Lecture 10: Cryogenic Systems
Part III: Systems Integration — SCE Futures
Learning Objectives¶
By the end of this lecture, you will be able to:
- Explain why superconducting electronics operate at 4K and the properties of liquid helium
- Compare wet (LHe) vs dry (cryocooler) cryogenic systems and select appropriate systems
- Design helium recovery systems and understand supply chain considerations
- Calculate cooling power budgets and thermal loads
- Identify cryogenic hazards and implement appropriate safety measures
- Set up and operate a cryogenic test lab
In [1]:
Copied!
# Setup
import matplotlib.pyplot as plt
import matplotlib.patches as patches
import numpy as np
COLORS = {
'primary': '#2196F3',
'secondary': '#FF9800',
'success': '#4CAF50',
'danger': '#f44336',
'dark': '#1a1a2e',
'light': '#f5f5f5',
'purple': '#9C27B0',
'cyan': '#00BCD4',
'cold': '#00BCD4',
'warm': '#FF5722',
}
plt.rcParams['figure.facecolor'] = 'white'
plt.rcParams['axes.facecolor'] = 'white'
plt.rcParams['font.size'] = 11
print("Setup complete.")
# Setup
import matplotlib.pyplot as plt
import matplotlib.patches as patches
import numpy as np
COLORS = {
'primary': '#2196F3',
'secondary': '#FF9800',
'success': '#4CAF50',
'danger': '#f44336',
'dark': '#1a1a2e',
'light': '#f5f5f5',
'purple': '#9C27B0',
'cyan': '#00BCD4',
'cold': '#00BCD4',
'warm': '#FF5722',
}
plt.rcParams['figure.facecolor'] = 'white'
plt.rcParams['axes.facecolor'] = 'white'
plt.rcParams['font.size'] = 11
print("Setup complete.")
Setup complete.
1. Why 4 Kelvin?¶
Superconducting electronics require temperatures below the critical temperature (Tc) of the superconducting materials used.
Critical Temperatures of Common Materials¶
| Material | Tc (K) | Typical Operating Temp | Notes |
|---|---|---|---|
| Niobium (Nb) | 9.2 | 4.2 K | Workhorse material, excellent properties |
| NbN | 16 | 4-10 K | Higher Tc, used in some detectors |
| NbTiN | 14-15 | 4-10 K | Good for kinetic inductance devices |
| Al | 1.2 | 0.1 K (mK) | Used in qubits, requires dilution fridge |
| YBCO | 93 | 77 K | High-Tc, but hard to make JJs |
Why Not Just Below Tc?¶
Operating well below Tc provides:
- Larger energy gap: Reduces thermal noise and quasiparticle generation
- Margin for heating: Chip dissipation raises local temperature
- Reproducibility: Properties are more stable far from the transition
- Convenience: 4.2 K is the boiling point of liquid helium at 1 atm
The 4.2 K Sweet Spot¶
4.2 K is the boiling point of liquid helium at atmospheric pressure, making it a natural operating point:
- Self-regulating: LHe bath stays at 4.2 K as long as liquid remains
- Large installed base: Decades of cryogenic infrastructure
- Reasonable cooling power: 1-2 W available from GM/PT cryocoolers
- Compatible with Nb: Well below Tc = 9.2 K
In [2]:
Copied!
# Visualize: Temperature scale and superconductor operating regions
fig, ax = plt.subplots(figsize=(14, 6))
# Temperature scale (log)
temps = np.logspace(-2, 2.5, 1000)
# Draw temperature bar
for i, t in enumerate(temps[:-1]):
color = plt.cm.coolwarm(np.log10(t) / 4 + 0.5)
ax.axvspan(t, temps[i+1], alpha=0.3, color=color)
# Mark key temperatures
markers = {
'Dilution fridge\n(10 mK)': 0.01,
'Pumped He\n(1.5 K)': 1.5,
'LHe\n(4.2 K)': 4.2,
'LN2\n(77 K)': 77,
'Room temp\n(300 K)': 300,
}
for label, temp in markers.items():
ax.axvline(temp, color='black', linewidth=2, alpha=0.7)
ax.text(temp, 1.15, label, ha='center', fontsize=9, rotation=0)
# Material Tc ranges
materials = [
('Al (Tc=1.2K)', 0.05, 1.2, COLORS['purple']),
('Nb (Tc=9.2K)', 2, 9.2, COLORS['primary']),
('NbN (Tc=16K)', 4, 16, COLORS['success']),
('YBCO (Tc=93K)', 40, 93, COLORS['secondary']),
]
for i, (name, t_op, tc, color) in enumerate(materials):
y = 0.3 + i * 0.15
ax.fill_betweenx([y-0.05, y+0.05], t_op, tc, color=color, alpha=0.7)
ax.plot([tc, tc], [y-0.05, y+0.05], 'k-', linewidth=2)
ax.text(t_op * 0.7, y, name, ha='right', va='center', fontsize=9)
ax.set_xscale('log')
ax.set_xlim(0.005, 500)
ax.set_ylim(0, 1.3)
ax.set_xlabel('Temperature (K)', fontsize=12)
ax.set_yticks([])
ax.set_title('Cryogenic Temperature Scale & Superconductor Operating Ranges', fontsize=14, fontweight='bold')
# Highlight SCE operating region
ax.axvspan(3, 5, alpha=0.2, color=COLORS['cold'], zorder=0)
ax.text(4, 0.1, 'SCE\noperating\nregion', ha='center', fontsize=10, fontweight='bold')
plt.tight_layout()
plt.show()
# Visualize: Temperature scale and superconductor operating regions
fig, ax = plt.subplots(figsize=(14, 6))
