Direct Answer
Helium mass-spectrometer leak testing is the standard method for locating and quantifying leaks in vacuum feedthroughs. The test works by exposing the feedthrough to helium, either by spraying helium on the exterior while the interior is connected to a leak detector (vacuum method) or by pressurizing the interior with helium and sniffing the exterior (sniffer method), and measuring the helium signal reaching the detector’s mass spectrometer. A useful leak-test requirement defines the test method, test direction, instrument sensitivity, calibration status, background level, dwell time, acceptance criterion, and whether the finished assembly or only a subcomponent is being tested.
A reported leak rate, commonly expressed in units such as atm cc/sec, is a flow-based quantity and should not be confused with a pressure reading. The correct acceptance value for a given application depends on its gas-load budget, pumping configuration, and reliability requirements rather than on a single generic threshold.
Vacuum Method vs. Sniffer Method
In the vacuum method, the feedthrough or assembly is connected to a calibrated leak detector so that its interior is under vacuum, and helium is applied to the exterior, typically with a fine probe, at suspect locations such as the ceramic-to-metal joint, the braze fillet, and the flange perimeter. This method generally offers the highest sensitivity and the clearest localization, because the response can be correlated directly with where the helium was applied.
In the sniffer method, the interior of the assembly is pressurized with helium or a helium-air mixture, and a sniffer probe connected to the leak detector is passed over the exterior surfaces to detect helium escaping through a leak path. This method is often used when the component cannot be directly connected to a vacuum-side leak detector, but it typically has lower sensitivity and is more affected by ambient helium background and probe speed than the vacuum method.
| Test method | Typical use case | Relative sensitivity | Key limitation |
|---|---|---|---|
| Vacuum method | Component or assembly can be connected directly to a leak detector | Higher, with clear localization | Requires vacuum-side access to the part |
| Sniffer method | Finished system or installed component where vacuum-side connection is impractical | Lower, more affected by background and technique | Sensitive to probe speed, distance, and ambient helium |
Preparing for a Helium Leak Test
- Confirm the leak-test requirement, including method, direction, sensitivity, acceptance criterion, and whether the requirement applies to the feedthrough alone or the finished assembly including flange, gasket, and connector hardware.
- Verify the leak detector’s calibration status using a certified reference leak of known leak rate, and record the calibration result before testing.
- Establish and record the background helium level in the test area before applying helium to the part, since ambient helium can produce a false response, particularly with the sniffer method.
- Clean and dry the part according to its handling requirements, since contamination or residual solvents can affect both the leak signal and the detector’s stability.
- Confirm that any gasket, flange hardware, and mounting interface involved in the test are installed and torqued according to the applicable procedure, since a test performed with incomplete hardware does not represent the finished configuration.
Performing the Test
- Pump down the system to the leak detector’s required operating pressure and allow the reading to stabilize before beginning helium application.
- Apply helium systematically and slowly at each suspect location, including the feedthrough body, the ceramic-to-metal or braze joint, the flange perimeter, and any weld, moving from areas least likely to leak toward areas most likely to leak to avoid contaminating the test chamber with a large early helium release.
- Hold the probe or spray at each location for a dwell time sufficient for the response to travel through the system and register on the detector, since response time depends on the detector’s pumping configuration and the distance from the test point.
- Record the leak rate at each tested location, along with the corresponding test point, so that a positive response can be localized to a specific joint or interface rather than reported only as a whole-assembly value.
- Compare each result against the defined acceptance criterion and record whether the part passes, fails, or requires further investigation at a specific location.
- After any corrective action, repeat the test on the affected location and, where required, on the full assembly to confirm the repair.
Interpreting Results Correctly
A leak rate below the detector’s minimum detectable level should be reported as such rather than as zero, since a leak detector has a finite sensitivity floor and cannot confirm the complete absence of any leak path. A leak-rate value should always be reported with its units and test conditions, because a value without units or without context about test direction and sensitivity cannot be meaningfully compared against an acceptance criterion or a different test’s results.
A passing helium leak test also does not, by itself, confirm that a feedthrough is suitable for a given application. A component can be fully hermetic and still be unsuitable if its materials, geometry, or trapped volumes create excessive outgassing, or if its electrical, thermal, or mechanical characteristics do not match the application’s requirements. Leak testing should be evaluated alongside, not as a substitute for, the other elements of a full qualification plan.
Common Sources of Test Error
- Elevated background helium in the test area, which can produce a false positive or mask a real leak, particularly with the sniffer method
- Probe movement that is too fast to allow the detector to register a response at each location
- Applying helium to multiple locations simultaneously, which prevents accurate localization of a positive result
- Testing before the system has stabilized at its base pressure, which can produce an unstable or misleading baseline reading
- Using a detector or reference leak with expired or unverified calibration
- Testing an incomplete assembly, such as a feedthrough without its intended gasket or flange hardware, when the finished-assembly configuration is what actually needs to be qualified
Related Reading
- Vacuum Feedthrough Failure: Causes, Diagnosis, Prevention, and Selection for UHV Systems
- How to Tell a Real Vacuum Leak From Outgassing or a Virtual Leak
- Vacuum Feedthrough Failure Modes: Cracked Ceramics, Braze Leaks, and Electrical Breakdown
- UHV Feedthrough Documentation and Qualification
Frequently Asked Questions
What leak rate is acceptable for a UHV feedthrough?
There is no universal acceptable leak rate. The required value should be established from the system’s gas-load budget, pumping configuration, operating environment, and any applicable customer or program specification, and it should not be assumed to carry over from a different application.
Is the vacuum method or sniffer method more accurate?
The vacuum method generally offers higher sensitivity and clearer localization because the part is directly connected to the leak detector. The sniffer method is useful when direct vacuum-side connection is not practical, but it is more sensitive to technique and ambient conditions.
Can a feedthrough pass a helium leak test and still fail in service?
Yes. A passing leak test confirms hermeticity at the time and under the conditions tested. It does not confirm suitability for outgassing behavior, electrical load, thermal cycling, or mechanical stress that the component will experience in service, nor does it guarantee that a joint under cumulative thermal-cycling stress will remain leak-tight indefinitely.
Why does a leak-test requirement need to specify units and test direction?
A leak rate reported without units or test direction cannot be compared meaningfully against an acceptance criterion or another test’s results, since sensitivity, response time, and reported values differ between the vacuum method and the sniffer method.
How often should a feedthrough be re-tested after installation?
Re-testing frequency depends on the application’s reliability requirements, bakeout and thermal-cycling exposure, and any program-specific maintenance plan. Retesting after a bakeout cycle, a mechanical disturbance, or a suspected performance change is common practice.
Next Steps
When specifying a leak-test requirement for a new feedthrough or evaluating a suspected leak in service, define the test method, sensitivity, and acceptance criterion before testing begins. Contact MPF Products with the application’s gas-load budget and reliability requirements to confirm an appropriate leak-test specification.