As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
A tracer gas test needs a recovery route for any remaining charge, a vacuum pump and a micron gauge, a cylinder of the five percent hydrogen and ninety five percent nitrogen mixture with a regulator to suit it, hoses and adapters that reach the service ports, and a gauge or manifold that can be read over a timed hold. The handheld detector is the last item in that sequence and the only one most listings sell you. Detector bullets routinely note that compatible tracer gases are supplied separately, which is an accurate description of where the rest of the equipment list begins.
The cylinder, the regulator and the pressure you are allowed to use
The mixture itself is the item people forget to order. It is a prepared blend rather than something assembled on site, because the hydrogen fraction has to sit well below any burnable concentration and has to stay consistent so that detector response means the same thing on every job. The cylinder arrives with a valve outlet thread that the regulator must match, and the regulator has to deliver a steady outlet pressure in the range the system under test allows rather than simply reducing from cylinder pressure to something approximate.
That allowable range is the part worth checking before anything is connected. Every sealed system has a weakest rated component, usually on the low side, and the equipment manufacturer publishes a maximum test pressure for it. The test pressure you choose sits under that limit. It is tempting to run high because a larger pressure differential drives more gas through a small path and makes the sweep easier, and that temptation is exactly how an intact flare or a tired service valve becomes a second fault to repair.
Hoses and adapters complete the delivery side. They have to seal reliably themselves, because a leaking test hose produces a signal at the detector that has nothing to do with the system, and a hose that leaks slowly produces a pressure decay that looks like a fault in the equipment.
Evacuation comes before the tracer gas, not after
Any refrigerant still in the system is recovered first, into a recovery machine and a cylinder rated for it. That step is a legal obligation in most jurisdictions and it is also practical, because residual halogen in the circuit contaminates nothing in a hydrogen test but residual pressure makes the fill unpredictable.
Evacuation then does two jobs. It removes the air that would otherwise dilute the five percent mixture into something the detector struggles to resolve, and it boils off water vapour that would stay in the system and form acid once oil and refrigerant are reintroduced. A vacuum pump sized for the system and a micron gauge are both required: the pump creates the vacuum and the gauge is the only thing that reports whether the vacuum is deep enough and whether it holds when the pump is isolated. A vacuum that rises steadily after isolation is itself a leak indication, found before a single cubic foot of tracer gas has been used.
A pressure hold and a leak sweep answer different questions
With the mixture in and the system isolated, the gauge becomes the instrument. A hold over a measured period tells you whether pressure falls and how quickly, which is a rate. It does not tell you where. The detector sweep tells you where, and it is slow work that you only want to do once, so the hold is what decides whether the sweep is needed at all.
The hold has one trap in it. Gas pressure in a fixed volume moves with temperature, so a system that cools after the sun goes off it will show a pressure drop that has nothing to do with a leak. The correction is the gas law, and in practice it means recording the ambient temperature at the start and the end of the hold and treating any drop smaller than the temperature can explain as a pass.
| Item | What it answers | What it cannot tell you |
|---|---|---|
| Recovery machine and cylinder | Whether the system is legally and physically empty | Anything about the integrity of the circuit |
| Vacuum pump and micron gauge | Whether moisture and air are out, and whether vacuum holds | The location of any leak the rising vacuum revealed |
| Tracer gas cylinder and regulator | Delivers a known mixture at a chosen test pressure | Whether that pressure is within the equipment limit |
| Gauge or manifold over a timed hold | Whether pressure falls, and at roughly what rate | Which joint is responsible |
| Handheld hydrogen detector | Which joint is responsible | The total loss rate of the system |
What the service kits put in the box
Packaged kits move part of that list into one case. The Robinair LD9-TGKIT service kit is sold as the tools needed to use tracer gas in place of refrigerant on vehicle air conditioning, and its bullets describe a detector built for the five percent hydrogen and ninety five percent nitrogen mixture, a full colour display with message and error screens, automatic calibration and reset to ambient air, a true mechanical pump, three ultraviolet inspection lights at 395 to 415 nm optimised for dye fluorescence, a high intensity inspection light and an audio mute.
