Every point on the graphic is green. Every value is plausible. The fan is proven, the valve is at 100%, the supply air temperature is right where it should be. And the room is 80 degrees. When the BAS looks entirely normal and the building doesn’t, the rule is simple: trust the building. Something is measuring the wrong thing, or nothing is measuring the thing that matters.
This is the call that separates the two kinds of technician. One keeps scrolling through graphics looking for the point that’s wrong. The other gets up and goes to the equipment. The BAS gives you an account of the building. The building gives you the building.
Four ways a normal-looking system hides a fault
It’s measuring the wrong thing
The number is accurate and the location is wrong. A zone sensor on an exterior wall, in afternoon sun, above a copier, or behind a door that’s always open. A discharge sensor mounted so close to the coil that it reads stratified air. An averaging point with one failed sensor holding a fixed low value, dragging the average down while every other sensor is fine. The sensor reports faithfully. It just isn’t reporting on the space the complaint is about.
Check: where the sensor physically is, and whether that spot represents what the sequence thinks it does
It’s configured wrong in a way that looks plausible
A thermistor configured as the wrong curve reads almost right near room temperature and gets worse toward the ends of the range. Measure the resistance at the sensor and it’s perfect for the actual temperature; the controller is simply translating it with the wrong table. It’s been wrong since the day it was commissioned, and it never alarmed because the numbers were never absurd.
Check: the reading against a reference thermometer, and the configured sensor type against the actual device
It’s describing the actuator, not the device
A slipped coupling lets the actuator stroke fully and report fully while the damper sits two-thirds open. A valve that doesn’t seat leaks hot water through a coil all night while its feedback reads a confident 0%. An actuator with stiction sits still until the signal builds, then jumps past where it was asked to go, and the loop gets blamed for it. Feedback proves the actuator moved. Only your eyes prove the damper or valve did.
Check: the blade, the shaft mark, the valve stem, and the temperatures on either side of the coil
Nothing is measuring the thing that failed
A fan proven by a current switch with a broken belt: the motor draws current, the switch closes, no air moves (unless the switch is set to catch the drop in current). A coil fouled on the air side: valve open, water flowing, air barely changing temperature. A drive commanded to 60 Hz and running at 48 because it’s current-limiting to protect itself. These are mechanical failures in places the control system has no sensor. The BAS isn’t lying. It’s blind there.
Check: rotation, belt, coupling, ΔT across the coil, the drive’s own display
There’s a fifth that isn’t a hardware problem at all: values that are old or overridden. A point held by a stale override, a frozen value from a dropped subscription, a schedule running a holiday nobody remembered. They look normal because they’re frozen at a moment when things were normal. Why “offline” isn’t always offline and your graphic should tell you when it doesn’t know both cover how that happens.
The go-look rule
Step four of the method is the step most often skipped and most often decisive: determine what the equipment is actually doing. Go to it. Look at it, listen to it, put a hand on it, put a meter on it.
- Is it running? Is the shaft turning? Is the belt intact and tight? Is the rotation right?
- Is the damper blade or valve stem where the graphic says? Not the actuator. The blade and the stem.
- Is the pipe hot where it should be hot, and cold where it should be cold?
- Does it sound normal? Surging, cavitation, a belt squeal, a bearing, silence where there should be airflow.
- Is the air coming out of the diffuser the temperature and volume the box says it’s delivering?
If what the BAS says and what the equipment does disagree, you’ve already found the fault’s neighbourhood: it lives in the layers between the point that reads right and the equipment that isn’t. That’s a very small place to search.
Go look even when the call seems to have already been diagnosed. “The zone sensor is bad” often arrives with a new sensor already installed and the room still warm. The caller took an observation and turned it into a conclusion. Keep the observation, leave the conclusion, and go see what the box, the damper and the diffuser are actually doing. Check the neighbours on the way: one unhappy zone next to satisfied ones points at that box or that room, while several unhappy zones on the same air handler mean you should stop working on the box and go upstream.
Prove the points against reality
You don’t have to prove every point. Prove the few that drive the complaint.
| Point | What proves it | Quick field check |
|---|---|---|
| Supply or discharge air temperature | A reference probe next to the sensor | Probe in the duct at the sensor location, then a few feet away to check for stratification |
| Damper position | The blade | Look at the blade or shaft indicator at 0, 50 and 100%. Check the coupling set screw |
| Valve position | Stem position plus water temperatures | Stem travel, and ΔT across the coil. A “closed” valve with a warm return line is passing |
| Fan status | Moving air | Shaft turning, belt intact, amp clamp, airflow at the discharge |
| Fan speed | The drive’s actual output | Output frequency and current on the drive display, compared to command. Check for active current limit |
| Space temperature | The space | A reference thermometer at the occupant’s desk, not at the thermostat |
| Airflow | An independent measurement | Compare against the balance report or a hood reading. Check the pickup tubing |
Let the physics check the points
Individual numbers can all be plausible while the combination is impossible. That’s your best tool for finding which point is lying without measuring every one.
- Mixed air has to land between outside air and return air. If it doesn’t, one of the three sensors is wrong, or the damper isn’t where it says. The mixed air fraction calculator does the arithmetic.
- A coil with its valve open and water flowing has to change the air temperature. Chilled water valve at 100% and a 3°F drop across the coil means water is moving and not picking up heat. Look at the coil, not the valve.
- A zone at maximum airflow that can’t cool is either not getting that airflow, not getting cold air, or carrying a load the box was never sized for.
- Heating and cooling should not both be doing work. If they are, somebody’s signal is wrong. Simultaneous heating and cooling covers the usual suspects.
Don’t tune your way out of a mechanical fault
When a loop won’t hold setpoint and the graphics all look reasonable, the temptation is to start adjusting gains. Resist it until you’ve watched the actuator. A damper with stiction, a linkage with slop or a valve that doesn’t seat produces exactly the jumpy, overshooting trend a badly tuned loop does, and no tuning will fix it. Trend command and feedback together, then go watch the device move. Hunting valves and dampers covers where tuning ends and mechanics begin.
What to do with what you find
- Fix the cause, not the symptom. Raising a setpoint so the room feels right with a slipped coupling is hiding evidence, not repairing anything.
- Record the discrepancy: what the BAS said, what you measured, and where.
- Recommend the missing point if the fault was invisible to the system. A discharge temperature sensor, an adjustable current switch set to catch a broken belt, a position switch on a critical damper. That’s a recommendation to the owner, not a change you make on your own.
- Verify in automatic, under the conditions where the complaint happened.
If the equipment isn’t running at all despite a command, start with tracing a command that produces no output. If something is holding it off, see interlocks and enables. And if it only goes wrong some of the time, intermittent problems is next.
The graphic is the last thing to trust, not the first. It’s a useful account of what the system believes. The building is the only thing that knows what’s actually happening, and the only way to ask it is to go look.
Stuck on a live call? The BAS Troubleshooter walks you through this method step by step, offline, on your phone.
Go deeper
The method behind this, chapter by chapter
Every sensor, drive and control loop in the building — how it works, how to prove it with a meter, and how it fails. BAS Field Fundamentals is the method this article uses, with 20 field cases and a browser diagnostic simulator to practice on.
Get BAS Field FundamentalsFound a wrong number or a broken link? Tell me and I’ll fix it.