# Temperature scale (log)
temps = np.logspace(-2, 2.5, 1000)
# Draw temperature bar
for i, t in enumerate(temps[:-1]):
color = plt.cm.coolwarm(np.log10(t) / 4 + 0.5)
ax.axvspan(t, temps[i+1], alpha=0.3, color=color)
# Mark key temperatures
markers = {
'Dilution fridge\n(10 mK)': 0.01,
'Pumped He\n(1.5 K)': 1.5,
'LHe\n(4.2 K)': 4.2,
'LN2\n(77 K)': 77,
'Room temp\n(300 K)': 300,
}
for label, temp in markers.items():
ax.axvline(temp, color='black', linewidth=2, alpha=0.7)
ax.text(temp, 1.15, label, ha='center', fontsize=9, rotation=0)
# Material Tc ranges
materials = [
('Al (Tc=1.2K)', 0.05, 1.2, COLORS['purple']),
('Nb (Tc=9.2K)', 2, 9.2, COLORS['primary']),
('NbN (Tc=16K)', 4, 16, COLORS['success']),
('YBCO (Tc=93K)', 40, 93, COLORS['secondary']),
]
for i, (name, t_op, tc, color) in enumerate(materials):
y = 0.3 + i * 0.15
ax.fill_betweenx([y-0.05, y+0.05], t_op, tc, color=color, alpha=0.7)
ax.plot([tc, tc], [y-0.05, y+0.05], 'k-', linewidth=2)
ax.text(t_op * 0.7, y, name, ha='right', va='center', fontsize=9)
ax.set_xscale('log')
ax.set_xlim(0.005, 500)
ax.set_ylim(0, 1.3)
ax.set_xlabel('Temperature (K)', fontsize=12)
ax.set_yticks([])
ax.set_title('Cryogenic Temperature Scale & Superconductor Operating Ranges', fontsize=14, fontweight='bold')
# Highlight SCE operating region
ax.axvspan(3, 5, alpha=0.2, color=COLORS['cold'], zorder=0)
ax.text(4, 0.1, 'SCE\noperating\nregion', ha='center', fontsize=10, fontweight='bold')
plt.tight_layout()
plt.show()
2. Liquid Helium Basics¶
Properties of Liquid Helium¶
| Property | Value | Implication |
|---|---|---|
| Boiling point (1 atm) | 4.2 K | Natural operating temperature |
| Latent heat | 2.6 kJ/L | Low cooling capacity per liter |
| Density | 125 kg/m³ | Light, easy to transfer |
| Cost | $15-30/L (2024) | Significant operating expense |
| Viscosity | Very low | Leaks through tiny gaps |
| Expansion ratio | 1:750 | 1L liquid → 750L gas |
Comparison: LHe vs LN2¶
| Property | LHe | LN2 |
|---|---|---|
| Boiling point | 4.2 K | 77 K |
| Latent heat | 2.6 kJ/L | 161 kJ/L |
| Cost | $15-30/L | $0.50-2/L |
| Availability | Limited | Abundant |
LN2 has 60× more cooling capacity per liter! This is why LN2 shields are used to reduce LHe consumption.
Helium Supply Chain¶
Helium is extracted from natural gas and is a non-renewable resource:
- Sources: US (Federal Helium Reserve depleting), Qatar, Algeria, Russia, Australia
- Supply volatility: Prices have spiked 3-5× during shortages (2012-2013, 2019)
- Grade matters: Research grade (99.999%) costs more than industrial
- Recovery is essential: Modern labs must recover and reliquefy
Dewar Types¶
| Type | Capacity | Hold Time | Use Case |
|---|---|---|---|
| Transport dewar | 30-60 L | 1-3 days | Moving LHe between buildings |
| Storage dewar | 100-500 L | 1-2 weeks | Lab storage |
| Research dewar | 50-100 L | Varies | Dip probe experiments |
3. Cryostat Types¶
Wet Cryostats (LHe Bath)¶
How it works: Sample immersed directly in liquid helium bath
Advantages:
- Simple, reliable temperature control
- High cooling power (limited by boil-off rate)
- Self-regulating at 4.2 K
- Fast cooldown (minutes to hours)
Disadvantages:
- Ongoing LHe consumption ($$$)
- Requires helium infrastructure
- Regular refills needed
- Vibration from boiling
Best for: R&D, university labs, quick experiments, high-power devices
Dry Cryostats (Cryocooler-Based)¶
How it works: Closed-cycle refrigerator cools sample through thermal link
Types of cryocoolers:
| Type | Base Temp | Power at 4K | Vibration | Cost |
|---|---|---|---|---|
| Gifford-McMahon (GM) | 2.5-4 K | 1-2 W | Moderate | $50-100K |
| Pulse Tube (PT) | 2.5-4 K | 0.5-1.5 W | Low | $80-150K |
| Stirling | 20-80 K | 5-50 W | High | $20-50K |
Advantages:
- No LHe consumption (just electricity + maintenance)
- Unattended operation for weeks/months
- Better for production/deployment
- Predictable operating costs
Disadvantages:
- Limited cooling power (~1-2 W at 4K)
- Longer cooldown (12-24 hours typical)
- Mechanical vibration (especially GM)
- Higher upfront cost
- Compressor maintenance (every 10-20k hours)
Best for: Production testing, deployed systems, long-running experiments
Dilution Refrigerators¶
For temperatures below 1K (needed for qubits, not typical SCE):
- Uses ³He/⁴He mixture
- Base temperature: 10-50 mK
- Cost: $500K-1M+
- Mostly for quantum computing applications
In [3]:
Copied!
