An electronic leak detector warms up in clean air, then sweeps a probe slowly below fittings and ports until the alarm signals a hit.
A system can lose refrigerant for months without a drip or stain. The first clues are usually a vent that blows lukewarm, a compressor that never rests, and a creeping power bill. Technicians using an AC leak detector to find refrigerant leaks run the same order every time: confirm enough pressure to make vapor, zero the sensor away from refrigerant, then sweep the probe low and slow under every joint until the alarm reacts.
Before The Probe Moves: Charge, Warm-Up, And Zero
An electronic detector senses refrigerant vapor, so a system that has lost its charge cannot be tested until some vapor exists. On an empty system, add a small holding charge: nitrogen for a pressure test, or a small amount of refrigerant to generate vapor. Charging to full capacity first buries the leak under the refrigerant you just added.
The Bacharach H-10G manual calls for a two-minute warm-up, has the operator check airflow at the probe tip with a calibration reference bottle, and sets sensitivity with a Leak Size switch — SM(cal) for calibration, with a rebalance required on other ranges. Then zero the detector in open air, well away from anything holding refrigerant; a zero taken beside a leaking fitting hides the leak. Match the sensor to the refrigerant as well, since a detector built for one refrigerant family will not reliably catch another.
| Detector Setting | Aim For | Why It Matters |
|---|---|---|
| Warm-up | 2 minutes | The H-10G manual requires it before the sensor reads true |
| Zero point | Open air, no refrigerant nearby | A zero taken near a leak hides that leak |
| Probe height | About 1/4 inch above the surface | Puts the tip in the vapor layer instead of above it |
| Sweep speed | About 2 inches per second | Slow enough for a seep to register |
| Minimum charge | Roughly 15% of capacity, or 50 psi | No vapor means nothing to detect |
| Sensitivity | Leak Size switch, starting at SM(cal) | Range changes need a fresh balance |
| Air movement | As still as the space allows | Drafts dilute and scatter the vapor |
Scanning With An AC Leak Detector: Where To Point The Probe
Refrigerant vapor is heavier than air and sinks, so the probe belongs where the gas settles.
Those numbers come from the RLD-H10PM manual and separate a hit from a miss — a probe waved along at walking speed slides past a slow seep. Bacharach’s H-10G operating manual is the reference many techs work from, and its airflow checks are worth repeating before each job.
Work in one direction and finish every point on the route:
- Service access ports and their caps
- Flare and brazed joints
- Coil seams and welds
- Hoses and crimped ends
- Any fitting wearing an oily film
Evaporator leaks need one extra stop: pull the condensate drain tube and check inside it, because vapor travels down the drain. Shut off fans while you scan — airflow dilutes the vapor and carries it away.
Why Do Detectors Miss Slow Leaks?
Four habits cause most missed leaks: sweeping too fast, probing above the leak, skipping the open-air zero, and testing a system with too little refrigerant to make vapor. Any one produces a false all-clear.
Treat an alarm as a claim, not a verdict. Recheck the spot on a second pass, and confirm a pressurized joint with soap bubbles — a reading that disappears the second time was probably ambient refrigerant from your gloves or a can nearby.
Model choice matters: sensitivity, refrigerant compatibility, and sensor clearing speed vary by unit, and a tested roundup of AC leak detectors is the quicker way to compare before buying.
Run the same loop every time: confirm pressure, zero in open air, sweep low and slow, and verify anything the alarm flags. The leak sits at a joint, port, or crimp more often than anywhere else — and is nearly always slower than the probe hunting it.
FAQs
Does A Leak Detector Work With Every Refrigerant?
Not automatically. Sensors are built for specific refrigerant families, and a unit tuned for one type can miss another or read falsely. Check the manufacturer’s specification against the refrigerant on the system nameplate. If they do not match, use a detector rated for the refrigerant actually in the system.
How Do You Confirm A Leak The Detector Flags?
Recheck the exact spot on a slow second pass and watch whether the alarm repeats. On a pressurized joint, back it up with soap bubbles, since escaping refrigerant pushes visible foam. A reading that appears once and never again is usually background vapor, not a confirmed leak.
Can You Test An Empty System For Leaks?
Not by vapor alone. Add a small holding charge first — nitrogen for a pressure test, or a small amount of refrigerant to create vapor — and leave the system short of full capacity.
References & Sources
- Hanna Instruments / Bacharach. “H-10G Refrigerant Leak Detector Manual.” Source for the 2-minute warm-up, the calibration reference bottle, and the Leak Size switch settings.
- Parker Hannifin (Sporlan). “023-064X Refrigerant Leak Detector Technical Manual.” Technical background on refrigerant leak detection hardware.
- Copeland. “Leak Detectors.” Manufacturer documentation on leak detection equipment.
