TF Coil Inductive Hipot Testing

Purpose

The April 17th, 2019, TF coil arc was believed to have been initiated by surface tracking across phenolic in the 1/8" gap between the center stack connections of TF Magnet C. It was deemed important that the cleaning and glyptal sealing intended to prevent further arcs should be validated with a turn-to-turn high-voltage test. Under normal operation, the full TF should have a maximum possible potential of 800V provided by the Robicon T1 and PEI power supplies in series.

It was necessary to use a high-frequency source to generate the high voltage needed for testing while minimizing the total energy available to dissipate in the event of breakdown. For this purpose, a capacitative surge tester was employed. With its 2μF capacitance and an administrative potential limit of 1kV, the maximum energy dissipated would be only 1J, compared with roughly 570kJ dissipated in the April arc due to the high current and long duration as the power supplies remained active until the preprogrammed voltage-control waveform extinguished.

Testing is governed by the LTX-β TF Coil Inductive Hipot Testing C-LTX-B-TP-TFHP-01 (updated July 10, 2019).

Testing History

Fig. 1: Example nominal TF ringdown from Test 1: 700V (left) and Test 3: 1kV #2 (right), showing voltage (black) and current (red) traces. The signature of the arc in Test 1 can be recognized in the faster initial voltage swing and short, very high spike of current.

Original data from TF inductive hipot tests is preserved in test-specific subdirectories within the /p/ltxdata/TFInductiveHipot/ directory, as well as any notes and plots generated from those tests.

Summer 2019

Test 1: July 15th

The first inductive hipot attempt was successful until an arc occurred at 700V (0.49J stored energy). The arc was faintly audible and clear on the voltage and current traces generated by the surge tester, but no arrangements were made to observe the arc visually at this time. The subdirectory for this test is test1/.

Test 2: July 16th

For the second test, the test cell lights were shut off for each discharge, and the web camera displays in the control room were watched for signs of arc light. This inductive hipot attempt was successful until an arc occurred at 900V (0.81J stored energy). The arc was faintly audible and clear on the voltage and current traces generated by the surge tester, but no arc light was observed, possibly due to limitations of the web cameras for observing such a low-energy discharge. The subdirectory for this test is test2/.

Test 3: July 23rd

For the third test, the test cell lights were shut off for each discharge, and additional personnel were in the test cell to watch for signs of arc light by eye and with better sight-lines. This inductive hipot attempt successfully reached 1kV and experienced no arcs after 10 consecutive 1kV (1J stored energy) discharges. The subdirectory for this test is test3/.

Aborted Test: August 1st

The first attempt at a post-glyptal test, during setup it was discovered that there was a 0.2V potential between the TF coil and ground. Investigation found that this was most likely due to the TF coil's DI cooling water, stagnant during the cleaning and painting process, changing its chemistry and effectively forming a weak electrochemical cell.

Test 4: August 2nd

The post-glyptal test successfully reached 1kV and experienced no arcs after 10 consecutive 1kV (1J stored energy) discharges. The subdirectory for this test is test4/.