How to Test a Vacuum Pump: Ultimate Vacuum, Pump-Down Time and Leaks
HCEM Engineering TeamShare
To test a vacuum pump, cap its inlet with a vacuum gauge, run it and compare the lowest reading with the ultimate vacuum on its datasheet. Then time how long it takes to pump down a known volume, check its flow, and close an isolation valve to see whether the vacuum holds. Those four checks tell you whether the pump is healthy, whether it is worn, and whether the real problem is a leak somewhere else in the machine.
This guide gives a practical test procedure for oil-free (dry) piston and diaphragm vacuum pumps of the kind built into OEM equipment: suction units, packaging machines, CNC tables, lab instruments. The worked examples use datasheet values from our own oil-free piston vacuum pumps.
How do you test a vacuum pump safely?
The forces in a vacuum test are large. At −90 kPa the atmosphere pushes on every square centimetre of a vessel with about 9 N, close to the weight of a 1 kg mass. Before you start:
- Use a vessel made for vacuum. A thin-walled tank, a plastic container or ordinary glassware can implode. Use a vacuum-rated chamber, receiver or desiccator, and wear eye protection.
- Check the supply against the nameplate. A 110 V/60 Hz pump on a 220 V/50 Hz supply, or a 12 V DC pump on 24 V, will overheat or fail. For DC pumps, check polarity and fusing.
- Never block the exhaust. A vacuum pump pushes everything it removes out of its outlet. If that is closed, the pump head is pressurised instead.
- Isolate the power before you change fittings, and mind the hot cylinder head after a long run.
- Keep liquid out. Dry pumps are built for air. If the application can carry liquid or dust, fit a trap or filter before testing on the real system.
What do you need to test a vacuum pump?
The essential tool is a vacuum gauge with the right range. Oil-free piston pumps work in the rough vacuum range, from atmospheric pressure down to a few tens of mbar absolute. Choose a gauge that reads clearly in that range:
- Mechanical (Bourdon) gauge, −100 to 0 kPa or −1 to 0 bar. Cheap and robust. A class 1.6 gauge on that scale is accurate to ±1.6 kPa, which is fine for a pass/fail check.
- Digital gauge in kPa or mbar. Easier to read near the end of the scale and often able to log readings for the pump-down test.
- Absolute digital gauge (for example 0–1100 mbar abs). The best choice for comparing pumps, because the reading does not change with the weather or altitude.
A micron gauge is the wrong tool here. It is made for the deep vacuum used to evacuate HVAC and refrigeration systems, where the target is often around 500 microns (0.67 mbar). That is the job of an oil-sealed rotary vane pump. A high-vacuum oil-free piston pump stops at about 33 mbar, roughly 25,000 microns, so a micron gauge sits at the top of its scale.
You will also need short, vacuum-rated hose or hard fittings, thread sealant, a tee, a ball valve to isolate the pump from the vessel, a stopwatch or logger, and a clamp meter. A vessel of known volume makes the pump-down and leak tests possible; a 10–25 L receiver suits most small and mid-size pumps.
Gauge or absolute: how do you read vacuum units?
Datasheets for oil-free piston pumps usually give vacuum as gauge pressure, measured against the atmosphere: −92 kPa means 92 kPa below atmospheric. Absolute pressure is measured against a perfect vacuum and is always positive. Keep these conversions to hand:
| Unit | mbar | kPa | Torr | inHg |
|---|---|---|---|---|
| 1 bar | 1000 | 100 | 750.06 | 29.53 |
| 1 kPa | 10 | 1 | 7.5006 | 0.2953 |
| 1 mbar (= 1 hPa) | 1 | 0.1 | 0.75006 | 0.02953 |
| 1 Torr (≈ 1 mmHg) | 1.3332 | 0.13332 | 1 | 0.03937 |
| 1 inHg | 33.864 | 3.3864 | 25.4 | 1 |
| Standard atmosphere | 1013.25 | 101.325 | 760 | 29.92 |
Here is the same scale from the gauge side, assuming a standard atmosphere of 1013 mbar:
| Gauge reading | Absolute pressure | Absolute in Torr | Vacuum in inHg |
|---|---|---|---|
| −80 kPa | 213 mbar | 160 Torr | 23.6 inHg |
| −86 kPa | 153 mbar | 115 Torr | 25.4 inHg |
| −90 kPa | 113 mbar | 85 Torr | 26.6 inHg |
| −92 kPa | 93 mbar | 70 Torr | 27.2 inHg |
| −95 kPa | 63 mbar | 47 Torr | 28.1 inHg |
| −98 kPa | 33 mbar | 25 Torr | 28.9 inHg |
One consequence catches many people out. A gauge reading depends on the air pressure on the day. A pump that reaches 93 mbar absolute shows −92 kPa when the barometer reads 1013 mbar, but only about −89 kPa on a stormy 980 mbar day, and about −81 kPa at 1000 m altitude. The pump has not changed; the reference has. Note the barometric pressure with every gauge reading, or use an absolute gauge. Our guide to altitude and ambient conditions covers this in more detail. For flow units, the CFM to L/min converter and our explainer on what CFM means help.