The detector sold on its own, the Robinair LD9-TG, publishes the numbers that matter for the sweep: leaks as small as 0.05 ounces per year for R-134a and 0.015 ounces per year for R-1234yf, compliance with SAE J2970, a sweep mode, a 128 by 160 colour display and a sensor life quoted at ten years. The same bullet list states that it is compatible with most hydrogen based tracer gases and that those are sold separately, which is the sentence this whole article exists to unpack.
Not every listing is that forthcoming. The Testo 0632 0336 gas leak detector carries no feature bullets at all on the listing supplied, so its sensitivity, its response time, its sensor type and its gas compatibility are all: not stated in the listing bullets. The Inficon 718-202-G1 combustible gas detector publishes four bullets and all four describe the package rather than the instrument, giving a category, a country of manufacture, a manufacturer name and a carton measuring 15 inches by 3.9 inches by 8.4 inches. Both may be perfectly capable instruments. Neither listing lets you check that before it arrives, and the manual is the only place the answer exists.
Building the rest of the bench around the detector
Treat the detector as one component in a chain that has to be verified end to end. Before the job, confirm the instrument responds to a known source rather than assuming it will, using the procedure in how to test a gas leak detector, and set a recurring interval for that check rather than doing it when something feels wrong. The reasoning behind choosing an interval is in how often to test a gas detector.
Add the safety equipment the workspace needs rather than the one the test needs. Nitrogen displaces oxygen in a confined space, and hydrocarbon refrigerants are increasingly common on the same bench, so a separate instrument that responds to combustible gas is worth having nearby even though it plays no part in the tracer measurement. The difference between that instrument and a refrigerant specific one is set out in what a combustible gas detector is, and the fixed sensing technology used for continuous area monitoring is described in infrared gas detection.
Finally, write the result down as a pair of facts rather than a verdict. The hold result with its start and end temperatures, and the sweep result with the sensitivity setting used. A test recorded that way can be repeated by someone else, and a test that cannot be repeated has proved nothing about the system.
Frequently Asked Questions
Is the tracer gas supplied with the detector?
Almost never. Detector listings commonly state that compatible hydrogen based tracer gases are sold separately, because a pressurised cylinder is a different kind of item to ship and to stock. Plan for the cylinder, a regulator that suits its valve thread, and a hose set with fittings that match the service ports you work on.
What pressure should a tracer gas test run at?
The limit is set by the weakest rated component in the circuit, which is usually the low side of the system or a service valve rather than the pipework. Equipment manufacturers publish a maximum test pressure and that figure governs. Pushing higher does make a small leak easier to find and also risks opening a joint that was intact, which turns a diagnosis into a repair.
Do I need a vacuum pump if the system is already empty?
Open to atmosphere is not the same as evacuated. A system that has been standing contains air and water vapour, and both dilute the tracer mixture and leave moisture behind for the charge that follows. Pulling a deep vacuum before the mixture goes in protects the result and the system, and a micron gauge is what tells you the vacuum was deep enough.
Can I reuse the tracer mixture after the test?
No, and there is no mechanism to. The mixture is vented or displaced when the test ends, which is one of the reasons the method is attractive in the first place: the gas being released is nitrogen with a trace of hydrogen rather than a refrigerant with a global warming potential attached to it.
Does a pressure hold replace the detector sweep?
They answer different questions and you normally want both. A hold tells you whether the system loses pressure and roughly how fast, which decides whether there is anything to look for. The sweep tells you where. A hold that passes removes the need for the sweep entirely, which is why it goes first.
Why do some service kits include ultraviolet lights?
Because the same technician also works with fluorescent dye, which is a separate leak finding method for systems that can be charged and run. A kit that bundles ultraviolet inspection lamps with a hydrogen detector is covering two methods in one case rather than using the lamp as part of the tracer gas test.