- Home
- What the reading means
What the reading actually means
Something in your house or your building is showing you a number, and the number does not match the way the place feels. Five versions of that complaint, crossed with the four kinds of instrument that produce it, and an honest account of what each one can and cannot settle.
Read these first, because they apply to every answer below
- A reading is a fact about one point at one moment.
Not about a room, and not about a house. A thermostat reads the air touching a thermistor behind a plastic cover on one wall. Everything it says is true of that spot and of nothing else, which is why a reading and a feeling can both be correct at the same time.
- When two instruments disagree, the disagreement is usually the instrument.
Consumer temperature and humidity sensors are commonly a few degrees or several percent out, and they are out by different amounts. Comparing two of them to each other tells you almost nothing. Comparing one of them to itself over three weeks tells you a great deal.
- Nothing on this page is a health finding.
We measure temperature, humidity, airflow, pressure, combustion and ventilation rates, and we will give you every number we take. What any of it means for somebody’s breathing, sleep or allergies is a medical question. We are not qualified to answer it and we will not pretend otherwise.
Work out what your reading can settle
Pick what is happening and what is doing the measuring. Every answer below is on this page whether you pick anything or not — the picker only narrows it down.
The inexpensive particulate sensors inside phone-linked monitors are optical, uncorrected for humidity, and carry no calibration certificate. On a foggy Santa Barbara morning many of them read high for no reason other than the moisture in the air. They are genuinely useful for one thing: noticing that something changed. They are not evidence of what changed, how much, or what it means for anybody’s health — and we will not treat a screenshot from one as a diagnosis, or sell you equipment on the strength of it.
Do not adjust anything yet. Take two temperatures.
The air leaving a supply register and the air entering the return, measured with the same probe within a minute of each other. The difference between them tells you whether the system is doing work at all, and it takes ten minutes. Separately, hold a probe next to the thermostat: a wall unit that is two or three degrees out is common, cheap to establish and changes the whole conversation.
Why a correct reading can still feel wrong
Three things move a thermostat’s reading without anything being faulty. Afternoon sun on the wall it is mounted on. A lamp, a television or a return grille immediately below it. And an interior wall with a duct or a draught behind it, which makes the thermostat read the inside of the wall rather than the room.
Then there is the part that is not the instrument at all: at the same air temperature, a room with higher humidity feels warmer, a room with a cold plaster wall radiating away from you feels cooler, and moving air feels cooler still. A thermostat measures one of those four things.
What to do with that: establish the thermostat’s own error first, then decide whether you have a comfort problem or a temperature problem. They have different fixes and only one of them involves equipment.
Log the humidity in the damp room, over days, before anybody proposes equipment.
One spot reading will not settle a humidity complaint. Put a sensor in the room that feels damp, leave it for a week, and look at what happens overnight and in the early morning. If the number climbs when the system is off and falls when it runs, that is a very different problem from a number that never moves.
Why the thermostat says the house is fine
Most thermostats either do not measure humidity or measure it in a hallway, and a hallway is not where damp is a problem. The bathroom, a north-facing bedroom, a closed-up room on a fog-belt street and a garage conversion will all sit several percent above the rest of the house.
The common cause in this climate is not a broken system but a system that is too large for the house. An oversized unit satisfies the thermostat quickly, shuts off, and never runs long enough to condense moisture out of the air. The house gets cold and stays damp, and adding capacity makes it worse. The other common cause is extraction that does not extract: a bathroom fan moving a fraction of its rated air because of a crushed duct or a blocked roof terminal.
Both are measurable. Neither is settled by a number on a wall.
Measure the airflow at that room’s register. This is not a temperature problem.
A room that never matches the rest of the house is, in the great majority of cases, a room that is not receiving the air it was designed to receive. Measure what is coming out of its registers and compare it with what the room needs. Until that number exists, every other explanation is speculation.
What we usually find in that one room
- A long or crushed flexible duct run, or one that has come off its boot in the crawl space or the attic and is now cooling the crawl space.
- No way for air to leave the room. Air that goes in has to get out, and a closed door with a carpet under it is not an exit. Pressure builds and the airflow drops.
- Registers balanced by whoever lived here before, usually in the opposite direction to what was wanted.
- A room added later, on the end of a duct system sized for the house before it existed.
- A ductless head in the next room being asked to condition around a corner, which it has no means of doing.
The fixes are ductwork, a transfer grille, rebalancing, or a second head. They are unglamorous and they are usually cheaper than the thing that gets proposed instead.
Find out what the instrument measures and how it was calibrated, before you act on it.