# Visualize: Comparison of wet vs dry cryostat architectures
fig, axes = plt.subplots(1, 2, figsize=(14, 8))
# Left: Wet cryostat (LHe dewar with dip probe)
ax = axes[0]
# Outer vacuum jacket
outer = patches.FancyBboxPatch((1, 0.5), 4, 7, boxstyle="round,pad=0.1",
facecolor=COLORS['light'], edgecolor='black', linewidth=2)
ax.add_patch(outer)
# LN2 shield
ln2 = patches.FancyBboxPatch((1.3, 1), 3.4, 5.5, boxstyle="round,pad=0.05",
facecolor='#E3F2FD', edgecolor='blue', linewidth=1.5, linestyle='--')
ax.add_patch(ln2)
ax.text(4.5, 3.5, 'LN2\nShield\n(77K)', fontsize=9, ha='left')
# LHe bath
lhe = patches.FancyBboxPatch((1.6, 1.5), 2.8, 4, boxstyle="round,pad=0.02",
facecolor=COLORS['cold'], alpha=0.5, edgecolor='black', linewidth=1.5)
ax.add_patch(lhe)
ax.text(3, 3.5, 'LHe Bath\n(4.2 K)', fontsize=10, ha='center', fontweight='bold')
# Sample
sample = patches.Rectangle((2.5, 2), 1, 0.8, facecolor=COLORS['secondary'], edgecolor='black', linewidth=2)
ax.add_patch(sample)
ax.text(3, 2.4, 'DUT', fontsize=9, ha='center', fontweight='bold')
# Dip probe
ax.plot([3, 3], [2.8, 8], color='gray', linewidth=4)
ax.text(3, 8.2, 'Dip Probe\n(wiring inside)', fontsize=10, ha='center')
ax.set_xlim(0, 6)
ax.set_ylim(0, 9)
ax.set_aspect('equal')
ax.axis('off')
ax.set_title('Wet Cryostat (LHe Bath)', fontsize=12, fontweight='bold')
# Right: Dry cryostat (cryocooler)
ax = axes[1]
# Vacuum chamber
chamber = patches.FancyBboxPatch((1, 0.5), 4, 7, boxstyle="round,pad=0.1",
facecolor=COLORS['light'], edgecolor='black', linewidth=2)
ax.add_patch(chamber)
# 40K stage
stage40k = patches.FancyBboxPatch((1.5, 4), 3, 1.5, boxstyle="round,pad=0.02",
facecolor='#FFECB3', edgecolor='black', linewidth=1.5)
ax.add_patch(stage40k)
ax.text(3, 4.75, '40 K Stage\n(30-50 W)', fontsize=10, ha='center')
# 4K stage
stage4k = patches.FancyBboxPatch((1.8, 1.5), 2.4, 2, boxstyle="round,pad=0.02",
facecolor=COLORS['cold'], alpha=0.5, edgecolor='black', linewidth=1.5)
ax.add_patch(stage4k)
ax.text(3, 2.5, '4 K Stage\n(1-2 W)', fontsize=10, ha='center', fontweight='bold')
# Sample
sample2 = patches.Rectangle((2.5, 1.7), 1, 0.6, facecolor=COLORS['secondary'], edgecolor='black', linewidth=2)
ax.add_patch(sample2)
ax.text(3, 2, 'DUT', fontsize=9, ha='center', fontweight='bold')
# Cryocooler cold head
ax.plot([3, 3], [5.5, 8], color='gray', linewidth=6)
coldhead = patches.FancyBboxPatch((2, 7.5), 2, 1, boxstyle="round,pad=0.02",
facecolor=COLORS['purple'], alpha=0.7, edgecolor='black', linewidth=2)
ax.add_patch(coldhead)
ax.text(3, 8, 'Cold Head', fontsize=10, ha='center', color='white', fontweight='bold')
# Compressor (external)
ax.text(5.5, 8, '\u2190 To\nCompressor', fontsize=10, ha='left')
ax.set_xlim(0, 6)
ax.set_ylim(0, 9)
ax.set_aspect('equal')
ax.axis('off')
ax.set_title('Dry Cryostat (Cryocooler)', fontsize=12, fontweight='bold')
plt.tight_layout()
plt.show()
print("Wet systems: Simple, high cooling power, but consume LHe")
print("Dry systems: No LHe, but limited cooling power (~1-2W at 4K)")
# Visualize: Comparison of wet vs dry cryostat architectures
fig, axes = plt.subplots(1, 2, figsize=(14, 8))
# Left: Wet cryostat (LHe dewar with dip probe)
ax = axes[0]
# Outer vacuum jacket
outer = patches.FancyBboxPatch((1, 0.5), 4, 7, boxstyle="round,pad=0.1",
facecolor=COLORS['light'], edgecolor='black', linewidth=2)
ax.add_patch(outer)
# LN2 shield
ln2 = patches.FancyBboxPatch((1.3, 1), 3.4, 5.5, boxstyle="round,pad=0.05",
facecolor='#E3F2FD', edgecolor='blue', linewidth=1.5, linestyle='--')
ax.add_patch(ln2)
ax.text(4.5, 3.5, 'LN2\nShield\n(77K)', fontsize=9, ha='left')
# LHe bath
lhe = patches.FancyBboxPatch((1.6, 1.5), 2.8, 4, boxstyle="round,pad=0.02",
facecolor=COLORS['cold'], alpha=0.5, edgecolor='black', linewidth=1.5)
ax.add_patch(lhe)
