Hardwired smoke detector that can ignore intermittent electrical faults?
October 12, 2021 4:43 PM Subscribe
I have a set of hardwired smoke detectors which tend to go off randomly around 3am. After much suffering, I believe I have identified the problem as an intermittently loose connection affecting one location in the network. Is there any hardwired detector that can be programmed to ignore electrical faults of less than some duration?
Right now I simply disconnected the problem location, and lo blessed slumber! If I connect any "normal" hardwired smoke detector at that location, it will cause the entire network to go off in error around 3am within a few weeks of being installed.
I have tried three different new dumb hardwired detectors, and all three went off in the middle of the night for no reason. Steam isn't a problem, dust isn't a problem, and bugs are unlikely (I have a pest control service that uses repellents at the perimeter of the house). I have even tried three different styles of detection technology, they all still failed in the middle of the night.
I have hired two different electricians, and neither was able to identify the fault. They both think the wiring is fine, so I'm guessing there's a connection that's loose somewhere that only becomes loose when they cool down in the middle of the night and the wires shrink fractionally. The overwhelming majority of electricians in my area are uninterested in pursuing this matter, including low-voltage specialists and alarm specialists. I have tried to check continuity myself, and everything seems fine. I have replaced all accessible wirenuts with levernut connectors so the probable issue is in a wirenut that is not easily accessible (I have no idea where this could be).
I like the idea of hardwired detectors, but am about to replace the problem location with a battery detector (this is against building code in my area).
So: are there any smoke detectors that are able to ignore transient power issues and continue operating without going off? I honestly don't care if they chirp at me, but thus far with three different brands they go into full alarm mode.
(I have checked prior questions, hence all of my details about testing etc)
Right now I simply disconnected the problem location, and lo blessed slumber! If I connect any "normal" hardwired smoke detector at that location, it will cause the entire network to go off in error around 3am within a few weeks of being installed.
I have tried three different new dumb hardwired detectors, and all three went off in the middle of the night for no reason. Steam isn't a problem, dust isn't a problem, and bugs are unlikely (I have a pest control service that uses repellents at the perimeter of the house). I have even tried three different styles of detection technology, they all still failed in the middle of the night.
I have hired two different electricians, and neither was able to identify the fault. They both think the wiring is fine, so I'm guessing there's a connection that's loose somewhere that only becomes loose when they cool down in the middle of the night and the wires shrink fractionally. The overwhelming majority of electricians in my area are uninterested in pursuing this matter, including low-voltage specialists and alarm specialists. I have tried to check continuity myself, and everything seems fine. I have replaced all accessible wirenuts with levernut connectors so the probable issue is in a wirenut that is not easily accessible (I have no idea where this could be).
I like the idea of hardwired detectors, but am about to replace the problem location with a battery detector (this is against building code in my area).
So: are there any smoke detectors that are able to ignore transient power issues and continue operating without going off? I honestly don't care if they chirp at me, but thus far with three different brands they go into full alarm mode.
(I have checked prior questions, hence all of my details about testing etc)
Response by poster: To clarify: I have tried a Kidde i2060A, which failed (new 9v), and a different brand of battery backup (First Alert iirc, unknown model #) which also failed.
...this is without AC running, furnace turned off, clean air (no forest fires), no steam, well above dew point (not condensing in unit), no dust (new detectors), no recent cooking, essentially zero interior household chemical usage and a largely bug-free environment (one harmless spider in the last year, who seems to have starved to death). I'm at a loss if it's not an electrical matter, are there additional failure modes?
posted by aramaic at 5:59 PM on October 12, 2021
...this is without AC running, furnace turned off, clean air (no forest fires), no steam, well above dew point (not condensing in unit), no dust (new detectors), no recent cooking, essentially zero interior household chemical usage and a largely bug-free environment (one harmless spider in the last year, who seems to have starved to death). I'm at a loss if it's not an electrical matter, are there additional failure modes?
posted by aramaic at 5:59 PM on October 12, 2021
Best answer: If you still have one of the models with battery backup, what happens if you connect it to the alarm wire (red) and neutral (white wire), but don't connect the hot (black) AC wire? Does the system still trigger at 3AM?
posted by RichardP at 6:19 PM on October 12, 2021
posted by RichardP at 6:19 PM on October 12, 2021
Also, do the other hard-wired smoke detectors in your system have battery backup with fresh batteries? What model are they? Have you tried placing one of the new hard-wired smoke detectors with battery backup at the problematic location, but not connecting any of the wires? Does that alarm still go off at 3 AM?
posted by RichardP at 6:48 PM on October 12, 2021
posted by RichardP at 6:48 PM on October 12, 2021
Best answer: Have you considered putting a power conditioner on the line? I'm not sure you would want to do this permanently (by far the cheapest way would be buying a plug-in one which would not play well with the electrical code) but it could be a good diagnostic tool to see if the electrical supply is really what's the matter.
