The sensitivity of a smoke detector is determined using a tool that has a cloud chamber. The tool has a calibrated measurement system to show the amount of smoke is needed to set the detector into alarm.
The full explanation, though, is a bit longer than the question, but the full explanation is needed to get a good picture of the process of sensitivity testing.
Types of Testing
There are four types of testing for smoke detectors:
- Magnet Testing - Testing the electronics to see what the electronics will do if the smoke chamber of the detector detects smoke
- Function Testing - Testing to see if the smoke detector detects smoke
- Sensitivity Testing - Testing to see how well the smoke detector detects smoke
- Internal Testing - Testing to see how dirty the addressable smoke detector has become over time
To mimic what the smoke chamber inside a smoke detector would do, magnet testing uses a magnetic reed switch inside the smoke detector housing. The magnetic reed switch does not detect smoke partials in the air (which is what smoke is made up of). Instead of detecting smoke particles, the reed switch detects a magnet. What a magnet test does not show is whether the smoke chamber of the detector will detect smoke; fires do not carry magnets in their back pocket.
Making sure the detector is properly connected to the fire alarm control panel, and making sure the detector and the control panel are programmed correctly, magnet testing is used during construction and installation. However, to show whether a smoke detector detects smoke, only artificial smoke from a can (which is particles in the air) or real smoke (which is particles in the air) is used.
To show if a smoke detector will detect smoke, function testing uses artificial smoke from a can. To find out how to use the canned smoke, read the instructions on the can.
Function testing is not calibrated; the person testing can use lots of extra canned smoke, the person testing can use the right amount of canned smoke, or the person testing might not be using enough canned smoke. Function testing, though, does test to make sure the detector can detect smoke. If the artificial smoke doesn't set the smoke detector into alarm, the detector is faulty.
To determine how much smoke it takes to set a smoke detector into alarm, sensitivity testing uses a calibrated, portable cloud chamber. It uses a photo-reader, kind of like the smoke sensing chamber inside a photo smoke detector, to sense the amount of smoke in the cloud chamber.
The cloud chamber, or testing tool, completely encloses the smoke detector and starts out with clean air. For the measurement, artificial calibrated smoke is slowly introduced into the chamber.
As the smoke is slowly being introduced, it's read by the cloud-chamber's smoke sensor and by the smoke detector. When the amount of artificial smoke inside the chamber is great enough, the smoke detector goes into alarm.
Because the sensor has been reading the amount of smoke inside the cloud chamber all along, the smoke sensor can say the equivalent of "It takes this much smoke to set the smoke detector into alarm".
To automatically adjust their sensitivity to a pre-set standard, many addressable fire alarm systems adjust a smoke detector's sensitivity downward as dust buildup sticks to the inside of the detector's smoke chamber.
Photo smoke detectors use a reflective principle to detect smoke. Inside the smoke sensing chamber, a light beam is projected by an LED onto the black wall of the chamber. The chamber is constructed so a photo-pickup transistor cannot detect this light beam.
When particles in the air get into the smoke chamber, each particle reflects a small amount of the beam into the photo transistor. When there's enough particles to reflect enough light, the smoke detector goes into alarm. Either the addressable smoke detectors, or the smoke threshold setting inside the addressable control panels, automatically adjust to compensate for the dust particles stuck to the wall of the smoke detector. This adjustment, along with the set sensitivity of the smoke detector, can be interpreted as the sensitivity of the detector.
In essence, the sensitivity for all the smoke detectors on the system can be printed out for the record.
Sensitivity Testing Issues
The sensitivity of a smoke detector can be measured using a cloud-chamber sensitivity tester. One place to purchase these sensitivity testers is at:
Yes, these testers are very expensive, and need to be sent back for calibration after testing a lot of smoke detectors. I don't know of a cheaper option, though, for properly testing the sensitivity of smoke detectors.
Also, these measurements take time. Be prepared to spend at least four or 5 minutes on each detector to obtain the sensitivity.
If you are required to test sensitivity on the smoke detectors, you are going to have to keep accurate records on each smoke detector in the building. You will have to show, in writing, the results of testing each smoke detector at the time it is installed, after it's been in use a year, and something like every five years after that.
With a new building, all the detectors can be sensitivity tested at the same time. Replacement detectors or added detectors still have to be tested when they are installed, and after they have been in use for a year. Because no one cares if sensitivity testing is done more often, after the year in use testing is done, the replacement or added detectors can be tested along with all of the other detectors.
Just make sure to show accurate records for each smoke detector because, if for no other reason, the AHJ (Authority Having Jurisdiction) has been known to want to look at the records.