GE HDxt 3.0T: cryostat pressure rising after cold head replacement
What happened
About a week after a cold head RDK-408A3 replacement (with the adsorber), the cryostat pressure started climbing to a reported 4.7 psi, while the shield and recondenser temperatures still read normal (reported 41.3 K and 4.3 K); the new head had only about two days of actual run time in the case.
What was checked
- Cryostat pressure trend compared with shield and recondenser temperatures (reported 41.3 K and 4.3 K) to see whether the pressure rise tracks a loss of cooling.
- Compressor working pressure checked; the reported dynamic value of about 2.1 raised the question of refrigerant charge after the head and adsorber change.
- Magnet monitor data requested to plot the pressure and cooling behaviour since the replacement.
- Inspection planned of the cold head to recondenser joint for a displaced indium spring, plus a look at cryostat vacuum quality as a possible cause.
Cause
Not confirmed in the case. The most likely suspect was lost cooling capacity at the recondenser — for example an indium spring displaced during installation, leaving the cold head without good thermal contact — since the shield and recondenser temperatures looked normal; degraded cryostat vacuum after the work was also considered.
The cause is probable and was not fully confirmed.
What was done
No confirmed fix was reported. The OEM service pushed for a compressor refrigerant top-up, while other engineers recommended first verifying the cold head to recondenser contact (the indium spring) and analysing the magnet monitor data; the outcome stayed probable.
Safety
The superconducting magnet stays energised and the cryostat holds liquid helium under pressure: keep ferromagnetic items out of the magnet room, treat any cryostat pressure rise as a quench and oxygen-depletion risk, and work on the cryogenic system only to the manufacturer's procedures.
Lesson: When cryostat pressure climbs after cold head replacement while the shield and recondenser temperatures look normal, suspect the joint you have just disturbed (indium seal/contact) and the vacuum first — confirm with magnet monitor data before touching the compressor charge.
Covered in the course: MRI Service — Level 2: Fault-Finding →
Written from an anonymised case shared by service engineers. It is not a manufacturer instruction and grants no authorisation. Independent note; names are trademarks of their owners.