SQUID Performance in a HV Environment Chen-Yu Liu Craig Huffer, Maciej Karcz, Josh Long Indiana University
Feb 05, 2016
SQUID Performance in a HV Environment
Chen-Yu LiuCraig Huffer, Maciej Karcz, Josh Long
Indiana University
Scenarios to study• HV Breakdown
– Induce HV from sparks (ESD)• could produce current exceeding the current limit of
the Josephson junctions, destroy SQUID (remedy: SQUID in a Faraday Cage)
– Induced current• could drive superconductor over its critical field, cause
flux trap, increase noise (no remedy required, might need to heat up SQUID periodically)
• Radio Frequency Interference (RFI)– minor: Increase the SQUID noise – moderate: flux jumps– serious: unable to lock SQUID – RF Source: micro-discharge, ground loop, switching mode
power supply– Remedy: SQUID in Faraday Cage, low pass filtered PS,
proper RF shield, proper ground
Experimental Setup
• Disk electrodes: 1.25” diameter, 0.25” thick
• Pb S.C. shield• HV feedthrough (ceramic)
rated for 20kV.• Star Cryoelectronic
magnetometer on chip
SQUID Noise Spectrum
• Star Cryoelectronics magnetometer prototype.– 8x8mm2 pick-up coil built in
on the SQUID chip.– 0.64 nT/0
– Intrinsic noise < 5/Hz
• no HV, SQUID sensor is placed in a faraday cage
(4 layers of Al coated mylar super-insulation)
• Measurements:– Noise ~ 30 0/Hz– S.C. Shielding should be
improved.– HV should also be better
shielded.
SQUID Noise Spectrumin HV environments
• Noise floor does not increase significantly with HV.
• Jumps add to 1/f noise and white noise. – E > 28 kV/cm (parallel
plates)
– E > 72 kV/cm (spherical HV electrode)
SQUID Response Under Large Current• To simulate a large current during breakdown
– Amplitude Modulated Sinusoidal Signal (1kHz) into a current loop (15.3 )
– Current loop is directly on top of the SQUID sensor
I=65 A→ 1.40
Observations
• Largest applied current– I=10Vpp/15.3 = 0.65A
• SQUID recovers to working condition right after the current is off.
• In nEDM system, assuming the discharge time is ~ micro-seconds, the spark current is about 23 A (~ 35 times bigger than the small system)– However, the SQUID is further away from the high
field region
SQUID ElectronicsD. Drung, Supercond. Sci. Technol. 16 (2003) 1320
Pickup coil
Input coil
Radio-Frequency Interference
• SQUID in flux lock mode (feedback circuit is on).
• Apply 50mVpp Sinusoidal Waveform into the current loop with 1k resistor in series.
• BW = 40kHz
1k
Radio-Frequency Interference
• SQUID in TUNE mode• Measure the amplitude of the
V- curve.• Apply 50mVpp Sinusoidal
waveform into the current loop with 1 k resistor in series.
1k
SQUID in FC
no FC
• Faraday cage – shields the high frequency
components.– Ensures the large V-
amplitude.
• f3dB~1MHz– Al thickness=85m– 4 layers of 0.0001 in =10 m
Micro-discharge vs Spark• Use a spherical HV electrode to ensure the breakdown
occurs in the field gap. (E up to 364 kV/cm)• Monitor the micro-discharge and spark currents
– Direct monitor on the ground electrode (through 1 in series).– Induced emf in the current loop.
> 4 0
Direct current
Induced emf
SQUID in SC shield~ 0.01 0
Frequency Spectrum of direct current measurement
• Major frequencies:– 30MHz, 85MHz, 145MHz
• Corresponds to – 6.6m, 2.35m, 1.37m
• System dimensions:– HV conductor: 0.66m– HV cable: 1.21m
• Due to impedance mismatch at various transitions.
Summary
• Destroyed one SQUID sensor in breakdown– Field = 15kV / 0.55mm = 273 kV/cm– Instantaneous spark current > 80A
• Micro-discharge– E> 7kV / 2.5mm = 28kV/cm (disk electrode)– E> 4kV / 0.55mm = 72kV/cm (spherical electrode)– I ~ 20 mA (4000 times smaller than the spark current)
• SQUID jumps – Increases the 1/f noise, corner ~ 200Hz– Starts at a lower field than the HV breakdown fields.
• Continuing study of effective RF shielding– Micro-discharge.– HV power supply (Glassman HV, series EH)
Current progress
• 3 squids : measured in a probe with a complete Pb can– Star Cryoelectronic magnetometer: 7.17 0/Hz– Quantum Design DC SQUID: 12.34 0/Hz– Supracon Blue2CE: 8.64 0/Hz
• Additional RF shielding (Al cage) around the high voltage input feedthrough.
• After the HV study in pressurized He, we are ready to carry out more RFI studies on these SQUID sensors.