Degaussing Allard Schnabel PhysikalischTechnische Bundesanstalt November 2014 PTB

Degaussing Allard Schnabel Physikalisch-Technische Bundesanstalt November 2014 PTB 8. 22 Allard Schnabel page

Outline • Introduction • Why degaussing • Degaussing principle • Degaussing ABC • Practical rules for the degaussing coil design • Limitations • Outlook November 2014 PTB 8. 22 Allard Schnabel page 2

Introduction National Metrology Institute of Germany Headquarter: Braunschweig Historical site: Berlin Measurement of the human magnetic field SQUID measurements inside magnetically shielded rooms (MSR) BMSR-2 Since 2001 -2014 the magnetically best shielded room on earth November 2014 PTB 8. 22 Allard Schnabel page 3

Why degauss a magnetic shield? • to get a better shielding factor • for a lower residual field and field gradient inside the shield • time stable magnetic conditions especially when magnetic field are used inside November 2014 PTB 8. 22 Allard Schnabel page 4

Principle of AC degaussing example: sin function ; linear envelope function H= magnetic field strength (magnetizing field) B = magnetic flux density (magnetic induction) November 2014 PTB 8. 22 Allard Schnabel page 5

Degaussing ABC A start from saturation => delete memory/history B infinite number of cycles with decreasing amplitude => reach B=H=0 C no DC magnetic field offset during degaussing => no magnetic field stored November 2014 PTB 8. 22 Allard Schnabel page 6

A start from saturation problems • the saturation density of MU-metal is 0. 6 T => no proper degaussing with a coil at the surface => closed magnetic pathway; coil through the shield Holes through the shield for the degaussing cable(s)! November 2014 PTB 8. 22 Allard Schnabel page 7

A start from saturation problems i • length of magnetic pathway => try to have equal magnetic pathway lengths => do different sides sequentially ; tested at PTB ☺ => all layers or sequential degaussing ? • weldings, overlaps, holes and door? Note! You get a local 50: 50 distribution of the domains in the applied magnetic field direction This is not the ideal random distribution in 3 D! November 2014 PTB 8. 22 Allard Schnabel page 8

B infinite number of cycles with decreasing amplitude Any AC function with continuously decreasing amplitude will work☺ choose the fastest degaussing function that provides a reproducible time stable remanent field with the requested homogeneity parameter frequency envelope type number of cycles max current sin-function? November 2014 boundary condition(s) penetration depth size of transformer equal B steps; energy/heat small field homogeneity duration, bit resolution, noise reach saturation get all parts into saturation precise extreme values even harmonics PTB 8. 22 Allard Schnabel page 9

C no DC magnetic field offset during degaussing • no surrounding magnetic field => surrounding field compensation (SFC) • no magnetic parts inside/within the shield => test everything to be non-ferromagnetic • no DC-offset on top of the AC degaussing current => use transformer as high-pass filter => avoid even harmonics of the degaussing frequency PC DAC Attenuator Power Amplifier Transformer degaussing coil ZL critical test: interchange wires at ZL November 2014 PTB 8. 22 Allard Schnabel page 10

C no DC magnetic field offset during degaussing effect of even harmonics on top of the degaussing function 1. 2 a sin(x) 1. 0 b sin(2 x+90°) 0. 8 0. 6 0. 4 0. 2 0. 0 -30 0 30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570 600 630 660 690 720 -0. 2 -0. 4 -0. 6 -0. 8 -1. 0 -1. 2 November 2014 x in degree PTB 8. 22 Allard Schnabel page 11

C no DC magnetic field offset during degaussing effect of even harmonics on top of the degaussing function 1. 2 a sin(x) 1. 0 b sin(2 x+90°) 0. 8 a sin(x) + b sin(2 x+90°) 0. 6 0. 4 0. 2 0. 0 -30 0 30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570 600 630 660 690 720 -0. 2 -0. 4 -0. 6 -0. 8 -1. 0 -1. 2 November 2014 x in degree PTB 8. 22 Allard Schnabel page 12

Practical rules for the degaussing coil design • diameter of degaussing cable as large as possible => less windings => low inductance => low ohm resistance => less heat => low voltage at power amplifier output, but large current • symmetrically distributed turns of degaussing coil => less sensitive against degaussing uncertainties nearly no magnetic field change in the centre of the shield during degaussing November 2014 PTB 8. 22 Allard Schnabel page 13

known degaussing limitations • • homogeneity of µ and mechanical stress of the MU-panels thermal stability of the shield => air condition noise of amplifier and noise picked up from the surrounding thermal magnetic field effects during/after degaussing PC DAC Attenuator Power Amplifier degaussing coil ZL Transformer A R V Transformer November 2014 PTB 8. 22 Allard Schnabel degaussing coil ZL page 14

residual field limitations the “quasi static shielding factor” is not a “static shielding factor”! shielding factor of a magnetically shielded room 100, 000 10, 000 BMSR-2 S 1, 000 100, 000 10, 000 0. 01 0. 1 1 10 f in Hz November 2014 PTB 8. 22 Allard Schnabel page 15

residual field limitations • static field inside hollow sphere with large µr shell is approaching “ 0” field inside => S gets infinite! Þ If the shape is close to a sphere we will get “zero” magnetic field inside after “degaussing” independent of the outside field ☺ ☺ ☺ => Static SFC not as useful as expected reason: “degaussing factor” shield becomes magnetic and “cancels” magnetic field inside to “zero”. but! slow outside magnetic field change suppressed only with quasi static shielding factor! => Dynamic SFC is useful consequence: after an outside or inside magnetic field change the magnetic field inside is meta stable!!!! November 2014 PTB 8. 22 Allard Schnabel page 16

residual field limitations a = initial magnetization b = saturation hysteretic loop f => 0 Hz (quasi static) c = idealized “curve” ≈ commutation curve shield is “stable” only at points on the idealized curve d! “lowest-energy state” with magnetic fields outside or inside the shield, the shield should be “idealized” or “equilibrated” and not degaussed. During “idealization” or “equilibration” internal fields (B 0) and static SFC on! After “idealization” a dynamic SFC should keep the magnetic field seen by the shield constant from DC to some Hz. November 2014 PTB 8. 22 Allard Schnabel page 17

final limitations A volume of 0. 5 m x 0. 5 m with < 100 p. T It is a problem to find materials with < 5 p. T residual field in 3 cm distance. Any conducting surface produces Johnson-Noise (Eddy current noise) Sources: rf shield, mylar foil inside the dewar, plates of the magnetic shield. Vibrations and magnetic field gradient Flux density noise of a low noise dc-SQUID inside the Berlin Magnetically Shielded Room BMSR-2 for two different magnetic field gradients: a) 1. 5 p. T/mm b) 0. 5 p. T/mm any vibrating “magnetic” piece inside the shield creates AC noise November 2014 PTB 8. 22 Allard Schnabel page 18

outlook simulation of the Degaussing process Problems: • mathematical definition of the hysteretic curve • many cycles • thin large permalloy sheets – small gaps We are optimistic that we can answer some of the open questions! November 2014 PTB 8. 22 Allard Schnabel page 19