Most refrigerant sniffers sold to HVAC techs are halogen detectors, and halogen sensors are blind to hydrocarbons. R290 and R600a are hydrocarbons. If you point a standard freon sniffer at a propane or isobutane circuit, it will read clean while the charge leaks out. The seven picks below are the ones whose listings state hydrocarbon coverage, and the page explains what to check before you buy.
Measure It
Compare It
Buy Once
As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
This comparison was assembled from the manufacturer listings supplied for each product. Only products whose listings state or strongly imply hydrocarbon coverage were included as picks for R290 and R600a work; the Inficon 718-202-G1 (P1), Inficon 705-202-G1 (P6), Elitech ILD-200 (P5), Elitech ELD-270's sibling listing (P4) and the RectorSeal Better Bubble (P10) were excluded because their listings describe halogen-only.
A heat diode sniffer whose listing explicitly covers A3 refrigerants R290 and R600.
Heated diode unit with the tightest stated sensitivity here and explicit R290 and R600a coverage.
Reads both ppm for locating and %LEL for flammable-atmosphere awareness.
Functionally the same instrument as the Simbow WJL-6000PRO under a different brand listing.
Heated diode unit with a stated 500 hour sensor life and R290 and R600A named in the title.
Metal-oxide sensor with a stated 24 hour battery and a combustible gas claim.
Corona sensor unit with a long probe, but the listing does not name R290 or R600a.
No pick matches that combination yet. Try a wider budget or a different measurement.
| Pick | Sensor type | R290 / R600a stated | Stated sensitivity | Readout | Battery | |
|---|---|---|---|---|---|---|
| Heat diode | Yes, A3 R290 and R600 named | 0.14 oz/yr adjustable | Concentration on 2.8 inch TFT | Rechargeable, runtime not stated | View | |
| Heated diode | Yes, R600A named | 0.07 oz/yr | 0 to 9 leak level | Not stated in the listing bullets | View | |
| Not stated in the listing bullets | Not named in supplied bullets | Not stated in the listing bullets | PPM and %LEL | Not stated in the listing bullets | View | |
| Not stated in the listing bullets | Not named in supplied bullets | Not stated in the listing bullets | PPM and %LEL | Not stated in the listing bullets | View | |
| Heated diode | Yes, named in title | 0.05 oz/yr max, 0.1 oz/yr normal | LCD leak level with peak hold | 3000mAh, 8 hours stated | View | |
| Metal-oxide semiconductor | Not named in supplied bullets | Not stated in the listing bullets | High and Low sensitivity, no numeric scale stated | 2500mAh, 24 hours stated | View | |
| Corona | No, halogenated refrigerants only | 0.21 oz/yr | Not stated in the listing bullets | Not stated in the listing bullets | View |
Sensitivity figures are quoted exactly as the listings publish them and are not directly comparable across sensor types. A detector that does not name R290 or R600a in its listing should not be relied on for those gases without written confirmation from the seller.
Six things to settle before you buy, in the order they will affect whether the tool works on the job.
Halogen sensors look for chlorine or fluorine. R290 and R600a have neither. Buy a heated diode, infrared or metal-oxide sensor and make sure the listing names the gas.
A locating tool tells you where the leak is. A %LEL readout tells you whether the air around you is approaching a flammable concentration. On R290 these are different jobs.
A stated maximum sensitivity is not the everyday figure. Where a listing gives both, as P9 does, the normal setting is the number to plan around.
A long flexible probe reaches compressor compartments and evaporator boxes. A metal wand survives being pushed into them. Warm-up time tells you how long you stand around before the first reading is valid.
Rechargeable is now standard, but runtime varies widely. A stated 8 hour figure covers a normal day; a stated 24 hour figure covers a long one. Disposable cells are a mid-job failure waiting to happen.
Heated diode and metal-oxide sensors are consumables. A stated life, like the 500 hours on P9, lets you schedule replacement instead of discovering drift on a call-back.
Most sniffers sold for HVAC work are halogen detectors. They respond to the chlorine or fluorine atom in refrigerants like R410A and R134a. R290 is propane and R600a is isobutane; neither contains a halogen, so the sensor has nothing to react to. The instrument reads clean because it genuinely cannot see the gas, not because the system is tight.
