29.3 Leak Detection & Refrigeration Test Instruments

29.3 Leak Detection & Refrigeration Test Instruments

Trade E still has to find the hole and prove the instruments. The outline’s remaining cluster in this chapter is chlorine/halogen leak detectors, the testing equipment used on a refrigeration call (velometer, Amprobe, VOM, vacuum gauges), and the standing rule that you already met in Section 25.4: never use oxygen as a pressure-test gas. Refrigeration & Air Conditioning Technology, 9th Edition (2021) is the detector-and-meter book. FBC Mechanical 2023 Section 1110 is the test-gas and vacuum sentence. EPA Section 608 (federal, not a substitute for the CILB card) still forbids venting. Class B still diagnoses through 25 tons / 500,000 Btu per system.

Quick Answer: Pressurize with oxygen-free nitrogen (never oxygen). Find leaks with soap bubbles and/or an electronic detector on R-410A. A halide torch is a chlorine tool (R-22 / R-12), not a 410A tool. Prove vacuum with a micron gauge (500 / 1,500 / 10 min). Amprobe = amps; VOM = volts/ohms; velometer = air fpm.

Never oxygen — then pick a detector that matches the refrigerant

FBC Mechanical 1110.3: field strength/leak tests use oxygen-free nitrogen, helium, or argon. Oxygen, air, and combustible gases shall not be used. The cutting-torch oxygen hose is the classic fatal shortcut: oxygen + compressor oil is an explosion. Shop air carries moisture and the same fire problem in an oily pipe. The small-bore exception (nameplate refrigerant as the test medium on tubing not over 5/8 inch OD) is not a license to skip nitrogen on a 7/8-inch suction, and it is not permission to spray extra 410A around a fitting “to listen for it” — that is venting under 608.

Standing-pressure decay (nitrogen, regulator, relief, 1110.4 gauge rules, 60-minute hold with no makeup gas) tells you whether the circuit leaks. Detectors tell you where.

Halide torch is the outline’s chlorine/halogen leak detector. It is an old, exam-favorite tool, and it is refrigerant-specific. R-22 has chlorine; the copper element forms copper chloride and the flame greens. R-410A is an HFC blend (R-32/R-125) with no chlorine — a halide torch is the wrong detector. Electronic heated-diode and IR tools are what you carry on a 410A Florida split. A helper who “always uses the green-flame torch” on 410A will pass a leaking liquid line and then add charge.

Electronic detectors still need a little refrigerant in the nitrogen if you are hunting on a repaired, empty circuit — follow the detector manual (some IR tools want a known trace; some ultrasonic tools do not). Calibrate, sniff from the bottom of a joint (refrigerant vapors are often heavier than air), and remember a 15 mph beach breeze will steal the sample. Soap bubbles remain the honest backup on a flare that the electronic wand cannot pin.

UV dye is a supplement. It does not replace 608 leak-repair thresholds on larger appliances, and it does not make a chronic leaker legal because “we can see it glow next season.” Recover, repair, nitrogen-test, evacuate, weigh in.

Worked leak call. A 3-ton R-410A change-out in Lee County loses 80 psig of nitrogen overnight on the 1110 hold. Helper A opens the oxygen hose “to get more pressure.” Stop — 1110.3 and the oil-film bomb. Helper B brings the pressure back with nitrogen, soaps the flares, and finds a wet Schrader. Helper C waves a halide torch and declares the circuit tight because the flame stayed orange — wrong detector for 410A. An electronic wand at the same Schrader screams. Replace the core, retest 60 minutes, then 500-micron vacuum.

Vacuum gauges vs the manifold compound gauge

Compound gauges on a manifold read in-Hg vacuum to some psig. They are fine for pump-down to a low-side psig or a rough “into vacuum.” They are not accurate in the micron range. 1,000 microns = 1 mm Hg ≈ 0.039 in. Hg. The last inch on a compound gauge is a barn door compared with 500 microns.