ax.text(3, 3.5, 'LHe Bath\n(4.2 K)', fontsize=10, ha='center', fontweight='bold')
# Sample
sample = patches.Rectangle((2.5, 2), 1, 0.8, facecolor=COLORS['secondary'], edgecolor='black', linewidth=2)
ax.add_patch(sample)
ax.text(3, 2.4, 'DUT', fontsize=9, ha='center', fontweight='bold')
# Dip probe
ax.plot([3, 3], [2.8, 8], color='gray', linewidth=4)
ax.text(3, 8.2, 'Dip Probe\n(wiring inside)', fontsize=10, ha='center')
ax.set_xlim(0, 6)
ax.set_ylim(0, 9)
ax.set_aspect('equal')
ax.axis('off')
ax.set_title('Wet Cryostat (LHe Bath)', fontsize=12, fontweight='bold')
# Right: Dry cryostat (cryocooler)
ax = axes[1]
# Vacuum chamber
chamber = patches.FancyBboxPatch((1, 0.5), 4, 7, boxstyle="round,pad=0.1",
facecolor=COLORS['light'], edgecolor='black', linewidth=2)
ax.add_patch(chamber)
# 40K stage
stage40k = patches.FancyBboxPatch((1.5, 4), 3, 1.5, boxstyle="round,pad=0.02",
facecolor='#FFECB3', edgecolor='black', linewidth=1.5)
ax.add_patch(stage40k)
ax.text(3, 4.75, '40 K Stage\n(30-50 W)', fontsize=10, ha='center')
# 4K stage
stage4k = patches.FancyBboxPatch((1.8, 1.5), 2.4, 2, boxstyle="round,pad=0.02",
facecolor=COLORS['cold'], alpha=0.5, edgecolor='black', linewidth=1.5)
ax.add_patch(stage4k)
ax.text(3, 2.5, '4 K Stage\n(1-2 W)', fontsize=10, ha='center', fontweight='bold')
# Sample
sample2 = patches.Rectangle((2.5, 1.7), 1, 0.6, facecolor=COLORS['secondary'], edgecolor='black', linewidth=2)
ax.add_patch(sample2)
ax.text(3, 2, 'DUT', fontsize=9, ha='center', fontweight='bold')
# Cryocooler cold head
ax.plot([3, 3], [5.5, 8], color='gray', linewidth=6)
coldhead = patches.FancyBboxPatch((2, 7.5), 2, 1, boxstyle="round,pad=0.02",
facecolor=COLORS['purple'], alpha=0.7, edgecolor='black', linewidth=2)
ax.add_patch(coldhead)
ax.text(3, 8, 'Cold Head', fontsize=10, ha='center', color='white', fontweight='bold')
# Compressor (external)
ax.text(5.5, 8, '\u2190 To\nCompressor', fontsize=10, ha='left')
ax.set_xlim(0, 6)
ax.set_ylim(0, 9)
ax.set_aspect('equal')
ax.axis('off')
ax.set_title('Dry Cryostat (Cryocooler)', fontsize=12, fontweight='bold')
plt.tight_layout()
plt.show()
print("Wet systems: Simple, high cooling power, but consume LHe")
print("Dry systems: No LHe, but limited cooling power (~1-2W at 4K)")
Wet systems: Simple, high cooling power, but consume LHe Dry systems: No LHe, but limited cooling power (~1-2W at 4K)
4. Helium Management & Recovery¶
Why Recovery Matters¶
At $20-30/L and typical consumption of 50-200 L/week in an active lab:
- Without recovery: $50K-300K/year in helium alone
- With recovery: $5K-30K/year (90%+ recovery rate)
Recovery pays for itself within 1-2 years.
Recovery System Components¶
Dewars/Cryostats → Collection bags → Recovery compressor →
Purifier → High-pressure storage → Liquefier → LHe storage
- Collection bags/bladders: Capture boil-off gas from dewars and cryostats
- Recovery compressor: Compress gas into high-pressure cylinders (200-300 bar)
- Purifier: Remove air and moisture contamination
- Liquefier: Convert gas back to LHe (optional, expensive)
- Storage: High-pressure gas cylinders or liquid dewars
Recovery Architectures¶
| Approach | Upfront Cost | Operating Cost | Best For |
|---|---|---|---|
| Bag + external liquefier | $10-50K | Pay per liter | Small labs |
| On-site liquefier | $500K-1M | Electricity only | Large labs |
| Closed-cycle (dry) | $100-300K | Maintenance | Production |
Contamination¶
Helium gas can become contaminated with:
- Air: From leaks in collection system (>1 ppm causes problems)
- Water vapor: Condenses and freezes, blocking lines
- Pump oil: From older recovery compressors
Most liquefiers require <1 ppm impurities; contaminated gas must be purified or discarded.
5. Cooling Power Budgets¶
Designing a cryogenic system requires careful thermal budgeting.