A combustible gas detector will respond to R290, because propane is a combustible gas. But it is built to warn you about a flammable atmosphere, not to pinpoint a 2 gram per year weep at a flare. The two jobs need different sensitivity and different probe behaviour. Some tools here, like P7 and P8, offer both modes in one housing, which is the practical compromise.
LEL is the lower explosive limit, the concentration in air below which a gas will not ignite because there is not enough fuel. %LEL is the reading expressed as a percentage of that threshold. A reading of 10 percent LEL means you are a tenth of the way to a flammable mixture. It is a safety number, not a leak size.
Both respond to hydrocarbons. Heated diode sensors are generally more sensitive to small concentrations, which suits leak locating. Infrared sensors are less prone to contamination and drift, which suits repeatable readings over a long service life. The listings here do not publish a direct comparison, so choose on the stated sensitivity and sensor life you can actually verify.
It depends on the sensor type and how much contamination it has seen. Most listings here do not publish a life figure. P9 states over 500 hours for its heated diode sensor, which is the only concrete number in this set. Where no figure is given, treat the sensor as a consumable and replace it when readings become erratic or fail a known-leak check.
No. A detector can find a leak, but it cannot prove absence. Probe speed, warm-up state, background contamination and sensor drift all affect whether a leak is detected. The listing for P11 gives the most useful guidance in this set: move the probe no faster than 1 inch (25 mm) per second. Even then, a clean sweep is evidence, not proof.
The ALD-200 is the only pick here whose listing names R290 and R600 as covered refrigerants rather than leaving the reader to infer it from a generic 'all refrigerants' phrase. That matters because the sensor type is the whole game. A heat diode sensor works by heating a ceramic element and measuring the change in conductivity when a gas breaks down on its surface. Hydrocarbon molecules break down on that element, so the sensor responds. A halogen sensor looks for the chlorine or fluorine atom instead, and R290 and R600a contain neither, which is why a halogen sniffer passes over a propane leak without a sound.
The listing quotes sensitivity adjustable up to 0.14 oz/yr, which is the practical floor for finding a weeping flare or a pinhole in a brazed joint before it becomes a call-back. The 16 inch probe is long enough to reach a compressor compartment on a reach-in or a rooftop unit without putting your face into the refrigerant stream, and the rechargeable battery removes the disposable cell problem that catches techs mid-job.
The display shows concentration rather than a bare yes or no, and the beep rate rises with it. That lets you sweep a joint and watch the number climb as you close in, which is faster than the hunt-and-peck you do with an alarm-only unit. The mute is a genuine convenience indoors, but remember that muting the buzzer does not mute the display, and it does not change the sensor's response time.
The ELD-270 lists R600A by name and includes HCs in its coverage string, which is the signal you are looking for. Its stated 0.07 oz/yr sensitivity is the tightest figure published across this set, and for R290 work that headroom is useful because hydrocarbon charges are small. A domestic fridge might hold 50 to 150 grams of R600a in total. A leak of 2 grams a year is a slow weep, but on a charge that size it is still worth finding before the system short-cycles.
The 0 to 9 leak level display gives you a relative scale rather than a concentration in ppm, so you cannot use it to judge whether an atmosphere is approaching a flammable limit. It is a locating tool, not a safety monitor. The three sensitivity levels let you sweep broadly on the low setting and then drop to the high setting once you have a direction, which reduces the false triggers you get from background solvent or exhaust fumes.
Automatic and manual zeroing is the feature most people ignore until they need it. Automatic zeroing drifts the baseline as you work, which is fine in clean air but can mask a slow leak in a contaminated space. Manual zeroing lets you fix the reference point yourself. The 30 second warm-up and 3 second response are both stated, and both are realistic for a heated diode element.
The WJL-6000PRO is the only tool in this set whose listing advertises a %LEL mode, and that changes what the instrument is for. PPM mode is a locating mode: you sweep a joint and watch the number climb as you close in. %LEL mode is a safety mode: it tells you what fraction of the lower explosive limit is present in the air around the probe. For R290, which is flammable, that second number is the one that decides whether you keep working or ventilate and step back.
