INTRO (0:00 to 0:30)
Darrel at the air handler, detector in hand: This system is low on charge, and the D24 work is done: superheat, subcooling, and pressures all agree, airflow is clean, the metering device checks out. So it has a hole. Today you watch the whole hunt: electronic sweep, bubble confirmation, and a nitrogen standing pressure test with the temperature math done in front of you. Nothing here is fast. Everything here is in order.
ON-SCREEN: D27 Field Demo: locate, confirm, prove
MAIN (0:30 to 10:30)
Beat 1: The setup and the search order (0:30 to 1:30)
Darrel holds up both detectors and gives each ten seconds: heated diode, hotter sensor, hyper-sensitive, eats itself over about a hundred hours and false-alarms on cleaners and glue; infrared, optical cell, lives ten times longer, recovers from a big hit in seconds. Either works if you respect it. He warms up his pick on camera, then zeroes it in clean air by the truck, away from the equipment, and says why: zero it next to the leak and you just taught it that refrigerant is normal. Then he states the map out loud before touching anything: 80 percent of leaks are in the A-coil, so we start there. Then service ports, then line set and brazes, then the condenser, last.
ON-SCREEN: Warm up. Zero in CLEAN air, away from the system
ON-SCREEN: Search order: A-coil first (80 percent), ports, line set, condenser last
Beat 2: Sweeping the A-coil (1:30 to 3:15)
The core teaching beat. Close-up camera on the probe tip the entire time. Darrel opens the coil access panel and narrates his hand: 1 to 2 inches per second, tip a quarter inch off the metal, and always the UNDERSIDE, because refrigerant falls. He traces the pattern for the camera: hairpin U-bends first, then the distributor and its capillary cluster, then the suction header, then the wet lower third of the slab. On vertical runs he starts high and works down, and says why: a leak above rains gas on everything below it; work upward and your first hit is the puddle, not the source. The detector ticks up near the condensate pan. Darrel freezes: hear that? Could be a leak. Could be the pool. He props the panel open, fans the cavity gently, waits a minute on camera, and re-sweeps. The pan area stays quiet, but the detector sings again at a distributor capillary, twice, from two directions. He marks it with a grease pencil. That is a locate. It is not yet a leak.
ON-SCREEN: 1 to 2 in/sec. 1/4 inch off the surface. UNDERSIDE of every joint
ON-SCREEN: Pooled gas alarms everywhere. Ventilate, wait, re-sweep
ON-SCREEN: Strongest REPEATABLE hit = your mark. Not the first beep
Beat 3: The wind demonstration (3:15 to 4:15)
Darrel moves to the condenser outside with the shop fan and stages the lesson honestly: I am going to lie to my own detector. Fan on low, three feet away, he sweeps the service valves and gets a hit on a joint he knows is tight, downwind. He kills the fan, waits, re-sweeps: silence at the tight joint. The plume was riding the breeze from the real source two feet upwind. The lesson, straight to camera: outdoors, the wind owns your plume. Block it with your body, a panel, the truck door, or hunt the outdoor unit in the still part of the morning. Never call an outdoor hit without confirming it in managed air.
ON-SCREEN: Wind carries the plume downwind of the real hole
ON-SCREEN: Block the wind or hunt in the calm. Confirm every outdoor hit
Beat 4: Bubble confirmation (4:15 to 5:30)
Back at the marked capillary. Darrel brushes commercial bubble solution generously over the joint and the two joints either side of it, then does the thing rookies cannot do: nothing. Timer in frame. He talks while it sits: a gross leak bubbles instantly, a weeper that loses ounces a month can take two or three minutes to raise its first dome, and wiping the joint at thirty seconds destroys the evidence. At about ninety seconds the macro camera catches it: a slow cluster swelling on the capillary braze. Cauliflower, he says. That is the finding. He photographs it with the grease pencil mark in frame: this photo is the diagnosis. The beep was the hint, the bubbles are the proof, and the picture goes in the job record before I touch a wrench.