Heat Load Sources¶
| Source | Typical Value | Notes |
|---|---|---|
| Radiation (300K→4K) | 50 mW/m² | With MLI shielding |
| Conduction (wiring) | 1-10 mW/wire | Depends on material, length |
| Conduction (supports) | 10-100 mW | Inter-stage; typically included in cryocooler specs |
| Chip dissipation | 1-100 mW | Depends on circuit type |
| Amplifiers (HEMT) | 5-15 mW each | If used; major heat source |
| Amplifier bias power | ~10 mW typical | ~50 µA × 50 Ω × number of amplifiers |
Heat Intercept Strategy¶
Use intermediate temperature stages to intercept heat before it reaches 4K:
300 K (room temp)
│
├── Radiation shields (MLI)
│
40-50 K (first stage) ← Intercept most conduction heat here
│ 30-50 W available
├── Thermal anchoring of wires
│
4 K (second stage) ← Only residual heat reaches here
1-2 W available
Wire Heat Load Calculation¶
Heat conducted along a wire:
$$Q = \frac{A}{L} \int_{T_{cold}}^{T_{hot}} k(T) \, dT$$
Use the thermal conductivity integral for common materials:
| Material | ∫k dT (300→4K) | ∫k dT (40→4K) | Use |
|---|---|---|---|
| Copper (OFHC) | 15,000 W/m | 800 W/m | Short runs, 4K only |
| Phosphor bronze | 1,500 W/m | 80 W/m | General wiring |
| Stainless steel | 300 W/m | 15 W/m | Structural supports |
| Manganin | 400 W/m | 20 W/m | DC wiring |
| NbTi (SC below 9K) | 150 W/m | ~0 W/m | Best for 4K stage |
In [4]:
Copied!
# Calculate: Example thermal budget for a cryocooler system
def wire_heat_load(n_wires, awg, length_m, material='phosphor_bronze', T_hot=40, T_cold=4):
"""
Calculate heat load from wires.
Using simplified thermal conductivity integrals.
"""
# Wire diameters (mm) for common AWG
awg_diameters = {36: 0.127, 32: 0.202, 28: 0.321, 24: 0.511}
# Thermal conductivity integrals (W/m) from T_hot to 4K
k_integrals = {
'copper': {'300_4': 15000, '40_4': 800},
'phosphor_bronze': {'300_4': 1500, '40_4': 80},
'manganin': {'300_4': 400, '40_4': 20},
'stainless': {'300_4': 300, '40_4': 15},
}
d = awg_diameters[awg] / 1000 # Convert to meters
A = np.pi * (d/2)**2 # Cross-sectional area
key = f'{T_hot}_4' if T_hot == 40 else '300_4'
k_int = k_integrals[material][key]
Q_per_wire = (A / length_m) * k_int
return n_wires * Q_per_wire * 1000 # Convert to mW
print("="*70)
print("THERMAL BUDGET EXAMPLE: Cryocooler System for SCE Testing")
print("="*70)
print("\nSystem: Sumitomo RDK-415D (1.5W at 4.2K, 40W at 45K)")
print("-" * 50)
# 4K stage heat loads
loads_4k = {
'Chip dissipation (AQFP)': 20,
'DC wiring (20x 36AWG PhBr, 0.3m)': wire_heat_load(20, 36, 0.3, 'phosphor_bronze', 40, 4),
'RF coax (4x SS, thermalized)': 4 * 5,
'Mechanical supports (G10)': 30,
'Radiation (with shield)': 10,
'HEMT amplifier (1x)': 12,
}
print("\n4K Stage Heat Loads:")
total_4k = 0
for source, load in loads_4k.items():
print(f" {source}: {load:.1f} mW")
total_4k += load
print(f" {'─'*45}")
print(f" TOTAL: {total_4k:.1f} mW")
print(f" Available: 1500 mW")
print(f" Margin: {(1500-total_4k)/1500*100:.0f}%")
# 40K stage heat loads
loads_40k = {
'DC wiring (20x 36AWG PhBr from 300K)': wire_heat_load(20, 36, 0.5, 'phosphor_bronze', 300, 40) / 10,
'RF coax (4x SS from 300K)': 4 * 200,
'Radiation shield': 500,
'Mechanical supports': 2000,
}
print("\n40K Stage Heat Loads:")
total_40k = 0
for source, load in loads_40k.items():
print(f" {source}: {load:.1f} mW")
total_40k += load
print(f" {'─'*45}")
print(f" TOTAL: {total_40k:.1f} mW = {total_40k/1000:.1f} W")
print(f" Available: 40 W")
print(f" Margin: {(40000-total_40k)/40000*100:.0f}%")
# Calculate: Example thermal budget for a cryocooler system
def wire_heat_load(n_wires, awg, length_m, material='phosphor_bronze', T_hot=40, T_cold=4):
"""
Calculate heat load from wires.
Using simplified thermal conductivity integrals.
