I installed two new smoke detectors in my house a few months ago and have been testing them using the 'test' buttons.

Today I decided to give them a more realistic test by placing a burning and then extinguished match underneath them. To my surprise neither was activated.

Thinking this may not have been enough to trigger a modern smoke detector I then removed them from the wall and placed them above a large saucepan of burning paper for around 3 minutes. There was lots of smoke and the flames were only inches below the smoke detectors. Yet again, neither detector sounded an alarm.

I contacted the manufacturer, and they say that this is 'normal' and that their smoke detectors are 'burned toast proof'.

To me this sounds like BS. If 3 minutes sitting directly above flames and plumes of smoke is not considered enough to set the alarm off then just how alight would my house have to be before I was alerted?

So, how should I properly test these smoke detectors in a way that the manufacturer cannot simply brush off?

  • 1
    @JamesNapier What type are they?
    – Jason C
    Commented Nov 29, 2015 at 22:47
  • 1
    They are optical smoke alarms made by FireAngel.
    – James
    Commented Nov 30, 2015 at 12:54
  • 1
    @james Fwiw, your test results are not unusual then, and the manufacturer's explanation is pretty reasonable and likely sound. Photoelectric alarms are inherently "burnt toast proof". I usually only use that type in the kitchen, with a combo type everywhere else. If you aren't choosing appropriate types to begin with, or you didn't check the type when you bought them, you're barking up the wrong tree a bit with the tests, anyways, as far as safety goes. :)
    – Jason C
    Commented Nov 30, 2015 at 13:44
  • 1
    possible duplicate of diy.stackexchange.com/questions/46351/… and diy.stackexchange.com/questions/37091/…
    – Adam Davis
    Commented Nov 30, 2015 at 14:47
  • 2
    I suppose setting fire to your house is out of the question?
    – Jon Story
    Commented Nov 30, 2015 at 16:39

7 Answers 7


Proper test methods depend on type, as there are different technologies employed depending on model. Some use a photo-sensitive element to "see" smoke, others use a radioactive element to detect products of combustion (even invisible ones), others "see" flames with an infrared detector and others detect heat.

You should test them exactly according to manufacturer's instructions and if they do not perform then replace them immediately. Like any other consumer product, you should deal with the seller and/or supplier and manufacturer for grievances related to defective products.

Exercise your right not to buy crap products and you will end up with reliable smoke/fire/CO detectors.

  • I mostly agree although I would add that the testing procedure for smoke and CO alarms in the US is very specific about what should and should not trigger an alarm, so I would expect all alarms that are UL listed to be pretty similar in how they behave, cheap or not.
    – Hank
    Commented Nov 29, 2015 at 22:21
  • @HenryJackson Inexpensive single sensor detectors will vary significantly in performance based on the sensor type they use. Ionization sensors respond faster to fast igniting, hot, and relatively smoke free fires. Photo-electric sensors respond faster to the large smoke particles caused by smoldering fires. Both types of fires are common in the home; if you see alarms marked as kitchen/bedroom models, the former will be ionization the latter photoelectric (because steam/burnt toast will readily false alarm a photo-electric sensor). Commented Nov 29, 2015 at 23:55
  • ...Higher end models will include both types of sensors allowing them to detect both types of fires while minimizing the false alarm rate. (The latter are often also bundled with a CO detector, but only as a value add/combination unit; not to assist in fire detection.) Commented Nov 29, 2015 at 23:56

Not enough information. Are they photoelectric or ionization type, or combo? Ionization alarms generally respond more slowly to smoldering fires, and photoelectric alarms are a bit less sensitive to small amounts of particles that cause false alarms (like bacon smoke or shower steam). Also various manufacturers may produce alarms of various sensitivities. The "burnt toast proof" claim is fairly reasonable, esp. for a photoelectric alarm, or a reduced sensitivity ionization alarm.

In any case there are only two valid tests: According to manufacturer's instructions, or light your house on fire. You will not be able to accurately test your smoke alarm using your own home-brewed methods as there are many different types of fires, and you cannot test all possibilities or know which type of fire may actually start in your home. It is unlikely that any results of such a test will represent the actual behavior of the alarm in a fire (for better or worse). If you are not confident in your smoke detector then it is reasonable to buy a new one from a different manufacturer.

It's all kind of all over the place, anyways (or, some less panicky relevant info).

My Kidde combo alarm trips just from sawdust from my table saw 20 feet away, but I want that sensitivity as I keep it in a room with a lot of flammable material. My kitchen photoelectric alarm, same manufacturer, doesn't register unless I'm actually burning something for a while, or cooking with a lot of smoke, e.g. it doesn't care about a little bit of bacon but searing a steak gets it every time. I chose these alarm types on purpose.

I guess it's also obligatory to point out that even the best, most accurate smoke alarm on the planet won't do you much good if it's not the correct type for the situation and/or not placed in an effective location. Always read and follow the manufacturer's recommendations for purpose and placement.


