INTRO (0:00 to 0:30)
ON CAMERA, Darrel at the wall control, screen dark, panel off the air handler
This system has never been powered up. By the end of this video it will be commissioned, configured, and running, and then I am going to break it the way Phoenix breaks them, and show you the hunt. Two halves, same lesson: a communicating system is a conversation between three computers, and the tech who can verify the conversation owns every one of these calls. Watch the order I work in, because on these systems the order is not optional.
MAIN (0:30 to 11:00)
Beat 1: The wiring walk, before any power (0:30 to 2:00)
OVERHEAD: bus terminals at the air handler, fingers on each terminal as named
Start where every communicating job starts: the bus. Three terminals. R, 24 volt power from this air handler's transformer. B, common, and it is also the reference the data is measured against. D, data, the wire every message rides on. I pulled four conductor 18 gauge; three land, one is a spare, taped off and grounded at this chassis only, never floating, because a floating conductor running next to a data line is an antenna.
ON CAMERA, walking the cable route from air handler toward the outdoor unit
Now walk the run. I want a foot or more between this cable and anything inductive: blower motor, air cleaner, panels, ballasts. No staples through the jacket. This junction box mid-run is where any splice lives: mechanical, tugged, protected, accessible. If I could not get my one foot of separation, I would use shielded cable and ground the shield at one end only. Two grounded ends make a loop that carries current and adds noise instead of blocking it.
OVERHEAD: outdoor unit bus connector, the cable held up but not landed
At the condenser I am landing the wires color for color, same convention as the air handler: one color per terminal for the whole job. Except I am not landing them yet. I want you to see what the system does when a device is missing, so the outdoor unit stays disconnected for the first power-up. That is deliberate.
Beat 2: Power-up and discovery, with a hole in the roster (2:00 to 3:30)
TIGHT ON SCREEN: comfort control booting
Power-up order: indoor unit first, because it carries the transformer and the data clock, then outdoor, which today is disconnected on purpose. Now we wait. Ninety seconds to two minutes of boot, and you do not diagnose a blank screen before boot finishes.
TIGHT ON SCREEN: discovery running, indoor unit appearing
There is discovery: the control polls the bus and the air handler introduces itself, type and size, automatically. And here is what a missing device looks like: no outdoor unit in the roster. On a real install this is your moment. The roster on this screen must match the equipment you bolted down, and a hole in the roster is a wiring or power problem you fix before the trim goes on, not after.
OVERHEAD: landing D and B and R at the outdoor connector, then TIGHT ON SCREEN as the outdoor unit appears in the roster
Bus landed at the condenser, D to D, B to B, terminals tight. Re-run discovery from the menu, and watch: outdoor unit, variable speed heat pump, four tons. It told the system what it is. That auto-fill is the gift of a fully communicating system, and the discipline is that you verify the gift. Photo of the completed roster right now; that picture goes in ServiceTitan.
Beat 3: The wizard and the setup groups (3:30 to 5:15)
TIGHT ON SCREEN: Installation Wizard pages advancing as narrated
First boot launches the Installation Wizard: date and time, installer setup, service reminders, dealer code. You can re-run it any time through the service menu. The gate you must learn is Technician Access: hold it five seconds, and the installer-level menus open. That same gate hides test modes and charging mode, so burn the path into your hands: Menu, Service, Technician Access, hold.
TIGHT ON SCREEN: setup groups, equipment page first
Setup groups. Equipment: it auto-configured from discovery, heat pump, variable speed, and I verify it anyway, because thirty seconds of reading catches the mismatch that becomes a winter no-heat. Sensors: onboard temperature and humidity active, no remotes today. Accessories: none on this trainer. Comfort settings: here is dehumidification, two tools. Airflow reduction drops blower CFM about 30 percent when humidity runs over target, and overcooling lets it cool past setpoint, a tenth of a degree per percent of humidity error, capped at one, two, or three degrees, your choice in the menu.
TIGHT ON SCREEN: airflow group showing disabled, then OVERHEAD: the CDA at the outdoor board
Now the trap I promised. Airflow group at the thermostat: grayed out. Not broken, disabled, because this is a TruComfort variable speed system and the outdoor board runs the modulation. Airflow and blower delays for this system get set right here, at the CDA, the communicating display assembly plugged into the outdoor control board. If you are hunting a setting and the menu is dead at the wall, walk outside.