How do you test a vacuum pump's ultimate vacuum?
The ultimate (blank-off) vacuum test shows the deepest pressure the pump itself can reach, with nothing else in the way. It is the quickest way to check whether a vacuum pump is working.
- Connect the gauge straight to the pump inlet with the shortest fitting you can, sealed with thread tape or a bonded seal. No hose, no machine, no extra valves. Leave the inlet filter in place if the pump normally runs with it.
- Check the supply: rated voltage and frequency, measured at the pump terminals while it runs.
- Run the pump and watch the needle. It falls fast at first, then creeps. Wait until the reading has stopped falling for about a minute. A few minutes of running brings the pump to a steady state.
- Record the lowest reading, the barometric pressure, the ambient temperature and the supply voltage.
- Compare with the datasheet, after allowing for the barometer and the gauge's accuracy.
If the pump reaches its rated ultimate vacuum, the pump is sound, whatever the machine is doing. If it falls clearly short at the right voltage, the cause is inside the pump: usually wear of the piston ring or seal, a damaged valve plate, or a leaking head gasket. For what the ultimate figure does and does not tell you in service, see ultimate pressure vs working vacuum.
What reading should a healthy vacuum pump reach?
The spec sheet is the reference. Here are datasheet values for five of our oil-free piston vacuum pumps. Several come in a standard and a high-vacuum version: the high-vacuum version goes deeper but moves less air. Absolute values assume a 1013 mbar atmosphere.
| Model | Supply | Version | Free-air flow | Ultimate vacuum (gauge) | About (absolute) | Rated power |
|---|---|---|---|---|---|---|
| HC80D | 12/24/48 V DC | Standard | 28 L/min (0.99 cfm) | −86 kPa | 153 mbar | ≤70 W |
| HC80D | 12/24/48 V DC | High vacuum | 12 L/min (0.43 cfm) | −95 kPa | 63 mbar | ≤70 W |
| HC100A | 220 V/50 Hz or 110 V/60 Hz | Standard | 45 L/min at 50 Hz, 50 L/min at 60 Hz | −90 kPa | 113 mbar | ≤110 W or ≤150 W |
| HC100A | 220 V/50 Hz or 110 V/60 Hz | High vacuum | 22.5 L/min at 50 Hz, 25 L/min at 60 Hz | −98 kPa | 33 mbar | ≤110 W or ≤150 W |
| HC480D | 12/24/48 V DC | Standard | 100 L/min (3.53 cfm) | −92 kPa | 93 mbar | ≤200 W |
| HC480D | 12/24/48 V DC | High vacuum | 50 L/min (1.77 cfm) | −98 kPa | 33 mbar | ≤200 W |
| HC580A | 220 V/50 Hz or 110 V/60 Hz | Standard | 120 L/min (4.24 cfm) | −92 kPa | 93 mbar | ≤340 W |
| HC580A | 220 V/50 Hz or 110 V/60 Hz | High vacuum | 60 L/min (2.12 cfm) | −98 kPa | 33 mbar | ≤340 W |
| HC1500A | 220 V/50 Hz or 110 V/60 Hz | Standard | 210 L/min at 50 Hz, 250 L/min at 60 Hz | −92 kPa | 93 mbar | ≤700 W or ≤750 W |
| HC1500A | 220 V/50 Hz or 110 V/60 Hz | High vacuum | 110 L/min at 50 Hz, 130 L/min at 60 Hz | −98 kPa | 33 mbar | ≤700 W or ≤750 W |
Know which version you have before you judge the result. A standard HC580A that stops at −92 kPa is healthy; a high-vacuum HC580A that stops at −92 kPa is not.