"""
# Wire diameters (mm) for common AWG
awg_diameters = {36: 0.127, 32: 0.202, 28: 0.321, 24: 0.511}
# Thermal conductivity integrals (W/m) from T_hot to 4K
k_integrals = {
'copper': {'300_4': 15000, '40_4': 800},
'phosphor_bronze': {'300_4': 1500, '40_4': 80},
'manganin': {'300_4': 400, '40_4': 20},
'stainless': {'300_4': 300, '40_4': 15},
}
d = awg_diameters[awg] / 1000 # Convert to meters
A = np.pi * (d/2)**2 # Cross-sectional area
key = f'{T_hot}_4' if T_hot == 40 else '300_4'
k_int = k_integrals[material][key]
Q_per_wire = (A / length_m) * k_int
return n_wires * Q_per_wire * 1000 # Convert to mW
print("="*70)
print("THERMAL BUDGET EXAMPLE: Cryocooler System for SCE Testing")
print("="*70)
print("\nSystem: Sumitomo RDK-415D (1.5W at 4.2K, 40W at 45K)")
print("-" * 50)
# 4K stage heat loads
loads_4k = {
'Chip dissipation (AQFP)': 20,
'DC wiring (20x 36AWG PhBr, 0.3m)': wire_heat_load(20, 36, 0.3, 'phosphor_bronze', 40, 4),
'RF coax (4x SS, thermalized)': 4 * 5,
'Mechanical supports (G10)': 30,
'Radiation (with shield)': 10,
'HEMT amplifier (1x)': 12,
}
print("\n4K Stage Heat Loads:")
total_4k = 0
for source, load in loads_4k.items():
print(f" {source}: {load:.1f} mW")
total_4k += load
print(f" {'─'*45}")
print(f" TOTAL: {total_4k:.1f} mW")
print(f" Available: 1500 mW")
print(f" Margin: {(1500-total_4k)/1500*100:.0f}%")
# 40K stage heat loads
loads_40k = {
'DC wiring (20x 36AWG PhBr from 300K)': wire_heat_load(20, 36, 0.5, 'phosphor_bronze', 300, 40) / 10,
'RF coax (4x SS from 300K)': 4 * 200,
'Radiation shield': 500,
'Mechanical supports': 2000,
}
print("\n40K Stage Heat Loads:")
total_40k = 0
for source, load in loads_40k.items():
print(f" {source}: {load:.1f} mW")
total_40k += load
print(f" {'─'*45}")
print(f" TOTAL: {total_40k:.1f} mW = {total_40k/1000:.1f} W")
print(f" Available: 40 W")
print(f" Margin: {(40000-total_40k)/40000*100:.0f}%")
====================================================================== THERMAL BUDGET EXAMPLE: Cryocooler System for SCE Testing ====================================================================== System: Sumitomo RDK-415D (1.5W at 4.2K, 40W at 45K) -------------------------------------------------- 4K Stage Heat Loads: Chip dissipation (AQFP): 20.0 mW DC wiring (20x 36AWG PhBr, 0.3m): 0.1 mW RF coax (4x SS, thermalized): 20.0 mW Mechanical supports (G10): 30.0 mW Radiation (with shield): 10.0 mW HEMT amplifier (1x): 12.0 mW ───────────────────────────────────────────── TOTAL: 92.1 mW Available: 1500 mW Margin: 94% 40K Stage Heat Loads: DC wiring (20x 36AWG PhBr from 300K): 0.1 mW RF coax (4x SS from 300K): 800.0 mW Radiation shield: 500.0 mW Mechanical supports: 2000.0 mW ───────────────────────────────────────────── TOTAL: 3300.1 mW = 3.3 W Available: 40 W Margin: 92%
6. Cryogenic Safety¶
Primary Hazards¶
| Hazard | Risk | Mitigation |
|---|---|---|
| Asphyxiation | He/N2 displace O2 | Ventilation, O2 monitors |
| Cold burns | Contact with cold surfaces/liquids | PPE, training |
| Pressure | Cryogen boil-off in sealed space | Pressure relief, no sealed containers |
| Embrittlement | Materials fail at cryo temps | Use appropriate materials |
Wet Systems (LHe/LN2) — Additional Hazards¶
These hazards apply when handling liquid cryogens:
Oxygen Deficiency — 1 liter of liquid helium → 750 liters of gas!
| O2 Level | Effect |
|---|---|
| 21% | Normal atmosphere |
| 19.5% | OSHA minimum safe level |
| 16% | Impaired judgment, rapid breathing |
| 10% | Loss of consciousness |
| 6% | Death within minutes |
Requirements for wet systems:
- O2 monitors in all rooms with cryogens (audible alarms at 19.5%)
- Adequate ventilation (basement labs especially at risk!)
- Never enter a space after large release without O2 testing
- Battery backup for O2 monitors
Pressure Safety (wet systems):
- Never seal a container with cryogens: It will build pressure and explode
- Pressure relief valves: Required on all dewars, check regularly
- Burst disks: Backup protection
- Ice plugs: Can block vent lines; use warming heaters on vents
- Transfer lines: Always have a path for gas to escape
Dry Systems (Cryocoolers) — Safety Considerations¶
Closed-cycle systems have fewer hazards but still require attention:
- No asphyxiation risk during normal operation (helium is sealed)
- Cold surfaces: 4K and 40K stages still cause cold burns if touched
- Vacuum window failure: Rare but can cause rapid pressure release
- Compressor: High-pressure helium lines; follow manufacturer guidelines
- Electrical: Compressors draw significant power; proper grounding required
PPE Requirements¶
| Item | Wet Systems | Dry Systems |
|---|---|---|
| Face shield | Required for transfers | Not typically needed |
| Cryogenic gloves | Required for LHe/LN2 handling | Only for cold hardware |
| Long pants/sleeves | Always | Always |
| Closed-toe shoes | Always | Always |
| Safety glasses | Always | Always |
In [5]:
Copied!
