A carbon monoxide meter is not an alarm. An alarm tells a household to get out; a meter tells a technician what the combustion appliance is doing, and whether the number is moving in the right direction after an adjustment. This page separates the instruments that can be used for low level investigation and calibration work from the ones that cannot.
Measure It
Compare It
Buy Once
As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
The source list contained twelve products. Six were excluded because they are residential life safety alarms or low level monitors rather than servicing meters: P9, P10 and P11 are UL 2034 residential alarms, and P12 is a low level monitor with a stated 10 ppm display start and a 25 ppm buzzer. They belong on an alarm page, not a servicing meter page.
Pump, probe, colour graphing display and USB data logging in one instrument.
Pumped analyzer with stated sensor tolerance, response time and time stamped alarms.
Certified handheld with a wide range, fast response and NIOSH based thresholds.
Factory calibrated, IP67 and a two year battery for straightforward point checks.
Rechargeable meter with three alarm modes and a zero and span calibration menu.
Two fixed alarm levels and a short term exposure limit alarm, nothing more.
No pick matches that combination yet. Try a wider budget or a different measurement.
| Pick | Range | Resolution | Alarm thresholds | Pump | Logging | Calibration stated | Protection | |
|---|---|---|---|---|---|---|---|---|
| 0 to 1000 ppm | 1 ppm | Adjustable, buzzer, LED, vibration | Yes, 4 ft probe | USB computer data logging | USA NIST traceable, Los Angeles | Not stated | View | |
| 0 to 1000 ppm | 1 ppm | Adjustable, audio, visual, vibration, time stamped | Yes, 0.5 LPM | Alarm digital time stamp and logging | USA NIST | Water, dust and explosion proof stated | View | |
| 1 to 1999 ppm | 1 ppm | 35 ppm and 200 ppm, LED, 80 to 85 dB, vibration | No | MAX peak only | CE certified EN 50271 | IP67 | View | |
| 0 to 1000 ppm | 1 ppm | Not stated | No | No | Factory calibrated, USA NIST certificate | IP67, 20 ft drop test | View | |
| 0 to 1000 ppm | Not stated | Two thresholds, light, sound, vibration | No | Alarm history | Zero and standard gas settings | Not stated | View | |
| Not stated | Not stated | 35 ppm and 200 ppm, STEL alarm | No | No | Not stated | Not stated | View |
Cells marked Not stated are absent from the listing bullets. The P1 response claim of 0.3 seconds is not shown here because it is a response claim rather than a range or resolution figure, and it is discussed in the review.
The right meter depends on whether you need to find CO, measure it, or prove what it did. Those are three different jobs.
A meter is for you, during work. An alarm is for the occupant, all the time. If you need to leave protection behind, buy an alarm as well; a meter left in a house is not protection.
A 1 ppm resolution lets you watch a reading fall as you adjust combustion air. A coarse display will sit still while the appliance is still spilling. All the meters here that publish resolution state 1 ppm.
A pump with a probe lets you sample a flue or a plenum you cannot reach. Without one you can only measure where the meter body can go. Pumped instruments need their sample line checked before every job.
A stated calibration trace and a zero and span menu are what let you defend a reading later. Factory calibration gives a starting point; a span gas check gives you confidence it is still there.
A CO meter detects carbon monoxide only. It does not detect natural gas, propane or refrigerant. For a leak search you need a different instrument, and for confined space entry you need a multi gas detector.
Rechargeable meters need charging before a call. Lithium primary cells last years and are the safer choice for an instrument that lives in a bag and is used occasionally.
No. A meter is an inspection tool. It measures while you are there and it does not sound an alarm to wake an occupant at three in the morning. The residential alarms in the source list were excluded from this comparison for that reason. If the household needs protection, it needs an alarm, and a meter is an additional tool for the technician.
It lets you see a reading change. In combustion work the useful question is often whether CO falls after an adjustment, and a 1 ppm display shows that movement. A coarse display can sit on the same number while the appliance is still producing CO. Resolution is not accuracy, though: a 1 ppm display does not mean the reading is within 1 ppm of the true value.
Only if you need to sample somewhere the meter body cannot reach, such as a flue test port or inside a plenum. A pump with a probe makes that possible. If all your work is standing in a room and reading the air, a pump adds a maintenance item you do not need.
The listings here do not publish a calibration interval, so follow the manufacturer's instructions for the specific instrument. What the listings do publish is the starting point: factory calibration with a stated trace on P3, P7 and P8, and a zero and span menu on P1 and P8. A span gas check is how you find out whether the sensor has drifted since it left the factory.
