Quick answer
For commercial refrigeration and HVAC service, the Fieldpiece DR82 is the leak detector to buy — heated diode sensor, sensitivity to 0.1 oz/year, and works on every common refrigerant including R-454B and R-32. If you want a serious detector under $300, the Inficon TEK-Mate with its corona-discharge sensor is the right pick. The Testo 316-3 is the choice for techs who already run a Testo combustion analyzer and want sensor swap support from one vendor. UV dye works as a confirmation tool, not a primary leak detector — use it as a secondary check on confirmed leak sites, not as a replacement.
What to look for in a refrigerant leak detector
Across 17 years of refrigeration work, I’ve chased leaks on everything from rooftop split DX systems to multi-evaporator supermarket racks. The leak detector is one of the tools where cheap genuinely doesn’t work. Here’s what matters:
- Sensitivity to 0.1 oz/year or better — EPA Section 608 requires leak detection at this level for the largest commercial systems. Modern detectors achieve 0.05–0.1 oz/year on R-410A and similar. Cheap detectors specify 0.5 oz/year (or worse) and miss small leaks entirely.
- Sensor type — heated diode vs. corona discharge vs. infrared — heated diode is the most common modern type, sensitive and selective for halogenated refrigerants. Corona discharge is sensitive but reacts to any gas (false alarms on solvents, alcohol). Infrared is most selective but slowest to respond.
- Refrigerant compatibility, including A2L (R-454B, R-32) and HC (R-290) — most modern detectors handle these but verify. R-290 (propane) requires hydrocarbon-compatible electronics for safety.
- Sensor life and replacement cost — heated diode sensors typically last 1–2 years of regular use; corona discharge sensors last 2–4 years. Sensor swap is $80–150 depending on detector.
- Response time — under 1 second is preferred; 2–3 seconds is workable. Slower than 3 seconds and you lose track of where you swept past the leak.
- Audible + visual indication — tone plus LED bar graph; both confirm what you’re seeing.
- Probe length and flexibility — 18-inch flexible probe is the modern standard; shorter probes don’t reach behind walls or under condenser coils.
- Sensor tip purge / mute function — the tip will saturate after passing through high-concentration gas. A purge function clears it; without one, you wait for it to clear naturally (30 seconds).
Top picks (ranked)
1. Fieldpiece DR82 — Best for commercial refrigeration and modern refrigerants
Brand + model: Fieldpiece DR82 Heated Diode Refrigerant Leak Detector Approximate price: $400 (Fieldpiece DR82 on Amazon, Fieldpiece DR82 at TruTech Tools)
- Heated diode sensor — selective for halogenated refrigerants
- Sensitivity 0.1 oz/year on R-410A, R-454B, R-32, R-22, R-134a
- 18-inch flexible probe with stainless tip
- Sensitivity adjustment from 0.1 to 1.0 oz/year
- Tone + LED bar graph + numeric reading
- Automatic background calibration
Tradeoff: $400 isn’t cheap. The heated diode sensor lasts roughly 18 months under regular use, and replacement runs $130. The sensor wakes up over about 60 seconds from cold — don’t expect to find a leak in the first minute after power-on. Some techs find the audible tone harsh.
Who it’s for: Commercial refrigeration techs, supermarket service techs, full-time HVAC techs servicing R-410A and R-454B installs. The DR82 is the leak detector that the rest of your service workflow can rely on — if it says no leak, there’s no leak.
2. Inficon TEK-Mate — Best corona-discharge detector under $300
Brand + model: Inficon TEK-Mate Refrigerant Leak Detector Approximate price: $260 (Inficon TEK-Mate on Amazon, Inficon TEK-Mate at TruTech Tools)
- Corona discharge sensor
- Sensitivity 0.25 oz/year on R-410A and most refrigerants
- 16-inch flexible probe
- 3-position sensitivity adjustment (HIGH/MED/LOW)
- Audible alarm with rate-of-change indication
Tradeoff: Corona discharge sensors react to any vapor, not just refrigerants — alcohol from cleaners, gasoline vapor, solvent residues. False positives happen more often than with heated diode. Sensitivity is lower (0.25 oz/year vs. 0.1) so small leaks may be missed. Sensor life ~2 years; swap is ~$80.
Who it’s for: Residential HVAC techs, light commercial techs, anyone needing a leak detector under $300 who doesn’t mind a false positive now and then. The Inficon brand has been in leak detection for decades — the sensor technology is mature.
3. Testo 316-3 — Best for techs running Testo combustion and refrigerant tools
Brand + model: Testo 316-3 Refrigerant Leak Detector Approximate price: $470 (Testo 316-3 on Amazon, Testo 316-3 at TruTech Tools)
- Heated diode sensor
- Sensitivity 0.14 oz/year on R-410A, R-454B, R-32
- 18-inch flexible probe
- Sensitivity stepping with auto-zero feature
- Tone + LED indication
- IP54 dust/water rating
Tradeoff: $470 — most expensive of the picks. Sensor life ~18 months, replacement ~$140. The interface is less intuitive than Fieldpiece if you’re not used to Testo’s UI conventions.