How do you measure vacuum pump-down (evacuation) time?
The blank-off test shows how deep the pump goes. The pump-down test shows how fast it gets there, which is what decides cycle time in a machine. Pump a vessel of known volume from atmospheric pressure to a target pressure and time it.
- Connect the vessel to the pump with a short hose and a ball valve, and put the gauge on the vessel, not on the pump.
- Open the vessel to atmosphere, then close it. Start the pump and the stopwatch together.
- Stop the clock when the gauge reaches the target, for example −80 kPa.
- Repeat three times and use the average.
To judge the result, compare it with a calculation. For an ideal pump with constant speed:
t = (V / S) × ln(p₀ / p₁)
where V is the volume, S is the pumping speed, and p₀ and p₁ are the start and target pressures. Both pressures must be absolute, and V and S must share the same volume unit (litres and litres per minute give minutes).
A real pump slows down as it approaches its ultimate pressure pu. A better estimate near the end of the curve is t = (V / S) × ln[(p₀ − pu) / (p₁ − pu)].
Worked example. A standard HC580A (120 L/min free-air flow, ultimate −92 kPa, about 93 mbar absolute) pumps a 20 L receiver from 1013 mbar to −80 kPa gauge, which is 213 mbar absolute. Hose volume (2 m of 8 mm bore is about 0.1 L) is small enough to ignore.
- Ideal formula: (20 ÷ 120) × ln(1013 ÷ 213) = 0.167 min × 1.559 = 0.26 min, about 16 seconds.
- With the ultimate pressure: (20 ÷ 120) × ln(920 ÷ 120) = 0.167 min × 2.037 = 0.34 min, about 20 seconds.
A measured 24 seconds would be close to that estimate; the hose, fittings and the pump's real speed curve add the rest. The test is most useful like for like: same pump, vessel, hose and target as last time, or as a new pump of the same model.
To choose a pump for a required cycle time, see our guide to evacuation time sizing for OEM buyers.
How do you check vacuum pump flow?
Free-air flow is what the pump moves with the inlet open to atmosphere. There are two practical ways to check it:
- Flowmeter on the inlet. Fit a low-restriction flowmeter (variable-area or thermal mass) sized for the pump's range on the open inlet. Any restriction in the meter lowers the reading, so choose one with a generous range.
- Back-calculate from the pump-down test. Rearrange the formula: S = (V / t) × ln[(p₀ − pu) / (p₁ − pu)]. In the example above, 24 seconds (0.4 min) gives (20 ÷ 0.4) × 2.037 ≈ 102 L/min of effective speed against a 120 L/min rating.
The effective figure includes the losses of the hose and fittings, so it is always below the datasheet. What matters is the trend, and how flow and ultimate vacuum move together. Low flow with a good blank-off reading points to a restriction: a clogged inlet filter, a long or narrow hose (see inlet piping pressure loss), a half-closed valve, or a sagging DC supply. Low flow and a poor blank-off reading point to wear inside the pump.
How do you check for leaks: is it the pump or the system?
When a machine will not reach its working vacuum, the first question is whether the pump or the system is at fault. Two tests answer it.
1. Blank off the pump. Run the ultimate vacuum test above at the pump inlet. If the pump reaches its spec there but the machine does not, the leak is in the system: hoses, fittings, seals, suction cups, valves or the chamber itself.
2. Rate-of-rise test. Pump the system down, close the isolation valve between pump and system, then stop the pump. Record the pressure in the system every minute.
- A steady, straight-line rise means a real leak: air is coming in from outside.
- A rise that slows and levels off usually means moisture or gas being released from surfaces, hoses and seals, not a leak. Pumping for longer, or drying the system, improves it.
- If the pressure rises only when the valve is open and the pump is stopped, air is coming back through the pump. See our guide to backflow prevention.
The leak rate is Q = V × Δp ÷ Δt. A 20 L system that rises from 200 to 230 mbar in 5 minutes leaks 20 × 30 ÷ 300 = 2 mbar·L/s. To find where the leak is, isolate sections of the system one at a time and repeat. An ultrasonic leak detector helps on larger machines; soapy water works poorly under vacuum because the bubbles are drawn inwards.