No. Carbon monoxide and natural gas are different gases and need different sensors. A CO meter will read zero in a room full of natural gas. If the call is a gas smell rather than a combustion problem, you need a combustible gas detector, and the two instruments do different jobs.
No. A low level alarm is a residential safety device that responds earlier than a standard alarm, but it still only alarms; it does not give you a calibrated reading you can act on for combustion adjustment. A meter gives you a number. The two are complementary, not interchangeable.
The reason to reach for this meter on a furnace call is the pump and the probe. A flue or a heat exchanger is not always reachable with the meter body itself, and drawing a sample through a 4 ft probe lets you put the sensor where the gas actually is while you read the display somewhere you can see it. The listing states a 0 to 1000 ppm range at 1 ppm resolution, which is the resolution you need for low level investigation: at 1 ppm steps you can watch a reading fall as you adjust air, rather than watching a coarse display sit still. The stated USA NIST traceable calibration is the part that matters if your readings ever have to stand up to a question, because it gives you a documented starting point rather than an unknown one.
The colour display and graphing are more useful than they sound. Combustion CO is rarely a single steady number; it drifts as the appliance warms, as the fan cycles, and as you change the air setting. A graph shows the shape of that drift, and USB logging means you can keep the trace instead of writing numbers on a pad. Adjustable alarm levels with buzzer, LED and vibration let you set a working threshold for the job rather than accepting a factory one.
The listing does not publish an accuracy figure, a response time, or a battery life, so treat those as unknown. It also does not state a protection grade. The pump is a maintenance item: it pulls dust and moisture through the sample path, and a pump that is not drawing properly will give you a low reading that looks like a clean appliance. Check the flow before you trust a zero.
This is the pick when the job needs a number you can defend. The listing publishes three things the others mostly do not: a sensor tolerance of within plus or minus 5 percent of full scale, a response time under 30 seconds, and a pump flow of 0.5 litres per minute. Those three together tell you what the instrument can and cannot resolve. Five percent of a 1000 ppm full scale is a 50 ppm window, so this is not an instrument for arguing about a 5 ppm difference; it is an instrument for establishing that a reading is 200 ppm and not 20, and for proving that it fell after you worked on the appliance. The stated response time tells you how long to hold a sample before the number means anything.
The pump makes it a point sampler or a continuous sampler, which is the distinction that matters in combustion work. Point sampling lets you put the probe in a specific place, such as a register or a flue test port, and read that place. Continuous sampling lets you leave it drawing while you watch the appliance through a cycle. Time stamped alarm logging is the documentation feature: when an alarm trips, the instrument records when, which is what you need if you are handing the job to somebody else.
The listing states water, dust and explosion proof, which is a strong claim for a pumped instrument and is worth treating as the manufacturer's statement rather than a guarantee about your particular environment. Zero and span calibration options are published, so the instrument can be adjusted in the field if you have span gas. As with any pump, the sample line is the weak point: a blocked or wet line reads low, and a low reading on a furnace is the one that gets people hurt.
The Inspector 2 is built around the way a technician actually moves through a building. The range starts at 1 ppm and runs to 1999 ppm, so the same instrument covers a clean basement and a badly spilling appliance without swapping devices. A response under 5 seconds is quick enough to sweep a room and watch the number change as you approach a source, which is how you find the appliance that is producing CO rather than the one you happened to point at first. The listing states EN 50271 certification for professional occupational use, which is the kind of published basis a commercial client may ask about.
The thresholds are set at 35 ppm and 200 ppm, described as NIOSH occupational levels. Those are working thresholds, not household alarm thresholds, and that difference is the whole point of carrying a meter rather than relying on the alarm on the wall. The MAX mode logs the peak reading, which is useful when you cannot watch the display continuously; you can leave the meter running and read the highest number it saw. IP67 means it survives being dropped in a puddle or hosed off, which matters in a mechanical room.
The listing is explicit that it is not intended for formally classified hazardous locations, so it is not a confined space instrument and it does not replace a four gas detector for entry work. It also does not publish an accuracy figure or a calibration interval. The 80 to 85 dB audible output is loud enough to notice in a plant room but it is not a substitute for an alarm that wakes a sleeping occupant.