One caution on the coverage claim. The listing says it sniffs CFCs, HCFCs, HFCs, HFOs, blends and PFCs. It does not name R290 or R600a in the bullet text supplied. The %LEL mode is the strongest hint that hydrocarbon response is intended, since %LEL is only meaningful for flammable gases, but a buyer should confirm hydrocarbon coverage with the seller before relying on this as the primary R290 instrument. If the seller confirms it, the dual readout makes it the most informative tool here.
The metal probe is a real durability point. Plastic probe tips snap when they are pushed past a fan shroud or into a tight evaporator box, and a snapped tip usually means a whole new instrument. A metal wand that bends and holds its shape survives that treatment. The mute is handy indoors, and the alarm tone changes to a siren rather than just speeding up, which is easier to notice when you are head-down in a cabinet.
This is the same instrument as P7 with a different brand name on the listing. The feature set is identical: PPM and %LEL modes, a metal probe, audible and visual alarms with mute. If you are choosing between them, choose on availability and on which seller will actually answer a question about hydrocarbon coverage, not on any imagined difference in performance. There is no basis in the supplied listings for claiming one is more sensitive than the other.
The one wording difference worth noting is that this listing frames the %LEL mode around A2L refrigerant safety rather than around flammability generally. A2L refrigerants are mildly flammable; A3 refrigerants like R290 are more so. The %LEL principle is the same either way, but the listing's own framing points at A2L, so a buyer who needs it specifically for R290 should treat the coverage as something to confirm rather than assume.
As with P7, the metal probe is the durability feature that separates this from the plastic-wand crowd, and the mute is a convenience rather than a capability. Neither listing publishes a sensitivity figure in oz/yr or a sensor type, so if you need a documented sensitivity number to satisfy a customer or a procedure, this is not the tool that gives you one.
The LD200 is the only pick here that publishes a sensor life figure: over 500 hours. That number matters because heated diode sensors are consumable. They degrade with use and with contamination, and a sensor that has drifted will under-read a small leak without telling you. A stated life gives you a replacement interval you can plan around instead of discovering the problem on a job.
The stated sensitivity is 0.05 oz/yr at maximum and 0.1 oz/yr at normal setting. Those are the tightest numbers in this set, and the listing is honest that the headline figure is the maximum rather than the everyday one. For R290 and R600a work, where the whole charge is small, that resolution is the difference between finding a weep and writing it off as background.
The peak value function is more useful than it sounds. When you sweep a joint, the reading rises and falls as the probe passes over the leak. Holding the peak lets you go back and confirm the same spot gives the same number, which is how you separate a real leak from a transient puff of solvent. The USB-C port and the stated 8 hour runtime cover a full shift, and the 3000mAh battery is a figure you can compare against other tools rather than a vague claim.
The SENSYX is the odd one out here because it is the only pick whose sensor type is metal-oxide rather than heated diode or infrared. Metal-oxide sensors respond to a broad range of gases, which is why the listing can call it a combustible gas detector as well as a refrigerant sniffer. That breadth is an advantage for R290 and R600a, because hydrocarbons are exactly the kind of gas a metal-oxide sensor is good at. It is also a disadvantage, because a sensor that responds to everything will respond to cleaning solvent, exhaust and alcohol too.
The listing is unusually candid about the warm-up. Metal-oxide sensors need to run hot for a period before they stabilise, and the listing says several hours in clean air when new or after storage. That is not a defect, it is how the chemistry works, but it means you cannot unbox this and trust the first reading. It also means the stated 24 hour runtime has to be read alongside that burn-in requirement.
The probe speed advice in the listing, no faster than 1 inch (25 mm) per second, is the single most useful sentence on the page. Every sniffer in this category has a response time, and moving the probe faster than the sensor can respond is the most common reason a tech declares a system leak-free when it is not. Whatever detector you buy, slow down.
