5)* 10 / 8 / 2 / 3
Voltage (V) / Current (nA)
100 / (1)* / 2 / < 1 / < 1
200 / (2000)* / 8 / < 1 / < 1
300 / 100 / 10 / 20 / 20
400 / 200 / 12 / 10 / 10
500 / 120 / 14 / 80 / 20
600 / 120 / 14 / 90 / 20
700 / 130 / 14 / 90 / 20
800 / 200 / 20 / 100 / 80
900 / 500 / 40 / 200 / 80
1000 / 50 / 100 / 80
1100 / 80 / 100 / 100
1200 / 90 / 100 / 100
1300 / 100 / 100 / 130
*The meter response on monitor channel 5 seemed unreliable, so we switched to monitor 10.
Anode Section / 4 / 5 / 6 / 7HV Channel / 0 / 1 / 3 / 2
Current Monitor / 10 / 8 / 3 / 2
Voltage (V) / Current (nA)
100 / 80 / 20 / 20 / 80
200 / 100 / 80 / 80 / 100
300 / 100 / 90 / 80 / 100
400 / 120 / 80 / 90 / 100
500 / 200 / 90 / 100 / 110
600 / 200 / 90 / 100 / 100
700 / 150 / 90 / 90 / 100
800 / 200 / 90 / 90 / 100
900 / 150 / 80 / 90 / 100
1000 / 120 / 80 / 80 / 110
1100 / 150 / 90 / 90 / 120
1200 / 120 / 90 / 120 / 120
1300 / 200 / 80 / 200 / 150
Anode Section / 8 / 9 / 10 / 11
HV Channel / 0 / 1 / 3 / 2
Current Monitor / 10 / 8 / 3 / 2
Voltage (V) / Current (nA)
100 / 90 / 30 / 12 / 90
200 / 100 / 90 / 60 / 100
300 / 120 / 90 / 90 / 110
400 / 130 / 100 / 90 / 120
500 / 120 / 100 / 90 / 110
600 / 110 / 100 / 90 / 110
700 / 110 / 100 / 90 / 110
800 / 120 / 90 / 90 / 120
900 / 120 / 100 / 90 / 130
1000 / 120 / 100 / 90 / 130
1100 / 110 / 100 / 80 / 200
1200 / 120 / 100 / 80 / 400
1300 / 110 / 110 / 80 / 500
Anode Section / 12 / 13 / 14 / 15
HV Channel / 0 / 1 / 3 / 2
Current Monitor / 10 / 12 / 3 / 2
Voltage (V) / Current (nA)
100 / 60 / 100 / 8 / 10
200 / 100 / 120 / 80 / 100
300 / 110 / 130 / 80 / 120
400 / 120 / 200 / 80 / 120
500 / 120 / 400 / 90 / 120
600 / 110 / 500 / 90 / 140
700 / 110 / 400 / 80 / 300
800 / 100 / 300 / 90 / 300
900 / 180 / 500 / 100 / 300
1000 / 200 / 500 / 100 / 800
1100 / 200 / 500 / 100
1200 / 200 / 800 / 120
1300 / 200 / 120
Contents / Anode Tests / Conclusion
Conclusion
We successfully brought the cathode plane to 10 kV and almost all the anode sections to 1300 V, the nominal operating points. There did not appear to be any dead shorts; the minimum effective anode resistance was 2 G. Four of the anode sections exceeded 200 nA at or below 1300 V. This was likely due to errors in the current measurement or the dirty gas composition.
Contents / Conclusion