This is the meter to buy when the job is a series of straightforward point checks and you want the instrument to survive the van. The listing states IP67, waterproof to 3 ft, and a 20 ft drop test pass, which is the kind of durability that matters when a meter lives in a tool bag and gets used in basements and on roofs. The 0 to 1000 ppm range at 1 ppm resolution is the same working range as the more expensive instruments here, and the electrochemical sensor is the standard technology for this measurement.
Factory calibration with a stated USA NIST certificate is the feature that separates it from the anonymous meters. A meter that arrives calibrated gives you a known starting point, and the certificate is what you keep if a client asks how the reading was obtained. The two year battery life on a 9V lithium cell means you are not charging it every night, which is one less thing to remember before a service call.
The listing does not publish an accuracy figure, a response time, or alarm thresholds. It also does not publish logging or a pump. So this is a read the display and write the number down instrument, not a documentation instrument. It has no alarm function described beyond the display, which means it is a meter and not an alarm; do not leave it in a home as protection.
The CT-580 is the cheapest instrument here that still gives you a calibration path. The listing states a zero and standard gas calibration setting in the advanced menu, which means you can adjust it against span gas rather than throwing it away when it drifts. That single feature is what makes a budget meter usable for servicing rather than just for curiosity. The 0 to 1000 ppm range covers the working span, and the three alarm modes, light, sound and vibration, mean you will notice a threshold even in a noisy mechanical room.
The listing states a response within 0.3 seconds, which is a very fast claim for an electrochemical sensor and is more likely describing the display update than the sensor's full response to a step change in gas. Treat it as a fast display, not as a verified 0.3 second sensor response. The alarm history is genuinely useful: if the meter tripped while you were working, you can go back and see what it saw. The TFT colour display makes the number easy to read at a glance.
The listing does not publish an accuracy figure, a sensor type, a battery life, or a protection grade. It is described as a manual inspection tool, which is the correct framing: it is not a life safety alarm and should not be left in a home as one. The rechargeable battery means you need to remember to charge it, unlike the lithium primary cells on some other picks.
The ET110 is the meter for somebody who wants a clear yes or no rather than a data set. The listing publishes two alarm levels, 35 ppm and 200 ppm, and a short term exposure limit alarm. Those thresholds match the occupational reference points used on the other instruments here, so the ET110 will tell you when a space has crossed the same lines that a more expensive meter would flag. It is portable and it is simple, which means it gets used rather than left in the case.
The listing is short, and that is the honest summary of the instrument. It does not publish a measurement range, a resolution, an accuracy figure, a response time, a battery type or a protection grade. It does not publish logging, a pump, or a calibration menu. So you should treat it as a threshold indicator with a display, not as an instrument for low level investigation or for documenting a combustion problem. If you need to know whether a reading is 8 ppm or 12 ppm, this is not the meter for that question.
The STEL alarm is the useful part for servicing, because it responds to a rising exposure over a working period rather than a single instantaneous spike. But it is still a meter, not an alarm: it does not protect a sleeping household, and it should not be left behind as one.
Short answer: For furnace servicing, the two picks that carry the most usable data are P7 and P8, because both publish a pump, a stated calibration trace and computer logging, which is what you need to prove a reading before and after an adjustment. P6 is the strongest all-round handheld for a technician who wants a stated certification and a fast response without a pump. P3 is the value pick for a rugged point-and-read meter. P1 is the budget option and P2 is the simplest alarm-threshold meter. The residential alarms in the source list, P9, P10, P11 and P12, were excluded: they are life safety alarms for occupants, not instruments for combustion analysis, and P12 is a low level monitor rather than a servicing meter.
An alarm is a device that decides for you. It sits on a wall, it watches continuously, and when carbon monoxide crosses a threshold it makes a noise loud enough to wake somebody. A meter is a device that tells you a number while you are standing there holding it. Those are different jobs, and confusing them is the most common mistake in this category.
In furnace servicing the meter is the working instrument. You use it to find out whether the appliance is producing carbon monoxide, how much, and whether the number changes when you adjust the air. The alarm is what protects the household after you leave. If you carry a meter and no alarm, the house has no protection. If you carry an alarm and no meter, you cannot diagnose anything. Most technicians need both, and they are not substitutes for each other. If you want the alarm side of the decision, the carbon monoxide detector comparison covers that ground, and the difference between a gas leak detector and a CO detector is worth reading before you buy anything for a call that might be either.
Carbon monoxide from a furnace is a combustion problem, and the reading is a diagnostic signal. A clean burning appliance produces very little CO, and the number should be low and stable. A number that climbs as the appliance warms, or that rises when the fan starts, points at a specific problem: a cracked heat exchanger, a blocked flue, or an air setting that is starving the flame. The meter does not tell you which of those it is, but the shape of the reading narrows it down.
