Skip to main content
Sandbox Physics
How physics instruments measure

Inside the instrument

Why a bigger vacuum pump may barely help the sample

The pump is running, its inlet pressure is low, yet the gauge beside the sample refuses to fall. Even a larger pump may make little difference. Gas leaving the chamber must first pass through a valve and a pipe. The ease of that journey determines how much pumping the sample actually receives.

Sandbox Physics ·

Molecules must find the exit

At high vacuum, gas molecules seldom collide with each other. A molecule is more likely to hit a chamber wall, rebound and strike another surface. The pump removes molecules that reach its inlet. A long narrow pipe limits how many arrive from the chamber.

The ease with which gas passes through a component is called conductance. A narrower opening admits fewer molecules, and a longer pipe introduces more wall collisions. For a sufficiently long circular pipe carrying the same gas at the same temperature, halving the diameter reduces molecular-flow conductance to about one eighth.

Suppose a pump offers 80 litres per second at its inlet and its connecting pipe has a conductance of 5 litres per second. Together they provide only about 4.7 litres per second at the chamber. Doubling the pump to 160 litres per second raises this to about 4.85. The narrow route still controls the journey. Litres per second describes pumping capability; the amount of gas removed also depends on the inlet pressure.

A shorter or wider pipe, or a pump closer to the chamber, can therefore help more than increasing pump size. Gauges at the chamber and pump inlet reveal the pressure difference along the route. Combined with its conductance, that difference gives the gas throughput through the pipe.

1Seff=1S+1C\frac{1}{S_{\mathrm{eff}}}=\frac{1}{S}+\frac{1}{C}
S is pump inlet speed, C is molecular-flow conductance of the connecting route, and S_eff is effective pumping speed at the chamber. All use the same volume-per-time units. This relation applies to the linear molecular-flow connection considered here.

Sources: [3][4]

A sealed wall can still supply gas

Closing a vacuum chamber does not instantly remove its gas sources. Metal exposed to air retains water and other molecules on its surfaces. Some gradually leave during pumping, a process called outgassing. As the rate supplied by the walls approaches the rate removed by the pump, pressure falls more slowly.

Gas can also hide deep inside a blind hole, screw thread or joint. The main chamber may already be evacuated while a small internal pocket empties through a narrow gap. This internal gas source is commonly called a virtual leak. Giving such pockets a clear evacuation route helps the gas leave sooner.

A crack connected to the outside can also hold pressure up. Wall gas, trapped gas and an external leak can contribute together. Similar falling-pressure traces can arise from different combinations. To locate an outside route, give the external gas a recognizable marker.

Sources: [3][5]

How a gauge turns gas into a reading

An ionization gauge sends electrons through the residual gas and collects the ions they create. More gas usually produces more ions. But gas species differ in how readily they ionize. Helium and nitrogen at the same pressure can produce different currents.

A gauge calibrated with nitrogen initially reports a nitrogen-equivalent pressure for an unknown mixture: the pressure of nitrogen that would produce the observed signal. Recovering the actual total pressure requires the mixture and its relative responses. Applying a helium correction to an unidentified gas mixture discards that distinction.

A capacitance diaphragm gauge reads the force of the gas on a thin membrane. Pressure bends the membrane and changes its separation from a fixed electrode. That separation changes the amount of electric charge the electrodes can store, their capacitance. Electronics turn the change into a pressure indication. Helium and nitrogen share the same pressure scale because this method senses mechanical force. At sufficiently low pressure, the signal falls close to zero drift and the instrument’s resolution; a useful reading then requires a gauge with a lower working range.

Known pressures provide calibration marks, and the original voltages allow later recalculation. A separate known sample, excluded from fitting, checks the calibration. Chamber and pump-inlet gauges each have their own response and need individual calibration before their readings can be compared.

Sources: [3][7][1][2]

Let helium trace the route

Apply helium outside a suspect joint and watch for helium inside. If a route connects them, the tracer enters the chamber, passes the valve and pipe, and reaches the leak detector. The detector ionizes the gas, selects the helium signal by ion mass and collects the resulting ions.

The signal takes time to rise. Helium must accumulate and travel through the system, and the detector has its own response. After the tracer is removed, helium already inside must be pumped away. A large chamber connected through a narrow pipe can require a surprisingly long wait.

Measure the background, then connect a known helium flow at the same location so it travels the entire route. Compare the background-subtracted unknown signal with this standard to obtain a flow. Changes to the pump, valve or pipe alter the fraction and timing reaching the detector, requiring calibration again. Conversion between moles per second and pressure-volume throughput also needs a stated gas reference temperature.

A helium response supports a tracer-transmitting route. Helium can also permeate some materials, so practical leak testing combines location, response time and material information. Retaining original signals, background, standards, tracer timing and settling time lets someone else reconstruct how the flow was measured.

Sources: [4][6][7][5]

Sources and further reading

  1. INFICON SKY CDG025D Operating Manual, tina49e1-f (2024-08)

    Printed pages 6 and 9: pressure deflects a ceramic diaphragm, capacitance supplies a linear analog signal, gas independence and finite resolution. Selected sections read; the Lab is a separate reduced model.

  2. INFICON SKY CDG025D Datasheet, tiba33e1-h (2024)

    Pages 2–3: full-scale options, resolution and suggested lower reading. These bounds explain why a diaphragm gauge loses useful low-pressure readings.

  3. Pfeiffer Vacuum, Vacuum Technology Book II, Part 2

    Printed pages 15–17, 19–20 and 94–95: throughput, molecular long tubes, wall gas and gas-dependent ionization gauges. Selected sections read; no full-book review.

  4. Pfeiffer Vacuum, Leak Detection Compendium (April 2013)

    Printed pages 11–15: units, calibration through the whole system, effective speed and tracer response. These relations are used within molecular flow here.

  5. Leybold, Definition and measurement of vacuum leaks

    Types of leaks, including gas from trapped cavities; outside-in testing, standard helium conditions and partial-flow measurements.

  6. P. Cruikshank and G. Bregliozzi, CERN CAS Leak Detection Tutorial (June 2017)

    Slides 25, 29–30 and 35–37: background and external-standard corrections, mass-selective detection, response and recovery time.

  7. C. Ehrlich and S. Tison, NIST SP 250-38 (January 1992)

    Printed pages 1–5: comparison of an unknown helium flow with a known flow after equilibration; temperature and molar-flow units. Historical method, not a claim of NIST traceability.

FeedbackSomething broken, confusing, or missing on this page?Tell us
What kind of feedback?

Sent with this page’s address, its Lab ID, the site language and your device type. Please leave out personal details. Feedback is kept for up to 12 months.Privacy policy