INTRO
0:00-0:30
SHOT: Darrel standing at the outdoor unit, panel still on, gauges NOT connected. He holds up his manifold, then deliberately sets it on the ground.
DARREL: No-cool call, variable speed system, hundred and nine outside. Twenty years of habit says grab the gauges. On this equipment, the gauges go on the ground first. This box has been keeping a diary of everything that went wrong, and we are going to read the diary before we touch anything. And before either of us touches the inside of that drive, you are going to watch me prove the capacitor bank is dead. That part is not optional. Let's work it.
MAIN
[0:30-2:00] Codes first, before power cycling
SHOT: Over Darrel's shoulder at the outdoor board display / service tool. Close-up insert on the display.
DARREL: First move, and notice the disconnect has not been touched: fault codes. Some of these platforms wipe the history when you cycle power, so the worst thing I can do right now is "reset it and see."
ON-SCREEN: text overlay "RULE 1: Codes before power cycling"
DARREL (reading): There is our history. Three stored events. I am photographing this display before anything else, that photo goes in the job record. Now read it like a story, oldest first. The oldest code points at a sensor circuit. The newer ones are the system protecting itself afterward. Rookies read the newest code because it is on top. The oldest one is usually the disease, the rest are symptoms.
SHOT: Darrel holds the tablet next to the display, service manual code page visible.
DARREL: And I decode against the manual for this exact model. Every brand numbers these differently, that is its own module. The habit is universal: codes first, photo, oldest to newest, decode.
[2:00-3:30] Watch it run: the ramp is not a fault
SHOT: Wide on the condenser, Darrel initiates a call from the indoor control. Insert: clamp meter on a compressor leg, then the manifold now connected.
DARREL: Now I let it try to run, and I watch how it fails, because where in the ramp it dies is evidence. Hear that? No clunk, no hard start. The drive ramps the compressor up from almost nothing. There is no locked rotor on this equipment, ever. The old inrush diagnostics from the diagnostics track do not exist here.
SHOT: Close on the manifold. The suction needle is visibly drifting downward.
DARREL: Look at the suction. Moving. Still moving. If you snap-judge that number you will invent a refrigerant problem this system does not have. While it ramps, those gauges are answering a question that keeps changing. The only honest way to read charge on this machine is test mode, a forced fixed speed, ten or fifteen minutes of stabilization, then compare to the manual's numbers for that mode. We are not there yet, because watch what it does.
SHOT: The unit ramps down and stops. Darrel points at the board display.
DARREL: Aborted its own ramp, logged another event, same sensor family. It is telling us where to look. But the next checks are inside the drive compartment, so now we do the part of this video you do not skip.
[3:30-5:30] Power down and prove the bus is dead
SHOT: Darrel at the disconnect. Deliberate pacing, each action shown clean and separate.
DARREL: Disconnect pulled. Locked, tagged. Power verified off at the line side, both legs, that habit is as old as module F1. And here is where this equipment is different from everything else you have worked on. Behind that panel is a DC bus, a bank of capacitors sitting at over three hundred volts DC when this thing runs. Pulling that disconnect stopped the refill. It did not touch the charge already stored. That stored charge can kill you, with the power off, ten minutes after the power is off, if the bleed circuit failed. And bleed resistors fail open with no warning.
ON-SCREEN: text overlay "DC bus capacitors hold LETHAL charge after power-off"
SHOT: Darrel checks the unit label, taps his watch.
DARREL: Label says wait. So we wait the labeled time. The wait is the minimum, it is not the proof. The proof is my meter.
SHOT: Tripod close-up, the money shot of the video: meter leads on the marked DC bus test points, display readable.
DARREL: Panel off, hands off everything else. Meter on DC volts, across the marked bus test points, the manual shows you where they are on this board. Reading: 31 volts... 28... 24. Below 50 and falling. That is a discharging bus. If that number were sitting high and frozen, the bleed circuit is dead, and I stand here re-measuring until it drains, however long that takes. What I never do is short it with a screwdriver. That is an arc flash and a dead board, best case.
SHOT: Darrel photographs the meter reading with the tablet.