This is why resolution matters more than range for most servicing work. A meter that reads in 1 ppm steps lets you watch the number move as you make an adjustment. A meter with a coarse display can sit on the same value while the appliance is still producing CO, which is worse than useless because it looks like a pass. Range matters for the other direction: you need enough headroom to read a badly spilling appliance without the instrument maxing out. The meters here that publish a range mostly state 0 to 1000 ppm, and P6 states 1 to 1999 ppm.
The thresholds you will see quoted, 35 ppm and 200 ppm, are occupational reference points rather than household alarm thresholds. They are useful as working lines: 35 ppm is a level at which you should be paying attention, and 200 ppm is a level at which you should be acting. But they are not the same as the thresholds a residential alarm uses, and a meter set to those levels is not a substitute for an alarm.
An electrochemical CO sensor drifts. It responds to the gas, but its output also changes with temperature, humidity and age. A meter that was accurate when it left the factory will not stay accurate forever, and the instrument does not know it has drifted. That is why calibration is the feature that separates a working instrument from a toy.
There are two levels of it. Factory calibration with a stated trace, which P3, P7 and P8 publish, gives you a documented starting point. A zero and span menu, which P1 and P8 publish, lets you adjust the instrument in the field against a known gas. If you do combustion work regularly, the span menu is the more valuable of the two, because it lets you check the instrument before a job rather than trusting a certificate from months ago. If you only need a spot check now and then, factory calibration is enough, provided you replace or recalibrate the instrument on the manufacturer’s schedule.
What none of these listings publish is a calibration interval. That is not a gap in the comparison; it is a gap in the listings. Follow the manufacturer’s instructions for the specific instrument, and if you are using the meter for work that has to be documented, keep the calibration record with the job.
The sensor gets all the attention and the sample path gets none. On a pumped instrument, the pump, the tubing and the probe are between the gas and the sensor, and anything that blocks or absorbs the gas will make the reading low. A low reading on a furnace is the dangerous direction, because it looks like a clean appliance. Check the flow before every job, and replace tubing that has been wet or contaminated.
On a non-pumped meter the equivalent problem is the sensor inlet. Dust, lint and moisture collect on it, and a blocked inlet slows the response. A meter that used to react in a few seconds and now takes half a minute is telling you something about its inlet, not about the gas. Keep the instrument in its case, keep it out of the rain even if it is rated for it, and do not store it in a hot van if you can avoid it.
Batteries are the other quiet failure. A rechargeable meter that was not charged is a meter that does not work, and the failure is silent until you need it. Lithium primary cells, like the 9V in P3 and the sealed cells in the alarms, avoid that problem but introduce a different one: they fail without warning at the end of their life. Whichever you choose, check the instrument before the call, not on the doorstep.
A carbon monoxide meter measures one gas. That is its strength and its limit. If the call is a gas smell, a refrigerant leak, or a suspected sewer gas problem, the CO meter will read zero and tell you nothing. For those jobs you need a different sensor, and the combustible gas detector comparison covers the leak search side. If the job is entering a confined space, a single gas meter is not enough: entry requires a multi gas instrument that also watches oxygen and the other hazards, which is a different category entirely.
And if the job is protecting a household, the meter is the wrong instrument. It measures while you are there and does nothing when you leave. The right answer for that job is an alarm, installed and maintained, and the meter is a tool you carry alongside it. Buying a meter to do an alarm’s job is the mistake this whole page is written to prevent.
The source list contained twelve products. Six were excluded because they are residential life safety alarms or low level monitors rather than servicing meters: P9, P10 and P11 are UL 2034 residential alarms, and P12 is a low level monitor with a stated 10 ppm display start and a 25 ppm buzzer. They belong on an alarm page, not a servicing meter page. The remaining six were compared on the specifications their listings publish: measurement range, resolution, alarm thresholds, pump, logging, calibration data and protection grade. Where a listing is silent, the table says Not stated rather than filling the gap. No laboratory testing was performed and no measurements of our own are published.
Seven calibration kits and components compared on blend, cylinder size, regulator type, and what the listing does and does not publish.
Read Guide→
Five electronic sniffers and one brass tank gauge, sorted by probe reach and stated detection range, for the four places a grill actually leaks.
Read Guide→
Six leak detectors that cover R410A, sorted by sensor technology, published sensitivity and what each one costs to keep running.
Read Guide→