The CLD-100 is included here as a caution rather than a recommendation for hydrocarbon work. Its listing describes a corona sensor and lists only halogenated refrigerants: HFCs, CFCs, HCFCs and blends. R290 and R600a are hydrocarbons, not halogenated, and they are not named. A corona sensor detects the halogen content of a gas, so a hydrocarbon leak gives it nothing to react to. On an R290 system this instrument is not a leak detector, it is a decoration.
That is not a criticism of the tool on its own terms. For R410A, R134a and R404A work, a corona sensor is a proven approach, and the stated 0.21 oz/yr sensitivity and six sensitivity levels are reasonable for general service. The 8.26 inch flexible probe reaches into places a rigid wand will not. The 6 second preheat is quick.
But if you are reading this page because you have R290 or R600a equipment, this is the wrong instrument, and the listing itself tells you so by omission. The absence of a hydrocarbon claim is the signal. Do not buy a halogen sniffer and hope; buy a sensor type that responds to hydrocarbons, and confirm the coverage in writing before you spend.
Short answer: For R290 and R600a work, the deciding factor is sensor chemistry, not brand. The Aprvtio ALD-200 (P2) is the most complete pick here because its listing states a heat diode sensor covering A1, A2L and A3 refrigerants including R290 and R600, with sensitivity adjustable to 0.14 oz/yr. The Elitech ELD-270 (P3) is the strongest value alternative, with a heated diode sensor, 0.07 oz/yr stated sensitivity and three sensitivity levels. If you want a detector that also reports concentration in ppm and %LEL rather than just alarming, the Simbow WJL-6000PRO (P7) and the Aprvtio WJL-6000PRO (P8) are functionally the same tool under two listings. The Inficon 705-202-G1 (P6) and the Inficon 718-202-G1 (P1) are halogen-only units; their listings do not claim hydrocarbon detection, so they should not be your primary instrument on an R290 system.
A halogen leak detector works by looking for a halogen atom. Chlorine and fluorine are the halogens that appear in the refrigerants the HVAC trade has used for decades: R22, R410A, R134a, R404A, R1234yf. The sensor, whether it is a corona discharge, a heated diode or an infrared cell, is tuned to react when those atoms pass over it. That is why a freon sniffer is so good at its job, and it is exactly why it fails on a hydrocarbon.
R290 is propane. R600a is isobutane. Both are hydrocarbons, made of carbon and hydrogen only. There is no chlorine and no fluorine in either molecule, so a halogen sensor has nothing to detect. Point a standard sniffer at a leaking R290 flare and it will sit silent. The tech concludes the joint is tight, tops up the charge, and the call-back is booked before the van leaves the street.
This is not a sensitivity problem you can fix by buying a more expensive halogen detector. It is a chemistry problem. The instrument is answering a different question from the one you asked. The fix is to choose a sensor type that responds to hydrocarbons: a heated diode, an infrared cell or a metal-oxide semiconductor. All three appear in the picks above. If you want the longer version of that sensor argument, it is covered in infrared versus heated diode refrigerant leak detectors.
Sensitivity is quoted in ounces per year, which is a mass flow rate. A stated 0.07 oz/yr means the sensor can register a leak that loses about two grams of refrigerant over twelve months. That sounds absurdly small, and it is, which is why the figure is worth reading carefully.
Two things distort the number. First, many listings quote the maximum sensitivity, achieved on the highest sensitivity setting, in still air, with a freshly calibrated sensor. The everyday figure is usually looser. P9 is unusually honest here: it states 0.05 oz/yr maximum and 0.1 oz/yr normal, so you can see the gap. Second, the figure is a detection threshold, not an accuracy. A detector that trips at 0.07 oz/yr does not tell you the leak is 0.07 oz/yr; it tells you it found something at least that big.
For R290 and R600a the small numbers matter more than they do on a large A1 system. A domestic refrigerator might hold 50 to 150 grams of R600a in total. A leak of a few grams a year is a slow loss, but on a charge that size it is a meaningful fraction of the whole. A detector that only trips at 0.25 oz/yr will miss it. That is the practical reason the tighter stated figures in this set are worth paying attention to.
