As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
Switch to tracer gas when the system has no refrigerant left in it to detect, when you have just opened it for a repair, or when the leak is slower than the sniffer can resolve in the air you are working in. A halogen sniffer finds refrigerant molecules, so refrigerant has to be escaping at the moment you sweep the joint. A tracer gas test puts a known inert mixture into an empty system and looks for that instead, which means the dividing line between the two methods is the state of the system in front of you rather than the quality of the instrument in your hand.
The question is whether there is anything left in the system
Every halogen sniffer works the same way at the level that matters here. Air is drawn across a sensing element, the element responds to halogenated molecules passing over it, and the instrument converts that response into a tone, a light or a number. Nothing in that chain functions if the refrigerant is already gone. A unit that lost its charge over a season, a line set that has been cut and brazed, a compressor that has just been swapped: in all three cases there is no halogen at the joint to find, and the sniffer will sweep the whole system in silence.
The usual workaround is to put a partial charge in and go looking. That wastes refrigerant, it vents the thing you are trying to contain, and on a system with several suspect joints it can take two or three charges before the source is found. A tracer gas test removes the refrigerant from the question entirely. The system is evacuated, pressurised with a mixture of five percent hydrogen in ninety five percent nitrogen, and swept with a detector that responds to hydrogen. The pressure can be set by the test rather than by the operating condition, which is why a slow leak that only weeps at working pressure can be pushed into revealing itself. The method and its pressure limits are set out in more depth in tracer gas leak detection explained.
Sensitivity figures the halogen sniffers actually publish
Where refrigerant is present, the sniffer is the faster instrument, and the listings put a number on how small a leak it will catch. The Elitech WJL-6000S states seven adjustable sensitivity levels set by a knob, a maximum sensitivity quoted at 0.21 oz/yr, a response within three seconds and a warm up of six seconds. The Aituzero rechargeable detector states three sensitivity levels and a sensitivity figure of 3 g/yr, which it also writes as 0.1 oz/yr, with a full colour display that shows a live parts per million reading, a funnel view for intensity and a trend graph for watching the level change as the probe moves.
The Aprvtio WJL-6000PRO takes a different approach to the same problem by publishing two measurement modes rather than a finer threshold: parts per million for small leaks and percent of the lower explosive limit for watching the safety margin around the mildly flammable A2L refrigerants. Its bullets quote a two second response and refrigerant identification within ten seconds, with an all metal probe in place of a plastic tip. None of the three publishes how long the quoted sensitivity survives in service, so for sensor life the honest entry is: not stated in the listing bullets.
The SENSYX rechargeable detector is unusual in spending a whole bullet on technique instead of on a number. It asks the operator to leave a new or stored instrument running in clean air for several hours to reach full sensitivity, to switch on away from the leak, and then to move the probe no faster than one inch per second at about a quarter of an inch from the joint with fans and drafts off. That instruction is worth more than a sensitivity figure, because it describes the conditions under which any published figure was obtained.
| Instrument | Sensitivity stated | Response and warm up stated | Sensitivity control |
|---|---|---|---|
| Elitech WJL-6000S | 0.21 oz/yr at maximum | Response within 3 s, warm up 6 s | Seven levels on a knob |
| Aituzero rechargeable | 3 g/yr, also written 0.1 oz/yr | Not stated in the listing bullets | Three levels, high, medium and low |
| Aprvtio WJL-6000PRO | Not stated as a leak rate | Response in 2 s, identification in 10 s | Parts per million and percent LEL modes |
| SENSYX rechargeable | Not stated in the listing bullets | Several hours to full sensitivity when new | Selectable high and low |
Hydrogen behaves differently in the air around a joint
Hydrogen is the smallest molecule there is, and the physical consequences all help a leak test. It passes through a restriction that a bulky halogenated molecule negotiates slowly, so a path that only weeps refrigerant can pass hydrogen at a rate a detector resolves. It is also far lighter than air, so it rises and disperses within seconds of leaving the hole rather than settling into a cloud. That is the opposite of refrigerant behaviour, and it is why a hydrogen sweep tends to point at the joint rather than at a region near the joint.
Safety follows from the mixture rather than from careful handling. Five percent hydrogen in nitrogen contains no oxygen, and any air it meets dilutes the hydrogen further, so the released gas moves away from a burnable composition rather than towards one. The nitrogen carrier is dry, which matters because moisture left in a refrigeration circuit reacts with oil and forms acid. A pressure test with a dry inert mixture leaves the system in the condition you want before evacuation and charging.