ON-SCREEN: Bubble solution + patience. Slow weeps need 2 to 3 minutes
ON-SCREEN: Macro: the cluster forming. THIS is a confirmed leak
ON-SCREEN: Photograph the finding before the repair
Beat 5: One minute on dye, and the decision (5:30 to 6:30)
Darrel holds up a UV dye kit from the truck and does NOT use it: you will be tempted to start with this, because it is easy. It needs the system to run, it needs days to circulate, it stains the oil and the next tech's gauges, and some manufacturers do not allow it in warranty equipment. It is for ghosts: the intermittent leak that beat a good electronic search. Today's leak took eleven minutes to find without it. He sets it down. Then the decision, stated plainly: this is a braze joint leak at the distributor, accessible, on a coil with no formicary signature, single location, clean copper elsewhere. This one is a repair. If this coil had pinholes scattered across the slab on a five-year-old system, we would be having the whole-coil conversation, and on an old R-410A system that conversation includes the A2L transition, which is module A31's territory.
ON-SCREEN: UV dye = the ghost-leak specialist, not an opening move
ON-SCREEN: One accessible joint = repair. Scattered pinholes = coil condition, bigger decision
Beat 6: Setting up the proof (6:30 to 8:00)
Jump cut, narrated in one sentence: the charge is recovered per C15 and the joint is repaired per C16, nitrogen flowing while brazing, and we pick up at the moment the repair must prove itself. Darrel connects the nitrogen regulator and holds up the tank fitting for the camera: dry nitrogen, through a regulator, and never oxygen or shop air, because either one meeting refrigerant oil under pressure can detonate. He checks the nameplate test pressure rating out loud, pressurizes to 150 psig, and then records, on camera, writing as he speaks: 150.0 psig, 95 degrees at the equipment, 2:40 pm. Pressure AND temperature, he says, because nitrogen pressure tracks absolute temperature and a reading without a temperature is half a number. He sprays the new braze and his own gauge connections with bubble solution: before I trust the clock, I prove my own rig is not the leak.
ON-SCREEN: Dry nitrogen through a regulator. NEVER oxygen or compressed air
ON-SCREEN: 150 to 200 psig, never above the nameplate rating
ON-SCREEN: Record FOUR numbers: start psig, start temp, end psig, end temp
Beat 7: The wait and the math (8:00 to 9:45)
The centerpiece. Time-lapse with the gauge and a clock in frame; Darrel uses the wait to teach the arithmetic on a whiteboard or tablet. The trap question first: you pressurize to 150 at 95 degrees in the evening, come back to 144 at 75 degrees in the morning. Leak? He works it in the open: gauge plus 14.7 is absolute pressure, Fahrenheit plus 460 is absolute temperature. 164.7 times 535 over 555 is 158.8 absolute, minus 14.7 is 144.1 psig. The expected reading IS 144. That system is tight; the cool night took the six psi. Then the reverse trap: same system, same swing, but the morning gauge still reads 150 after warming 20 degrees. That system LEAKS, and the gauge never moved. Rule of thumb at this pressure: about 1 psi per 3 degrees. Back to real time: 30 minutes elapsed, equipment temperature unchanged at 95, gauge dead on 150.0. He writes the end numbers next to the start numbers and calls it: no temperature change, no pressure change, corrected verdict TIGHT. This repair has earned a charge.
ON-SCREEN: Expected P2 = (P1 + 14.7) x (T2 + 460) / (T1 + 460), minus 14.7
ON-SCREEN: 150 at 95 F reading 144 at 75 F = TIGHT. Same swing reading 150 = LEAK
ON-SCREEN: Verdict: 150.0 to 150.0, temp steady. TIGHT, corrected and recorded
Beat 8: Document and hand off (9:45 to 10:30)
Darrel photographs the gauge with the thermometer and his written four-number record in frame: the bubble photo from Beat 4 and this photo are the two ends of the story, the hole and the proof it is gone. Both go in the ServiceTitan close-out with the method, location, and repair written in one line. From here the job is C15 and C17: evacuate to 500 microns with a decay test, then weigh the charge in. He taps the gauge once: a standing test you did not write down did not happen.
ON-SCREEN: Two photos tell the story: the bubble cluster and the corrected verdict
ON-SCREEN: Then C15: 500 microns + decay test. Then C17: charge by weight
OUTRO (10:30 to 11:00)
Darrel, packing the detector case: The whole hunt was three jobs in order. Locate with the detector, slow and low, A-coil first. Confirm with bubbles and a photograph. Prove with nitrogen, four numbers, and the temperature math. Your practical is this exact sequence with a leak I planted and a clock I will pressure you on. The system knows where it is leaking. Move slow enough to hear it.
ON-SCREEN: D27 practical: find the planted leak, confirm it, prove it, document it