How much leakage is acceptable depends on the machine. For acceptance criteria and test conditions, see our vacuum system leak rate acceptance test guide.
How do you check current draw, temperature and noise?
These three readings catch a failing pump before its vacuum drops.
- Current draw. Measure with a clamp meter during the pump-down and at blank-off. For DC pumps the arithmetic is simple: an HC480D rated at ≤200 W on 24 V should draw no more than about 8.3 A. For AC pumps, compare with the nameplate current or a known-good pump, since the motor's power factor makes current higher than watts divided by volts. Check the voltage at the pump terminals at the same time: long DC cables lose voltage under load. More on wiring in our electrical integration guide.
- Temperature. After the pump has run long enough to stabilise, measure the cylinder head and motor housing with an infrared thermometer or a contact probe, and note the ambient temperature. A pump that runs hotter than before, at the same duty and ambient, needs a look at cooling, ventilation and load.
- Noise. Datasheet noise is given in dB(A) at a stated distance, for example ≤60 dB(A) for the HC580A at 50 Hz. Measure at the same distance and conditions. A new knock, rattle or rising hum matters more than the number: it often means loose mounting, a worn bearing or a broken valve reed.
Why is my vacuum pump not pulling enough vacuum?
Put the test results together and most problems point to a few causes.
| Symptom | Likely cause | How to confirm |
|---|---|---|
| Blank-off reading short of spec, slow pump-down | Worn piston ring or seal, leaking valve plate or head gasket | Blank-off test at the pump inlet at rated voltage |
| Pump reaches spec blanked off, machine does not | Leaking fittings, hoses, seals, valves or suction cups | Rate-of-rise test; isolate sections one by one |
| Good blank-off reading, slow pump-down, low flow | Blocked inlet filter, narrow or long hose, partly closed valve | Flow check with and without the filter and hose |
| Slow, weak or stalls; trips the supply | Wrong voltage or frequency, voltage drop on DC cables, motor or bearing fault | Measure voltage at the terminals under load; check current |
| Runs hot, current higher than before | Poor ventilation, high ambient, restricted exhaust, worn bearings | Temperature and current against your baseline |
| New knocking, rattling or vibration | Loose mounting, worn bearing, broken valve reed | Inspect mounts; listen at the head and motor |
| Vacuum falls off soon after the pump stops | Air returning through the pump, or a system leak | Rate-of-rise test with the isolation valve closed |
| Reading falls short only on some days or sites | Barometric pressure or altitude, not the pump | Read an absolute gauge, or correct for the barometer |
Should you repair or replace a vacuum pump?
The parts of an oil-free piston pump that wear are the piston ring or seal, the valve plates and reeds, gaskets, bearings and the inlet filter. If tests show the motor and cylinder are sound and only these parts are worn, replacing them usually brings the pump back to its blank-off figure.
Replacement makes more sense when the motor winding has failed, the cylinder bore is scored, several problems appear at once, or the downtime of repeated repairs costs more than a new pump. For OEM fleets, make sure the replacement matches the original configuration, voltage and version exactly: our replacement parts and configuration control checklist explains why.
Whichever you choose, record a baseline on every new pump: blank-off reading, pump-down time into your test vessel, current, temperature and noise. Later tests are only meaningful against that baseline. Our startup and commissioning checklist shows where those readings fit into acceptance.
Common questions
How do I check that a vacuum pump is working?
Fit a vacuum gauge directly to the inlet, run the pump for a few minutes and read the lowest pressure. If it reaches the ultimate vacuum on its datasheet, allowing for the day's barometric pressure, the pump is working.
Can I test a vacuum pump without a gauge?
Only roughly. A hand over the inlet tells you the pump is moving air, not how deep it pulls or how fast. Any judgement about wear needs a gauge reading compared with the spec.
Why does my gauge show less vacuum than the datasheet?
Check the simple causes first: low barometric pressure or altitude, gauge accuracy, low supply voltage, a hose or filter between pump and gauge, or the wrong version (standard instead of high-vacuum). If a direct blank-off reading at the right voltage is still short, suspect wear.
For background on how these pumps work and how to read a spec sheet, see how vacuum pumps work, or browse our oil-free vacuum pumps. If you are testing pumps for a new machine, send us your flow, vacuum level, supply voltage and duty cycle and our engineers will suggest a model, including OEM versions for your equipment.