FBC Mechanical 1110.5 method 2: achieve 500 microns, isolate the pump, remain at or below 1,500 microns for at least 10 minutes. That number is read on a micron (vacuum) gauge piped to the system, valve-off from the pump so you are not reading the pump’s happy display through a wet hose.

Worked vacuum. Micron gauge on the suction service port: 400 microns, pump isolated. Ten minutes later 1,200 microns — pass (still ≤1,500). Same start, 2,800 microns in three minutes — leak or moisture. If the rise is fast and nitrogen standing-pressure already failed, it is a leak. If standing nitrogen held and microns climb slowly toward a plateau, it is often water in the oil/drier — keep pumping, change the drier, do not charge.

Deep vacuum is also how you boil remaining moisture. A 29-inch compound “vacuum” at a 40°F coil is not 500 microns. Buy the micron gauge.

Velometer, Amprobe, VOM — the rest of the Trade E kit

The outline lists these with refrigeration test equipment because a “no-cool” is not always a leak.

Velometer — Section 29.2. If cfm is 280 per ton, fix air before you hunt a 2-ounce leak that is not there.

Amprobe (clamp-on ammeter) — measures current without opening the circuit. Compare compressor RLA (rated load amps) on the nameplate to running amps. High amps + high head = dirty condenser, overcharge, or bad outdoor airflow (Florida grass clippings in the coil). High amps + locked rotor that never drops = seized compressor (Chapter 32). Low amps + high SH = unloaded or starved compressor, not a motor that “needs more gas.” Clamp one conductor; clamping the whole whip reads near zero because the currents cancel.

Worked Amprobe. 3-ton 410A compressor RLA 16.8 A. Running 21.5 A, liquid 480 psig on a 95°F day, condenser packed with palmetto trash. Clean the coil; amps fall to 15.9 A. That was not a leak and not a bad capacitor yet.

VOM (volt-ohm-milliammeter) — volts at the disconnect and contactor, ohms on windings and heater elements (power off, isolated), milliamps on some flame-safety and control loops. A 24-volt thermostat that is actually 19 V under load is a transformer or shorted damper, not a 410A leak. Continuity across a fused disconnect that is off is a dead meter, not a dead compressor. Chapter 30 drills electrical diagnosis; Trade E’s point here is that the VOM and Amprobe sit in the same bag as the gauges so you do not recover a full charge to “check the compressor” when L1 is dark.

Manifold gauges still must be rated for the refrigerant. R-410A high-side hoses and gauges are a different pressure class than R-22. Low-loss fittings and 608 still apply: you are not paid to vent the hose into a Naples driveway.

Florida scenario

Sunset Coast Mechanical, Class B in Lee County, has three no-cools on one 90°F afternoon. Job 1 is a 2-ton 410A that equalizes at 90 psig at 75°F indoor — far below the typical ~218 psig standing saturation (29.1). Nitrogen hold fails; soap bubbles show a cracked flare; electronic detector confirms; halide torch is left on the truck because 410A has no chlorine. Nitrogen retest, 500-micron hold, weigh-in. Job 2 is a 4-ton with a full standing charge, 4°F superheat, and 1,100 cfm on a velometer/free-area calc that wanted 1,600 — Amprobe shows low compressor amps because the coil is iced (low load), not a leak. Job 3 is a 30-ton packaged 410A on a condo roof: same leak method, same oxygen ban, same micron gauge — but 489.105 makes the contract Class A. The Class B candidate still owes the detector theory; the qualifier does not take the job.

Traps: (1) Oxygen or air as the test gas. (2) Halide torch on R-410A. (3) Certifying 500 microns on a compound gauge. (4) Clamping an Amprobe around a whole power whip. (5) Using dye or extra refrigerant as a substitute for repair and 608 recovery. (6) Hunting a leak on a coil that is actually a 280 cfm/ton air problem.