God-damned smoke alarms

I've had a few that false alarmed. I suspected a bug got into the works because they eventually cleared themselves, but I'm not certain it was that. Mostly it's the battery thing. The newer ones have permanent batteries claimed to last 10 years, but that's bull. One went out after just a small number. Then Costco would not accept just one, I had to return both of the two pack! That visit was the end of my Costco membership. They also would not stand behind their batteries even when they leaked in the package before ever going into equipment!!!

It will be interesting to see what you come up with... however, you may be greatly increasing the failure rate, that is the false negative reporting! You might need to use three and use a majority voting circuit. It will need to report the odd man out in order to preserve the reliability as units crap out. Testing may be more complex too. I guess we know why they don't do this in the first place.

Sensor ageing?

How about a camera confirming the smoke?

The smoke detector detects smoke long before it can be seen by a camera. It is critically important that smoke detectors be at the very top of the ceiling to gain those few extra seconds. If you aren't out of the house in the first few minutes of a fire, in many cases you won't make it out at all. The smoke fills the home very quickly, obscuring your way out and suffocating you on the way.

It should see the same if camera is next to detector on ceiling.

I am not saying to replace the sensor, but to confirm it before sounding the alarm. If all you need to see is smoke or not, should be easy to do with camera module. i.e. OV76XX. You only need 1 bit data and your FPGA perhaps.

Was it a smoke+CO detector?

CO detectors seem to be the problem. Maybe some other chemical vapor gets to them.

We get the 10-year smoke+CO ones, and they tend to break after a few years.

There will be nothing for the camera to see! The smoke detector is designed to be much more sensitive to smoke than a camera will be. Read up on the subject and stop making silly posts with no knowledge of the subject.

The one that was impolite last night is based on this chip:

If I use two, then I'm about doubling the false-negative failure rate, plus whatever failure rate my own circuit has. If I use three, with majority voting, then I'm probably ahead.

I'd use one LED per smoke alarm, that flashes when my circuit has prevented an alarm from sounding, indicating that a unit needs to be replaced, or during testing when one unit failed to detect smoke that the other two did.

The place it's installed is quite dark, so a flashing LED will be conspicuous.

Lookout for the next Darwin awards: "Home owner implements circuit to prevent false smoke alarms. Circuit catches fire, and burns the house down, with all inside lost."

Sylvia.

I don't know how sensitive it need be. If i can see smoke, camera should be able to see smoke. So, what is your definition of detectable but invisible smoke?

No, just an optical smoke detector.

Sylvia.

I was very surprised to find that the four alarms that were installed at the same time in my appartment all conked out within half a year after about 10 years. Very consistent quality of both battery and electronics.

Sylvia Else wrote: ===============

A look inside the shop revealed how absurd the installation was. The PIR unit was being triggered by passing traffic at night, whenever a vehicle shone it headlamps in the front window. The microwave unit was triggered by any of a number of cardboard signs hanging on single strings which rotated in the slightest draft.

One Sunday, I was able to corner the shop's handyman and asked him to wire the two so they operated in series. We were able to test that there was no way to move about in the covered area without BOTH alarms triggering at once.

There was never another false alarm.

.... Phil

On a sunny day (Thu, 19 Aug 2021 13:09:47 +1000) it happened Sylvia Else snipped-for-privacy@email.invalid wrote in snipped-for-privacy@mid.individual.net>:

Mine sounds when the 9V battery is low I did not even hear the smoke alarm upstairs until I needed to get something from the attic...

So later I bought a gas alarm that runs on mains...

What can you do? Run and watch... Oops, my cellphone is still in there...

My smell is perhaps better.... than the smoke and gas alarms.

Ed Lee snipped-for-privacy@gmail.com wrote in news: snipped-for-privacy@googlegroups.com:

You do not understand how they work then. They detect particulate, which is what smoke is, except they see it long before we do or a visible spectrum camera would.

Optical smoke detection fails as well, and so programming has to be done to differentiate false alarms from bona fide smoke events and that makes optical "detectors" far too expensive for the masses. And they would still false alarm more than the current method.

There is some Australian regulation that requires wired (house powered) alarms to be cross-wired so that all sound if one activates. Not the other way around. I don't know the precise rules but our electrician mentioned it when he installed new wiring a couple of years back. He also chucked out a couple of cheap ones that were prone to false alarms and installed better ones that aren't.

You might want to check the regulations, and ask what applies to you.

CH

Particles under 1 micron don't affect visible light, but they're still present, and are harmful because they get into the far reaches of your lungs.

Most of the really harmful stuff in wood dust is invisible, sub-micron stuff - which is why woodworking plants have their dust filters outside.

some might do it intentionally since they should be replaced after 10 years

Fit a mains powered one...

It's not a battery issue.

Sylvia.

It's not a chirping sound. It's a full-on alarm sound.

Sylvia.