I have two Exelguard photoelectric smoke detectors that were installed at the same time about four years ago. They've both started, within a couple of days of each other, giving frequent spurious alarms. This seems a remarkable coincidence, and makes me wonder whether they're programmed to do exactly that. Even aging of the components would surely not be that consistent.
Sylvia.
A programmed obsolescence clock would perhaps not be that consistent either.
True enough; I can't see any crystal on the board, and why would there be?
But I don't have a statistically significant sample. A clock variation of +/- one percent gives a +/- 14 day spread over four years. An occurrence with 2 days could be chance.
The alternative is a common cause effect, but nothing unusual has happened in the house.
Sylvia.
Can you smell smoke?
Nup, and the detectors don't give spurious alarms at the same time. If there is a common cause effect, it's damaged the detectors somehow.
Sylvia.
A discharged battery will do that, unless they are powered from the mains.
New batteries.
And changed again, just in case.
Still misbehaving.
Sylvia.
Flat batteries?
Colin
Even mains powered units have a battery backup that needs replacing.
Consistent faulty manufacture in the same batch?
It would have to be very consistent indeed to achieve useful lives that were the same to within a fraction of a percent. As McAvity observed, even getting a clock this accurate is a challenge without a crystal.
Sylvia.
How about VOC's?
Acer had the misfortune to sell a pile of new computers to a government dept that moved into a newly built building with complete new fittings. They soon gained the reputation of being crap. The cause was VOCs from the new stuff. Not widely known or understood at the time.
It still seems rather improbable that the degradation would somehow combine with the inherent variation in components to produce a better than 1% similarity in failure time.
Its the similarity that's important. If they'd just failed within a few months of each other, I'd have reached the obvious conclusion - rubbish smoke alarms - but the better than 1% result is suspicious. Such results would usually only arise by design.
Sylvia.
N=2 doesn't support that conclusion, just a coincidence. They usually wait until you murder three husbands before they start looking for signs of murder.
Have something to add? Share your thoughts — no account required.
Ask the community — no account required