# Visualize: Oxygen deficiency hazard zones
fig, ax = plt.subplots(figsize=(12, 6))
# O2 levels and colors
zones = [
(21, 23, 'Normal (21%)', COLORS['success'], 1.0),
(19.5, 21, 'Acceptable (19.5-21%)', COLORS['success'], 0.5),
(16, 19.5, 'DANGER: Impaired (16-19.5%)', COLORS['secondary'], 0.7),
(10, 16, 'SEVERE: Incapacitation (10-16%)', COLORS['danger'], 0.7),
(0, 10, 'FATAL: (<10%)', COLORS['dark'], 0.9),
]
for o2_min, o2_max, label, color, alpha in zones:
ax.barh(0, o2_max - o2_min, left=o2_min, height=0.5,
color=color, alpha=alpha, edgecolor='black')
ax.text((o2_min + o2_max)/2, 0, label, ha='center', va='center',
fontsize=9, fontweight='bold', color='white' if color == COLORS['dark'] else 'black')
# Alarm threshold
ax.axvline(19.5, color='red', linewidth=3, linestyle='--')
ax.text(19.5, 0.35, 'ALARM\n19.5%', ha='center', fontsize=10, color='red', fontweight='bold')
ax.set_xlim(0, 23)
ax.set_ylim(-0.5, 0.5)
ax.set_xlabel('Oxygen Level (%)', fontsize=12)
ax.set_yticks([])
ax.set_title('Oxygen Deficiency Hazard Levels', fontsize=14, fontweight='bold')
plt.tight_layout()
plt.show()
print("\nKey safety rules:")
print("1. O2 monitors with alarms are MANDATORY")
print("2. Never work alone with large cryogen quantities")
print("3. Know your escape routes")
print("4. If alarm sounds, LEAVE IMMEDIATELY")
# Visualize: Oxygen deficiency hazard zones
fig, ax = plt.subplots(figsize=(12, 6))
# O2 levels and colors
zones = [
(21, 23, 'Normal (21%)', COLORS['success'], 1.0),
(19.5, 21, 'Acceptable (19.5-21%)', COLORS['success'], 0.5),
(16, 19.5, 'DANGER: Impaired (16-19.5%)', COLORS['secondary'], 0.7),
(10, 16, 'SEVERE: Incapacitation (10-16%)', COLORS['danger'], 0.7),
(0, 10, 'FATAL: (<10%)', COLORS['dark'], 0.9),
]
for o2_min, o2_max, label, color, alpha in zones:
ax.barh(0, o2_max - o2_min, left=o2_min, height=0.5,
color=color, alpha=alpha, edgecolor='black')
ax.text((o2_min + o2_max)/2, 0, label, ha='center', va='center',
fontsize=9, fontweight='bold', color='white' if color == COLORS['dark'] else 'black')
# Alarm threshold
ax.axvline(19.5, color='red', linewidth=3, linestyle='--')
ax.text(19.5, 0.35, 'ALARM\n19.5%', ha='center', fontsize=10, color='red', fontweight='bold')
ax.set_xlim(0, 23)
ax.set_ylim(-0.5, 0.5)
ax.set_xlabel('Oxygen Level (%)', fontsize=12)
ax.set_yticks([])
ax.set_title('Oxygen Deficiency Hazard Levels', fontsize=14, fontweight='bold')
plt.tight_layout()
plt.show()
print("\nKey safety rules:")
print("1. O2 monitors with alarms are MANDATORY")
print("2. Never work alone with large cryogen quantities")
print("3. Know your escape routes")
print("4. If alarm sounds, LEAVE IMMEDIATELY")
Key safety rules: 1. O2 monitors with alarms are MANDATORY 2. Never work alone with large cryogen quantities 3. Know your escape routes 4. If alarm sounds, LEAVE IMMEDIATELY
7. Lab Setup & Operations¶
Essential Lab Infrastructure¶
| Item | Wet Systems | Dry Systems |
|---|---|---|
| O2 monitors | Required (battery backup) | Optional (no open cryogens) |
| Ventilation | Critical | Standard lab ventilation |
| Dewar storage | Required, away from exits | Not needed |
| Helium recovery | Essential for cost | Not applicable |
| Compressed gas | For recovery system | For cryocooler charge |
| Electrical | Standard | Clean power for compressor |
Dry System (Cryocooler) Operations¶
Pre-Cooldown Checklist:
- Vacuum system leak-checked (<10⁻⁵ mbar)
- All wiring connections verified at room temp
- Sample mounted with good thermal contact (use Apiezon grease)
- Thermal links verified
- Compressor helium charge checked
Cooldown Procedure:
- Verify vacuum (<10⁻⁵ mbar)
- Start compressor, verify helium flow
- Monitor temperatures (40K stage: ~2 hrs, 4K stage: 6-12 hrs)
- Verify all temperature sensors reading correctly
- Begin device testing only after thermal equilibrium
Common Failure Modes:
| Symptom | Likely Cause | Debug |
|---|---|---|
| Won't cool below 40K | Vacuum leak | Check pressure, leak check |
| 4K stage too warm | Heat load too high | Check wiring, reduce cables |
| Temperature unstable | Poor thermal contact | Check mounting, add grease |
| Slow cooldown | Low helium charge | Check compressor pressure |
| Compressor cycling | Contaminated helium | May need purge/recharge |
Wet System (LHe) Operations¶
Pre-Cooldown Checklist:
- Vacuum system leak-checked (<10⁻⁵ mbar)
- All wiring connections verified at room temp
- Sample mounted securely
- O2 monitors active and calibrated
- Recovery bags/system ready
- LHe supply confirmed, transfer line available
- Cooldown schedule communicated to lab
Cooldown Procedure:
- Pre-cool with LN2 to 77K (saves LHe)
- Pump out LN2, warm briefly to avoid ice
- Begin LHe transfer slowly
- Monitor boil-off rate to estimate fill level
- Top off as needed during measurements
Common Failure Modes:
| Symptom | Likely Cause | Debug |
|---|---|---|
| High boil-off rate | Vacuum degradation | Leak check, check pressure |
| Ice blocking lines | Moisture contamination | Warm and purge |
| Can't transfer LHe | Blocked transfer line | Check for ice plugs |
| Rapid pressure rise | Blocked vent | Clear vent immediately! |
7.1 Electrical Ground Isolation¶
Dry cryocoolers introduce electrical noise that wet systems (LHe baths) do not have. Understanding the noise sources and isolation options is critical for sensitive measurements.
