As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
Yes. An optical bench ages in ways that all push the reading in the same direction, so without correction it reads progressively high. Most consumer instruments handle this with an automatic baseline routine that assumes the lowest concentration seen over the past few days was outdoor air, which works in a room that empties and fails in a room that does not. Where the assumption fails, or where there is no such routine, the correction is a manual fresh air calibration: settle the instrument outdoors in shade and moving air, away from traffic and flues, then trigger the routine and leave it alone.
Why an infrared bench drifts even with no moving parts
Nothing inside the instrument wears mechanically, which is why the drift surprises people. The ageing is optical. The infrared source is a heated element whose radiant output falls as its hours accumulate. The windows at each end of the chamber collect a film of dust, cooking grease and condensate, and a film absorbs. The band pass filter in front of the detector degrades slowly. The detector itself changes responsivity with temperature and, over long periods, with age.
Every one of those effects reduces the energy arriving at the detector. The instrument interprets less arriving energy as more absorption, and more absorption as more gas. That is why the drift is one sided: an uncorrected bench creeps upward over months and essentially never downward. A user who has watched a monitor climb from a plausible indoor figure to an implausible one, with no change in how the room is used, is usually watching this and not a change in the air.
A reference channel removes most of the source ageing by watching a wavelength the gas does not absorb and dividing it out. It cannot remove what affects only the measurement band, including degradation of the filter itself. Instruments that state a dual channel arrangement are claiming the first half of that protection, as the Autopilot desktop monitor and data logger does with a two channel low drift infrared sensor, alongside a stated operating envelope of thirty two to one hundred and twenty two degrees Fahrenheit and zero to ninety five percent relative humidity without condensation.
Automatic baseline correction and the rooms it gets wrong
The routine is simple and the simplicity is the problem. Firmware keeps a rolling record of the lowest concentration observed over a window of several days, then asserts that this minimum was outdoor air and adjusts the calibration so it reads the outdoor figure stored in the firmware. In a workplace that empties every evening, the assumption is sound and the instrument quietly maintains itself for years without anyone touching it.
Three kinds of room break it. The first is any space occupied continuously, such as a bedroom used every night with the door closed, where the overnight minimum is a real accumulation rather than fresh air and the routine subtracts a genuine signal. The second is an enriched space such as a greenhouse or a grow room held deliberately above outdoor concentration, where the minimum may be far above outdoor air and the routine eventually decides that the enriched level is zero. The third is a space with a persistent indoor source, such as an unvented appliance, where the same thing happens for a reason nobody wants hidden. Neither Autopilot listing says whether such a routine is running at all, which is Not stated in the listing bullets and has to come from the manual before you decide whether to intervene.
The outdoor reference itself is not a constant either, which is the overlooked half of the problem. Concentration outdoors rises in a street canyon, downwind of combustion, inside a courtyard and under a plant canopy overnight. A firmware figure that assumes a clean rural value will set a low zero for an instrument living beside a busy road, and a manual calibration performed in the wrong spot does the same thing deliberately. For the specific case of a room that never empties, see using a monitor in a bedroom.
Running a fresh air calibration properly
The routine is only as good as the air you give it, so the preparation matters more than the button press. Choose a spot outdoors in open moving air, well away from a road, a flue, a vent, a compost heap and dense planting. Keep the instrument in shade, because direct sun heats the enclosure and a warm optical block reads differently from a cool one. Stand upwind of it or walk away, since your own breath is a large local source at the scale being measured.
Then give it time. The instrument has to reach outdoor temperature, the chamber has to flush the indoor air it carried out, and the display has to stop moving. Half an hour is a sensible floor and longer is better in cold weather. Only once the reading is genuinely stable should the routine be triggered, and where a device documents a fixed settling period the whole of that period has to pass undisturbed. Bring it back indoors afterwards and compare against whatever second instrument you have before trusting the result, which is the practical check our walkthrough of calibrating a carbon dioxide monitor goes through step by step.
Fixed installations introduce a physical problem the procedure assumes away, because a wall mounted display cannot be carried outside. A remote sensing head solves it. The Autopilot monitor with a remote sensor probe states a sensing probe on a cable of fifteen feet with a photocell, which means the element can be positioned deliberately rather than wherever the display happens to be convenient, and that separation is exactly what makes a controlled reference possible without moving the whole instrument.