Beat 4: Charging mode and close-out (5:15 to 6:45)
OVERHEAD: CDA and gauges on the service valves
Charge verification on a variable speed system has one approved path. A modulating compressor chases the load all day; it never holds a steady state, so gauges in normal operation are noise. Charging Mode-Cooling pins it to a known state. Technician Access, Test Mode, Variable Speed, Charging Mode-Cooling. The window matters: outdoor between 55 and 120, indoor between 70 and 80, and the subcooling target gets corrected for line length and lift from the table in the service facts. Today the trainer holds its factory charge, so this is a verification, and the numbers land on target.
TIGHT ON SCREEN: system report screen, then ON CAMERA
Close-out: run a cycle in each mode and let the system narrate itself, demand, compressor speed, CFM, right on the screen. Then the proof set: photo of the roster, photo of the report screen, and one more reading I have not shown you yet, the most important meter habit in this module. That one gets its own beat, because now I am going to break this system.
Beat 5: The staged fault, and the alert that reports it (6:45 to 8:15)
ON CAMERA at the junction box mid-run, opening it
Phoenix special. A landscaper, a wire staple, a bougainvillea. I am simulating it here in the junction box: the D conductor comes apart inside a splice that still looks fine from the outside. Jacket intact, box closed, invisible.
TIGHT ON SCREEN: waiting, then the alert appearing
Now watch the timing, because it teaches. The comfort control sends its demand message every minute and declares the fault after three missed messages, so this takes a few minutes to appear. There it is: Err 91.02, loss of communication with the system, and the alert history logs it with a timestamp. First rule on arrival at any communicating call: read and photograph the alert and the history before touching anything. Timestamps are witness statements. A cluster of comm errors at 2 a.m. during a monsoon storm is a different diagnosis than three weeks of scattered ones.
ON CAMERA
Second rule: sort it. 89, 90, 91, those families are the conversation breaking. Everything else is equipment news that arrived over a healthy bus. This is a 91. Bus first, equipment second, because until the bus works, nothing the silent half of the system says can be trusted.
Beat 6: The hunt, one DC reading at three doors (8:15 to 10:00)
OVERHEAD: meter set to DC volts, probes on D and B at the air handler, display closeup
Here is the reading that does most of the work on these systems. DC volts, not AC, across D and B. At the air handler: about 12 volts. Twelve means alive and talking; this end of the bus is healthy. Sixteen would mean powered but silent, nobody transmitting, think dead clock or mid-boot. Zero means dead.
OVERHEAD: probes on D and B at the outdoor unit bus connector, display reading near zero
Same reading at the condenser: zero. Twelve at one end, zero at the other, the fault lives in the wire run between them. No board on this system is a suspect right now, and the last company that quoted this customer a control board never took these two readings.
ON CAMERA at the junction box, opening the splice, the broken conductor visible
Walk the run, and the only place hands have been is this box. There is the break, inside a splice that looked fine. One more move before I fix it, because you need it for the practical: if the whole bus had read zero everywhere, I would isolate by disconnection, lift one device's bus leg at a time and watch the meter. The literature gives you the reversed-wire signature: D and B swapped at the outdoor unit reads zero connected and about twelve on the field wires the moment you disconnect them, because the bus springs back without the clamping device. Zero even when disconnected says the D line itself is grounded in the run.
OVERHEAD: splice remade, then TIGHT ON SCREEN: system rediscovering, alert clearing
Proper splice, box closed, and watch the screen: the outdoor unit comes back, the alert clears, the conversation resumes. A comm fault that comes back in minutes was never fixed; this one stays gone.
Beat 7: Board discipline, the sixty second sermon (10:00 to 11:00)
OVERHEAD: pointing out the personality module on the outdoor control board
Last lesson, no fault staged, just the rules, because board swaps are where communicating callbacks are born. This little memory device is the personality module. It holds this unit's identity and configuration. A replacement board ships generic, so the PM moves from the old board to the new one, always; leave it in the recycle pile and the Err 114 family locks the compressor out until a good one is installed. After any swap: re-run discovery, clear any device showing offline in the summary table, and verify firmware versions match, because a fresh board can carry newer software than the rest of the system and mismatches act like ghosts: wiring tests clean and the system still misbehaves. Document all of it. Board is still last on these systems, and now the board has a memory you are responsible for.
OUTRO (11:00 to 11:30)
ON CAMERA
One system, both halves of the job: commissioned it, configured it, charged it in charging mode, then cut the conversation and found the cut with two DC readings. Your practical is exactly this: you commission this trainer end to end, I stage one bus fault, and you find it with the meter and your eyes. Twelve talking, sixteen silent, zero dead. Come ready.