DARREL: And that reading gets photographed into the job record. At this company the discharge verification is documented on every inverter job. Now, and only now, my hands go in this cabinet.
[5:30-7:00] Sensor checks: cheap parts, expensive misdiagnoses
SHOT: Darrel at the open control compartment, following the sensor harness.
DARREL: Codes pointed at the outdoor ambient sensor circuit, so before anybody says the word "board," we check the cheap part. These systems run entirely on what their sensors tell them. A bad sensor makes a healthy machine act insane: aborted ramps, limited speed, endless protection logic. Looks exactly like a drive problem. Costs forty times less.
SHOT: Close-up, Darrel unplugs the sensor connector, meter on ohms, manual's thermistor table visible on the tablet.
DARREL: Sensor unplugged, meter on ohms, compare against the resistance table in the manual for today's actual temperature, which my IR gun says is 109 at this coil. And... open circuit. There is the whole call. Follow the harness... and look at this connector. Backed out at the plug. Reseat, reading again: now it matches the table for this temperature.
DARREL: Notice what almost happened. Aborted ramps, weird gauge readings, drive-flavored fault codes. Half this trade would have quoted an inverter board today. The board was never the suspect, the diary said sensor circuit from the first photograph we took.
[7:00-8:45] Drive versus compressor: the discrimination
SHOT: Darrel still at the open compartment, compressor leads visible.
DARREL: Before I button up, the two-minute discrimination you run any time a drive or compressor is genuinely in question, and always before any board ever gets replaced. Bus is verified dead, so: compressor leads off the drive.
SHOT: Close-up on the meter, three measurements shown in sequence.
DARREL: This compressor is a three phase motor, even though this house is single phase, because the drive builds its own three phase power. So no common-start-run, no sum check. Three windings, and the readings should be equal. Phase to phase: 0.9... 0.9... 0.9 ohms. Leads zeroed first, because at under one ohm your lead resistance can condemn a good compressor. Equal, low, healthy. Each lead to ground: nothing. The megohm rules from D26 still apply if there is any doubt, pressure in the shell, never under vacuum.
DARREL: Here is why this check is sacred: a shorted compressor winding takes the drive's output stage with it. Hang a new board on an untested compressor and the compressor executes the new board at first start, and now you have an angry customer and a denied warranty claim. Windings before boards. Every time.
SHOT: Darrel reconnects the leads, gestures at the board.
DARREL: And the verdict logic, say it with me: sensor fault shows as a sensor disagreeing with your own instruments. Compressor fault shows in the windings with the drive disconnected. Drive fault is what is left when supply, sensors, and windings are all proven and the drive still will not make balanced output on those three legs. The board is condemned by elimination plus evidence. Never by default. And if the evidence dead-ends, the manufacturer tech line, with the model, the serial, that fault photo, and my measured values written down before I dial.
[8:45-9:45] Restore, stabilize, verify
SHOT: Panel on, power restored, system running. Darrel at the board display, then at the manifold.
DARREL: Power back, call initiated. Soft start, clean ramp this time, no abort, and the code history shows our event cleared, no new entries. Now I will give you the charge verification, the right way: test mode, locked speed, and we wait. Ten to fifteen minutes minimum at that fixed speed before these gauges mean anything.
SHOT: Time-lapse insert, then a settled manifold next to the manual's test mode targets on the tablet.
DARREL: Settled, at a known speed, compared against the manual's targets for this mode, not the generic numbers in your head. In range. That is the only refrigerant judgment I have made all day, and it is the only one this machine allows.
OUTRO
9:45-10:15
SHOT: Darrel at the closed unit, tablet in hand.
DARREL: Walk it back. Codes before power, photo first, oldest code is the story. Watch the ramp fail, do not judge gauges while it moves. Power down and prove the bus is dead, under 50 and falling, photographed. Cheap sensors before expensive boards. Windings before any board, equal times three. And a stable forced speed before any charge call. One monsoon surge season from now, this flow will be the difference between you and the guy who replaces three good boards a summer.
ON-SCREEN: text overlay "Codes. Discharge verified. Sensors. Windings. Then verdicts."