Two point calibration and when it is worth the cylinder
| Route | What it corrects | What it leaves uncorrected | What it needs |
|---|---|---|---|
| Automatic baseline routine | Slow upward drift of the zero | Gain errors higher up the scale, and everything if the room never empties | A room that reaches outdoor air regularly |
| Manual fresh air calibration | The zero, against a known outdoor assumption | Gain errors, and any error in the assumed outdoor figure | A settled instrument in clean outdoor air and patience |
| Single point span with certified gas | The gain at one working concentration | The zero, if it has moved independently | A cylinder, a regulator and a calibration adapter |
| Two point calibration | Both the zero and the gain, bounding the whole range | Non linearity between the two points | Nitrogen or scrubbed air plus a certified span mixture |
For a display somebody glances at, the first two rows are enough and the cylinder is hard to justify. The calculation changes when the reading operates equipment. A controller that switches injection and ventilation acts on its own number without a human reviewing it, so an error in the gain becomes an error in how much gas is released, and the only way to bound that is a reference at a concentration near where the equipment will actually work.
Note also that the zero on an infrared device cannot be taken in room air the way an electrochemical zero can, because ambient concentration is a substantial fraction of the working range. The low point has to come from nitrogen, from scrubbed air or from the outdoor assumption, which is why the fresh air routine exists in the first place and why it is the compromise it is.
How often, and what should reset the clock
An instrument running an automatic routine in a suitable room needs no scheduled intervention, since it is correcting itself every few days. An instrument without one, or one in a room where the routine is disabled, benefits from a manual fresh air calibration a couple of times a year. Beyond the calendar, several events should move the date forward regardless of when the last one happened: transport to a different altitude or climate, a long period in storage, a battery removal that cleared the configuration, a large temperature change in the room, and any physical knock.
The strongest trigger is disagreement. When two instruments that used to track each other diverge by more than their combined published tolerance, something has moved, and the check costs nothing but time. Our look at a multi channel room monitor and our review of a compact carbon dioxide and humidity unit both deal with devices where this comparison is practical because the history is retrievable.
Finally, keep other channels out of the operation. A particle instrument such as the Eujgoov particle counter, which states six size channels and a built in micro vacuum pump drawing the sample, drifts through chamber fouling and flow rate change rather than through optical ageing in a gas band. Clean air tells it nothing useful about its own gain, and a fresh air routine designed for a gas bench has no bearing on it at all.
Frequently Asked Questions
How do I know whether my monitor has an automatic baseline routine?
The manual is the reliable answer, and the behaviour is the next best one. A device with the routine running tends to converge on a familiar low value after a few days in a room that empties regularly, even if it started somewhere implausible. A device without it holds whatever offset it has indefinitely and only moves when you trigger something. A listing that documents a manual field calibration procedure is usually describing a device where the user is expected to supply the reference.
Should I switch the automatic routine off?
Only where the assumption behind it is false. In an office, a classroom or any room that stands empty for part of every few days, leaving it on is the easiest way to keep a bench honest without a cylinder. In a continuously occupied bedroom, a greenhouse, a grow space or anywhere the concentration is deliberately held above outdoor air, the routine will converge on the wrong reference and should be disabled in favour of a manual calibration performed somewhere sensible.
Can I calibrate by taking the monitor outside for five minutes?
Five minutes is rarely enough and the location matters as much as the time. The instrument has to reach the outdoor temperature, the optical block has to stabilise with it, and the sensor has to flush the indoor air it arrived with. Half an hour in shade, in moving air, away from traffic, flues and dense vegetation is a reasonable minimum, and a device with a documented settling period should be left undisturbed for all of it.
What does a calibration not fix?
It does not fix a fouled optical path, a failing source or a sensing principle that was never measuring the gas. Calibration shifts the relationship between the raw signal and the displayed number, so it can compensate for a weaker signal up to a point and then stops. It also does nothing for placement, which is frequently the real cause of a reading somebody does not believe.
Two instruments differ by a hundred and fifty parts per million. Which one needs attention?
Neither, if both publish a tolerance and the gap falls inside the sum of the two. That arithmetic is the first thing to do, and it resolves a surprising number of disputes. If the gap does exceed the combined figure, take both to the same outdoor location and let them settle, then the one further from the outdoor baseline is the one whose zero has moved.
Do the humidity and particulate channels get calibrated at the same time?
No, and treating them as one operation is a common mistake. A fresh air routine adjusts the carbon dioxide zero and has no effect on anything else in the box. A humidity element is usually checked against a saturated salt reference instead, and an optical particle channel drifts through chamber fouling and flow rate change, which a reference in clean air does nothing about.