The detector itself is a different instrument, not a sniffer with a new label, and the units built for this work are collected in the guide to tracer gas leak detectors.
Where a sniffer still finds the leak first
On a charged system that is running and losing refrigerant slowly, the sniffer wins on time. There is no recovery, no evacuation, no cylinder and no pressurisation step: the instrument warms up and you sweep. Warm up is short on these units but it is not optional, and the detail of why the first minute matters is covered in gas sniffer warm up time.
Selectivity is the second reason to keep one. A halogen sniffer that responds only to halogenated compounds will ignore the solvents, exhaust and fuel vapour that sit around a workshop floor. That is a real advantage over a broad band instrument, and it is also the boundary that trips people up when hydrocarbon refrigerants appear, since propane and isobutane carry no halogen at all. The SENSYX listing is explicit that its SFD201 variant adds a response to hydrocarbons and combustible gases including methane, ethane, propane and isobutane, while both variants cover the halogenated family. The distinction between the two instrument classes is set out in refrigerant leak detector against combustible gas detector.
The third consideration is age. A sniffer that alarms on a deliberate test leak is proving only that it responds to a large concentration, not that it still meets the ounces per year figure it was sold with. Sensing elements are consumable, and the signs that one has reached the end are described in when to replace a gas detector.
Choosing the method before you open the cylinder
Work through the system state first. If refrigerant is still in it and the loss rate is high enough to build a detectable concentration, sweep with the sniffer on its highest sensitivity in still air and slow the probe right down. If the system is empty, has been opened, or has already been swept without result, evacuate it and go to tracer gas rather than charging it to look again. If the suspect joints are in a bay where halogen has already been released, the sniffer is compromised by the background and tracer gas is the cleaner measurement regardless of the state of the charge.
Two habits improve either method more than a better instrument does. Record the sensitivity setting you used alongside the result, because a quiet sweep on a low setting is not the same finding as a quiet sweep on the highest one. And make the air still before you start, since both methods are defeated by a draft that carries the escaping gas away faster than the probe can gather it.
Frequently Asked Questions
Does a tracer gas detector respond to refrigerant as well?
Not usually, because it is built around a sensor that responds to hydrogen rather than to halogen. A hydrogen detector sweeping a charged system will stay quiet over a joint that a halogen sniffer would alarm on immediately. Instruments that claim both jobs are carrying two sensing principles, and the listing should say so explicitly rather than implying it.
Is a hydrogen and nitrogen mixture safe to put into a sealed system?
The standard mixture is five percent hydrogen in ninety five percent nitrogen, and it carries no oxygen of its own. Diluting it into room air can only lower the hydrogen fraction further, so the mixture cannot reach a burnable concentration on its own. Nitrogen is also dry and chemically inert, which is why it leaves nothing behind in the pipework.
How small a leak can a halogen sniffer actually find?
Listings quote a figure in ounces or grams per year at the highest sensitivity setting, and that figure assumes still air, a clean sensor and a slow sweep. The same instrument on a windy forecourt with a sensor that has been in storage will miss leaks an order of magnitude larger. Treat the published number as a laboratory ceiling rather than a field expectation.
Why does the sniffer alarm two inches away from the joint?
Refrigerant is heavier than air and lingers, so a leak builds a pool of vapour that extends well beyond the hole. Sweeping quickly through that pool triggers the alarm somewhere in the cloud rather than at the source. Slowing the probe to roughly an inch per second and working in still air shrinks the cloud to something that points at the joint.
Can I pressure test with nitrogen alone and skip the hydrogen?
You can, and a pressure hold with dry nitrogen is a genuine test, but it tells you only that the system loses pressure somewhere. It gives you a rate and not a location. The hydrogen fraction exists so that a handheld detector has something to follow back to the exact joint once the hold has told you there is a leak to find.
Does a sniffer lose sensitivity as the sensor ages?
Yes, and that is the quiet failure mode behind most missed leaks. Heated diode and metal oxide elements both drift as the sensing surface is poisoned or consumed, and the instrument keeps switching on and indicating readiness long after its real threshold has risen. Sensor replacement intervals belong in the service routine, not in the drawer with the manual.