The most common failure in leak detection is not a bad instrument. It is probe speed. Every sensor in this category has a response time, and if you move the probe past the leak faster than the sensor can react, you get a clean reading over a real leak. The listing for P11 states the rule plainly: no faster than 1 inch (25 mm) per second. That is slower than most people naturally sweep.
The second mistake is skipping the warm-up. Heated diode sensors need 30 seconds or so to stabilise, which most listings state. Metal-oxide sensors need far longer, and the P11 listing says several hours in clean air when the unit is new or has been in storage. Starting a job before the sensor is ready gives you a baseline that drifts while you work, and a drifting baseline hides slow leaks.
The third mistake is zeroing in contaminated air. If you zero the instrument next to a solvent rag or in a room full of exhaust, the baseline is set wrong and every subsequent reading is offset. Zero in clean air, away from the system, then walk to the leak. The automatic and manual zeroing modes on P3 exist precisely so you can take control of this rather than letting the instrument decide.
Heated diode and metal-oxide sensors are consumables. They degrade with use, and they degrade faster when they are exposed to contamination: silicone sealants, compressor oil mist, cleaning solvents, exhaust. A sensor that has drifted does not fail loudly. It under-reads. You sweep a joint, get a low number, and conclude the leak is small or absent when the sensor has simply lost its edge.
The defence is a known-leak check. Keep a small cylinder or a charged fitting with a deliberate, controlled leak, and confirm the instrument still trips on it before you trust a clean sweep. Do this at the start of a job, not after you have already declared the system tight. P9 is the only pick here that publishes a sensor life figure, over 500 hours, which gives you a replacement interval to work to. Where no figure is published, treat the sensor as a consumable and replace it when the known-leak check starts to look weak.
Probe care matters too. A plastic wand that snaps leaves you with an instrument you cannot use. The metal probes on P7, P8 and P9 are built for the way techs actually work, bending into tight evaporator boxes and engine bays without breaking. Wipe the probe after use and store it capped.
A leak detector finds leaks. It does not make a flammable atmosphere safe, and it does not replace a gas monitor. If you are working on an R290 system in an enclosed space, the question you need answered is not where the leak is but whether the air around you is approaching a flammable concentration. That is a %LEL measurement, and only P7 and P8 offer it in this set. Even then, a handheld sniffer is not a fixed gas detection system and should not be treated as one.
Nor is a leak detector a substitute for pressure testing or for a proper evacuation and charge procedure. A sniffer helps you find a leak you already suspect. It is poor at proving a system tight, because a clean sweep has too many ways to be wrong. If you need to certify a system, use the detector to locate and a pressure or vacuum test to confirm.
And if your work is entirely on A1 or A2L refrigerants, you do not need a hydrocarbon-specific tool at all. A good halogen detector will serve you better and often costs less. The hydrocarbon requirement is specific, and it is worth buying for only when you actually have R290 or R600a equipment in front of you. For the broader category, see best refrigerant leak detectors, and for the A3 context specifically, A3 refrigerant leak detectors and compatible tools. If you are weighing a sniffer against a combustible gas monitor for general work, refrigerant leak detector versus combustible gas detector sets out where each one stops being the right answer.
This comparison was assembled from the manufacturer listings supplied for each product. Only products whose listings state or strongly imply hydrocarbon coverage were included as picks for R290 and R600a work; the Inficon 718-202-G1 (P1), Inficon 705-202-G1 (P6), Elitech ILD-200 (P5), Elitech ELD-270's sibling listing (P4) and the RectorSeal Better Bubble (P10) were excluded because their listings describe halogen-only detection or, in the case of the bubble solution, a different method entirely. Where a listing does not publish a figure, the table says so rather than estimating. No laboratory testing was performed and no measurements of our own are published.
A comparison of seven ultrasonic leak detectors for HVAC technicians, covering frequency range, audio and visual output, and the jobs each one suits.
Read Guide→
Six CO meters for furnace servicing, compared on range, resolution, alarm thresholds, calibration data and logging, with the limits of each stated plainly.
Read Guide→
Seven calibration kits and components compared on blend, cylinder size, regulator type, and what the listing does and does not publish.
Read Guide→