Why Cryocoolers Are Electrically Noisy¶
| Noise Source | Mechanism | Amplitude |
|---|---|---|
| GM valve motor | Stepper motor switching at ~1 Hz | 50-200 mV ground bounce |
| Compressor switching | 5-15 A motor on 50/60 Hz | 100+ mV ground spikes |
| Helium pressure pulses | Piezoelectric effect in cold head | µV-mV microphonics |
The GM (Gifford-McMahon) cycle uses a motorized valve to switch helium flow direction. This valve motor draws significant current and creates ground noise that propagates through the cryostat structure.
Wet systems don't have this problem - a LHe dewar is electrically passive.
Two Ground Domain Architecture¶
For sensitive measurements (AQFP, low-noise analog), isolate the measurement ground from the cryocooler:
DOMAIN 1: Earth Ground DOMAIN 2: Measurement Ground
───────────────────────── ───────────────────────────
• Compressor chassis • Isolated instrument rack
• Cryocooler cold head • DAQ and bias electronics
• Cryostat vacuum chamber • Sample backplane
• Building electrical • On-chip ground planes
┌──────────────────┐
Cold Finger ───┤ Alumina Disk ├─── Sample Stage
(Domain 1) │ (ceramic) │ (Domain 2)
└──────────────────┘
Good thermal contact
Electrical isolation
Alumina (Al₂O₃) ceramic disk:
- Thermal conductivity: ~30 W/m·K at 4K (good heat transfer)
- Electrical resistivity: >10¹² Ω·cm (complete isolation)
- Typical thickness: 2-5 mm
- Mounts between cold finger and sample backplane
Isolation Transformer Placement¶
Option A: Isolate compressor (ideal for small compressors <5 kW)
- Medical-grade isolation transformer on compressor power
- Breaks ground current path at source
- ~$1000-2000 for 5 kVA unit
Option B: Isolate measurement electronics (practical for large compressors)
- Smaller isolation transformer (0.5-2 kVA) on instrument rack
- Easier and cheaper than isolating large compressor
- Requires attention to DAQ cables (see below)
Maintaining Isolation Through Cables¶
The isolation fails if any cable creates a galvanic path:
WRONG - Ground loop through USB:
[Isolated Rack] ←─ USB ─→ [Computer on building power]
│ │
└────── both grounded ───────┘
= ISOLATION DEFEATED
RIGHT - Use USB isolator or fiber:
[Isolated Rack] ←─ USB ─→ [USB Isolator] ←─ USB ─→ [Computer]
(ADuM4160 or similar)
Signal cable checklist:
- USB: Use USB isolator ($50-200) or fiber converter
- Ethernet: Already transformer-isolated (OK as-is)
- GPIB: Ground at one end only, or use fiber extender
- Coax/SMA: Shield grounded at one end only (usually instrument end)
Feedthrough Considerations¶
The vacuum feedthroughs must maintain Domain 2 isolation:
- RF feedthroughs: Use types rated for high voltage isolation (>500V)
- DC feedthroughs: Floating design, not grounded to flange
- Fiber feedthroughs: Inherently isolated (preferred for sensitive signals)
When Isolation Is NOT Needed¶
Most measurements work fine with star grounding to the cryostat chassis:
| Application | Isolation Needed? |
|---|---|
| SCE digital (RSFQ/ERSFQ) | No - digital margins absorb noise |
| SQUID magnetometers | No - commercial systems use chassis ground |
| Standard I-V curves | No - signal levels are mV |
| AQFP with sensitive analog | Yes - zeptojoule signals need low noise |
| STM/tunneling measurements | Yes - femtoamp currents |
Start with chassis grounding and add isolation only if you have documented noise problems that survive proper star grounding.
Noise Debug Protocol¶
- First: Verify star ground topology (all grounds meet at single point)
- Measure ground-to-ground voltage between isolated and building ground
- <10 mV: Good isolation
100 mV: Find the galvanic path
- Never probe Domain 2 (isolated measurement ground) with a standard scope - the scope's ground connects to building earth through its power cord, defeating the isolation. Use differential probes or power the scope from the same isolation transformer.
- Compare measurements in dry system vs. LHe dip probe - if LHe works but dry doesn't, the chip is good; focus on the dry system grounding
8. Summary¶
Key Concepts¶
Why 4K?
- Nb-based SCE requires T << Tc = 9.2 K
- 4.2 K is LHe boiling point: convenient, self-regulating
- Reasonable cooling power available from cryocoolers
Cryostat Selection:
- Wet (LHe bath): High cooling power, simple, but consumes helium
- Dry (cryocooler): No LHe, but limited to ~1-2 W at 4K
- Helium recovery essential for wet systems (90%+ recovery possible)
Thermal Budgeting:
- Use intermediate stages to intercept heat (40K has 30-50W, 4K has 1-2W)
- Every wire is a heat leak; use low-conductivity materials
- Thermalize all wiring at each temperature stage
Electrical Grounding:
- Dry systems introduce GM valve motor and compressor noise (50-200 mV)
- Two ground domain architecture: earth ground vs. measurement ground
- Alumina ceramic disk provides thermal link with electrical isolation
- Start with star grounding; add isolation only if needed
Safety:
- O2 monitors are mandatory (alarm at 19.5%)
- Never seal containers with cryogens
- PPE: face shield, cryo gloves, closed shoes
Key Numbers¶
| Parameter | Value |
|---|---|
| LHe boiling point | 4.2 K |
| LHe latent heat | 2.6 kJ/L |
| LHe expansion ratio | 1:750 |
| GM/PT cooling at 4K | 1-2 W |
| GM/PT cooling at 40K | 30-50 W |
| O2 alarm threshold | 19.5% |
| Cu thermal integral (300→4K) | 15,000 W/m |
| PhBr thermal integral (300→4K) | 1,500 W/m |