I can't vouch for them, but there are aerosol cans of some sort of mist sold specifically for the purpose of testing smoke detectors. Theoretically, following their instructions should give you a better test of the smoke detector. Doing better than that probably requires a real test setup capable of producing a precisely calibrated mist or smoke.

  • 1
    I've seen those. The only thing it really tests is if your particular smoke detector responds to that particular type and concentration of mist being sprayed at it for however long you sprayed it, which may or may not be representative of the infinite types of fire and smoldering patterns that could pose a danger in your home. I would not use home brew test results as any sort of quality judgment for a smoke detector, good or bad.
    – Jason C
    Commented Nov 29, 2015 at 21:51

Simple Testing

For ionization and photoelectric smoke detectors you can purchase aerosol sprays that provide the combustion products and particles typical to a residential or commercial fire which should exercise your smoke detector.

To do this properly, you also need an enclosure so that the concentration of particles is sufficient to trigger the alarm. You can simply point and spray, but you'll end up using a significant amount of the testing aerosol per detector to trigger it.

An enclosure and integrated dispensing device, such as the Solo 330, will allow you to test multiple detectors with one can of aerosol:

enter image description here

Aerosol test sprays are available for CO alarms as well.

Heat rise detectors and other more specialized alarms require different testing techniques. If you need to test such devices, contact the manufacturer for testing instructions.

UL Testing

UL 217 defines the requirements and test methods for UL listed smoke alarms. Section 38 gives a complete description of the test method, which is not easily summarized. If you want to find out if your smoke alarms meet UL requirements you can read UL 217 and follow the instructions to replicate their testing.

UL 217 (1993)

  • Is this different than diy.stackexchange.com/a/78985/20671?
    – Jason C
    Commented Nov 30, 2015 at 13:40
  • I've seen those. The only thing it really tests is if your particular smoke detector responds to that particular type and concentration of mist being sprayed at it for however long you sprayed it, which may or may not be representative of the infinite types of fire and smoldering patterns that could pose a danger in your home. I would not use results of a test like that as any sort of quality judgment for a smoke detector, good or bad.
    – Jason C
    Commented Nov 30, 2015 at 13:48
  • @JasonC Please consider adding your distrust of professional smoke alarm testing products and devices to your existing answer, where you can expand on how your expertise in this subject leads you to believe these tests are invalid.
    – Adam Davis
    Commented Nov 30, 2015 at 13:58
  • It is mentioned in my existing answer to an extent that I am satisfied with.
    – Jason C
    Commented Nov 30, 2015 at 14:00

I would not trust an alarm I cannot test with real fire and smoke. Detector should detect fire while it is small, before whole house is in fire. Testing this is not too difficult. We have perhaps 30 years old ionization detectors. Currently one of those seems to be on the table instead of near the roof. It has again reacted to small amount of smoke coming from oven. I guess those would react to a warm breeze from the oven door, if placed above it. (If oven is not absolutely clean, it could produce very small amount of smoke.) Occasional false alarm is not too bad. These do not alarm without some smoke source. Of course there should not be alarm every time oven door is opened.


Lighting a fire or using aerosol spray near them will coat their sensors with smoke or other particles, and they might not work properly after that. They are designed to be tested by depressing the button; that's all you need to do. If they're ten years old, they should be replaced.


You can take them to your local fire department and they will assess the smoke detectors. What you are describing doesn't sound right to me. Usually a warm breeze from the oven door is enough to set off a modern smoke detector.

  • 7
    If a "warm breeze" from the oven door sets off your smoke detector, I would look at getting a better one. A smoke detector that puts out false positives at least twice a day is not going to be a very effective first warning system for a real fire. "Oh, my husband must be warming up some croissants again. I can simply ignore that sound." Commented Nov 29, 2015 at 22:48
  • my smoke detectors aren't set off by that, it was just an illustration to point out they should be more sensitive. Commented Nov 30, 2015 at 1:09
  • 3
    +1 for the fire department suggestion but -1 for the warm breeze part. Try not to instill wrong info into people. Do proper research first :D
    – Aloha
    Commented Nov 30, 2015 at 12:36
  • Having the fire department assess is probably the best answer in this thread.
    – Jason C
    Commented Nov 30, 2015 at 13:58
  • 1
    @LightnessRacesinOrbit: I wish there were some standard for multiple sensitivity levels with a facility for temporary and permanent adjustments, as well as a local alarm which could be easily temporarily silenced but if not silenced would trigger a system-wide alarm within a short time. If there shouldn't be any sort of combustion products in an area, an immediate alarm from even a tiny level of combustion products may enable someone who is within earshot to e.g. unplug an overheating appliance and thereby prevent a fire.
    – supercat
    Commented Nov 30, 2015 at 16:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.