Ask two questions of whatever produced the number: what is it actually sensing, and against what reference? A great many consumer air readings are a single optical particulate sensor with no humidity correction, and some are a software estimate rather than a measurement at all. Once you know which, the number becomes either useful or ignorable. Then measure the ventilation rate, which is the thing that can actually be changed.
What we can measure, and what we will not conclude
We can measure ventilation rates, airflow, filter pressure drop, humidity, and carbon monoxide and combustion products around gas appliances. Carbon monoxide is the one reading on this page that is a straightforward safety matter, and if it is present we will say so plainly and act on it.
We can also tell you what filtration will and will not change: a better filter captures more particulate and costs airflow, and on a system already short of airflow that trade is a real one. We will measure the static pressure before and after so the trade is visible rather than assumed.
What we will not do is attach an outcome to any of it. Not asthma, not allergies, not sleep, not headaches. Anybody in this trade who connects an air reading to a symptom in order to sell you a device is making a claim they cannot support, and it is the single most common dishonest sale in indoor air.
Read the schedule and the enrolled programs before you assume a fault.
Open the settings and look for three things: the schedule itself, any learning or auto-adjust feature, and any utility or manufacturer program the thermostat is enrolled in. A system that changes its own setpoint on hot afternoons is very often doing exactly what somebody agreed to when it was installed, and no equipment is faulty.
Sorting a setting from a fault
Things that look like a fault and are not: a schedule with a setback nobody remembers; an auto-changeover band that flips between heating and cooling in the same day; a learning feature adjusting the start time; an eco or away mode triggered by a phone’s location; and demand-response enrolment, where a utility program raises the setpoint at peak times. That last one is voluntary, it is usually disclosed at sign-up, and it can be left.
Things that are a fault: short cycling on and off every few minutes, which is a control, sizing or airflow problem and hard on the equipment; a system that runs with no call for it; and a thermostat losing its configuration, which normally means wiring or a common-wire power problem rather than the thermostat itself.
We will read the configuration with you and tell you which of those you have. It is frequently a five-minute answer and it is worth having before anybody is paid to replace anything.
What the thermostat on the wall can settle
It can tell you the setpoint, whether the system is calling, and — on most models made in the last fifteen years — how many hours it ran. Run time is the most under-used number in the house: a system running far longer or far shorter than it used to has changed, and that is worth investigating before anything fails.
It cannot tell you the temperature of the room, only of one spot on one wall. It cannot tell you whether the air is moving, whether the humidity is comfortable, or whether the room next door is anything like it. And it has its own error, which nobody ever measures and which is worth establishing once.
What a shop hygrometer can settle
It can tell you that something changed, in one place, over time. Left in the same spot for a fortnight, an inexpensive sensor is a genuinely useful instrument, because whatever its error is, it is the same error every day. Overnight rises, morning peaks and the difference between a run cycle and an idle hour all show up clearly.
It cannot tell you an absolute humidity figure you should act on. Several percent out is normal and unremarkable. It also cannot be compared to a different sensor in a different room: two cheap sensors reading 48% and 56% may be measuring identical air. Trust one device against itself, never two against each other.
What a smart thermostat with room sensors can settle
It can tell you a great deal, and this is the most useful consumer setup on the list. Several rooms measured at once, over months, with run time alongside. It will show you which room lags, at what hour, and in which season, which is exactly the evidence an airflow problem leaves behind.
It cannot tell you which room is in charge unless you check the averaging settings, and this catches people out. Depending on the configuration the system may be following one sensor, an average of several, or a different one at different times of day. Find out which before you conclude the equipment is misbehaving. And a sensor in a room is still a sensor at one point in that room.
What a phone app or consumer air monitor can settle
It can tell you that today is not like yesterday. As a change detector in one location it has real value, and a sustained shift is worth a phone call.
It cannot tell you a number you should spend money on. No calibration reference, optical particulate sensing that reacts to humidity, and in some products a modelled estimate rather than a measurement. Outdoor air-quality apps are a separate matter again: they report a regional monitoring station that may be miles away and tell you nothing about your own rooms. We will not diagnose from one, and we will not price work from one.
Where this page stops
We will measure anything on this page and give you every figure, with the date and the conditions it was taken in. We will tell you what we think it means mechanically, and where we are not sure we will say that instead of guessing.
We will not attach a health outcome to a reading, we will not sell a device on the strength of a symptom, and we will not quote a price for work before we have seen the building. If a reading turns out to indicate carbon monoxide, that is the one item here that is urgent and it gets dealt with immediately. Everything else on this page can wait for a measurement.