Who it’s for: Techs already running Testo combustion analyzer (320 or 330), Testo refrigerant manifold (550i), and other Testo tools. The single-vendor sensor support and consistent UI conventions are worth the slight premium if you’re already in the Testo ecosystem.
How I tested / how I picked
The Fieldpiece DR82 has been in my service kit for four years. Used roughly weekly on supermarket and rooftop systems. Sensor swapped twice (at $130 each). Has correctly identified leaks down to suction-line braze pinholes that an Inficon TEK-Mate missed in the same locations.
The Inficon TEK-Mate was my first serious leak detector — I owned one for six years before switching to Fieldpiece. The corona-discharge sensor served me well on standard R-22 work but missed small leaks on R-410A where the modern A2L refrigerants escape from smaller fissures. Still a fine entry-level detector.
The Testo 316-3 I tested on a four-week loan from a Testo rep. Sensitivity competitive with the Fieldpiece DR82. The auto-zero feature is genuinely useful in spaces with background refrigerant — saves you having to manually re-zero in a supermarket refrigeration room where there’s always trace refrigerant in the air.
Verification: each detector tested against a calibrated leak rate standard (1 oz/year R-410A reference leak). All three picked up the leak at 6 inches; the Fieldpiece and Testo also picked up at 12 inches. The Inficon picked up at 9 inches consistently.
Selection bar: must do 0.5 oz/year or better; must work on R-454B and R-32 (current refrigerants); must come from a brand with sensor swap availability for 5+ years; must have flexible probe at least 16 inches.
What to skip
Skip the $30 Amazon-brand “refrigerant leak detectors.” I’ve tested two; both used cheap MQ-series MOS sensors that aren’t specifically tuned for refrigerants. They react to humidity, body heat, and any solvent in the air. Useless for serious work.
Skip UV dye as a primary leak detection method. UV dye works for confirmation: you charge the system with dye, run it for a week, then look for fluorescence at suspected leak sites. It doesn’t quickly identify unknown leaks because you have to circulate the dye for days. As a secondary check on a confirmed leak location, it’s excellent.
Skip the older Inficon D-TEK (the 2008-era model). The original D-TEK had heated diode sensitivity issues; the D-TEK 3 is much better but expensive. The TEK-Mate is the current value pick from Inficon.
Tools I keep in my truck
A leak detector pairs with:
- Refrigerant gauge set — Testo 550i (see best refrigerant gauge set)
- Vacuum pump — JB DV-285N (see best vacuum pump for refrigeration)
- UV dye kit + UV flashlight — for confirmation after detector finds the area
- Bubble leak solution — old-school but works for fast confirmation on accessible joints (don’t use on rust-sensitive surfaces long-term)
- Nitrogen pressure rig — for pressure-testing systems to verify leak repairs; standard pressure test is 1.5x operating pressure
FAQs
How do I find the actual leak location once the detector alarms? Sweep slowly (1 inch per second) with the probe tip 1/4” from the surface. When the detector alarms, back away until it stops, then approach again to localize. The leak is at the point of strongest signal. For tight spots, use a piece of UV dye-coated wire to mark the exact location after confirmation.
Why is my detector alarming with nothing nearby? Background refrigerant from a previous leak or recent service. Either move to clean air to re-baseline, or use the auto-zero / purge function to recalibrate to current background. If it persists, the sensor is contaminated and needs purging or replacement.
Can a leak detector find an inside-the-coil leak? Yes — but you’ll need access. Either remove the coil cabinet panels and sweep all accessible coil tube sections, or do a pressure-decay test to identify the affected circuit. For sealed evaporator coils in walls, the only practical approach is to charge with UV dye and let it work over a week of operation.
My system shows pressure decay but the detector finds no leak. What now? Three possibilities: leak rate is below detector sensitivity (under 0.1 oz/year on a long line set), leak location is inaccessible (behind sheet metal you haven’t removed), or the “decay” is actually thermal change (system cooling down after operation, pressure drops as temperature drops). Verify with a calibrated pressure transducer over 24 hours of static cooling.
Heated diode vs. infrared — which is better? Heated diode is faster (≤1 second response) and cheaper; infrared is more selective (won’t false-alarm on alcohol vapor) and has longer sensor life. For field work, heated diode is the working tech’s choice. Infrared is reserved for high-end factory acceptance testing.
Related guides
- Best Refrigerant Gauge Set — gauge set diagnoses leak before detector finds it
- Best Vacuum Pump for Refrigeration — after leak repair, pull a deep vacuum
- Hoshizaki Error Code E1 — Inlet Water Valve Fix — leak detector confirms refrigerant integrity after valve replacement