For that matter I am wondering if there is any large electrical load that could be kicking on in your house at 3am. It seems unlikely that it would be happening repeatedly at such a similar time if it was heat related, since the timing would depend on how cold it was each night, right?
posted by goingonit at 7:02 PM on October 12, 2021
For that matter I am wondering if there is any large electrical load that could be kicking on in your house at 3am. It seems unlikely that it would be happening repeatedly at such a similar time if it was heat related, since the timing would depend on how cold it was each night, right?
posted by goingonit at 7:02 PM on October 12, 2021
Can you tell us anything more about the problematic location? Is it in the basement or attic?
Do you have a water softener or water filter or anything else that would have something scheduled at 2 or 3am? Do you have a sump pump? A cat? Do you run the dishwasher or clothes dryer before bed?
Are those smoke detectors with the 10 year battery up to code?
posted by cali59 at 7:26 PM on October 12, 2021
Do you have a water softener or water filter or anything else that would have something scheduled at 2 or 3am? Do you have a sump pump? A cat? Do you run the dishwasher or clothes dryer before bed?
Are those smoke detectors with the 10 year battery up to code?
posted by cali59 at 7:26 PM on October 12, 2021
Check with the fire department and see what they say. When we had a similar problem one of the firemen told me they had multiple battery-operated smoke detectors up in their house, and thought it was adequate, even though not up to code. They might not come right out and direct you to do something against code, but if you asked them for their feelings about it they might not think that battery powered detectors are more dangerous than up-to-code ones that go off at 3 am.
posted by Vatnesine at 7:55 PM on October 12, 2021
posted by Vatnesine at 7:55 PM on October 12, 2021
Checking with your local jurisdiction is a great idea, but the new 10 year sealed lithium battery detectors are what we’ve been installing in people’s houses for years. I would recommend trashing the old detectors and placing the new units in the bedrooms and great room(s). Make sure to write the date on the bottom!
posted by sara is disenchanted at 8:18 PM on October 12, 2021
posted by sara is disenchanted at 8:18 PM on October 12, 2021
Best answer: Some questions about your troubleshooting. Hopefully, this will just serve to confirm that you've already gotten the correct answers. I do, however, doubt your diagnosis, since if detectors were designed to detect power faults, the manuals would mention it, and if this were a common reason for nuisance alarms, I'd assume discussion about it would be far more common.
How long after installation did it take for each of the new ones to trigger the first time? And after the first occurrence, how long until the next occurrence? You note "within a few weeks", so I'm presuming this is somewhat intermittent. How long have you had the problematic location disconnected? With intermittent issues, it's unfortunately quite easy to mis-diagnose the cause.
Is the interconnect activating? (Are all the alarms in the house going off, or only the one?) If they're all going off, have you looked at all the detectors to determine which one was the originating detector? The Kidde 12060A you linked says that only the originating detector will have a flashing red LED, all others will be dark. 3am isn't the most fun time to be roaming around the house, but if you can check every detector, you can verify which one is initiating the alarm. If you have, for example, a CO detecting unit in the basement that's activating, that's going to change your next steps...
Are all the detectors in the house of compatible brands/models? Not all detectors are compatible. Kidde's compatibility chart is here; note that "Connection of Kidde products to a non-specified manufacturer's interconnect system, or connection with non-specified equipment from another manufacturer into an existing Kidde system could result in nuisance alarming, failure to alarm, or damage to one or all of the devices in the interconnect system." [emphasis added]
If you press the "Hush" button on the originating detector, does it reactivate after the 8 minute hush period, or does it remain deactivated once the hush period expires? Ongoing alarming would tend to rule out a wiring issue. I'm not sure if it's possible to hush at an alarm that's not the originating alarm, so you may have already done this.
Are all the detectors installed per the manufacturer's requirements? The Kidde ones require all the detectors to be on the same circuit. I presume your electricians would have checked this. I'd also expect that this would be more likely to cause damage or prevent the interconnect from working than cause nuisance alarms.
Assuming all the detectors are in fact compatible, if you disconnect the interconnect (red) wire at the location in question, but connect the power (black/white), does the detector still go off? That would isolate the problem to the detector itself, or the power to the detector, not the interconnect wiring. Coupled with RichardP's idea to disconnect the line(black) and leave the interconnect (red) connected, you could rule out some of the wiring.
posted by yuwtze at 8:30 PM on October 12, 2021 [2 favorites]
How long after installation did it take for each of the new ones to trigger the first time? And after the first occurrence, how long until the next occurrence? You note "within a few weeks", so I'm presuming this is somewhat intermittent. How long have you had the problematic location disconnected? With intermittent issues, it's unfortunately quite easy to mis-diagnose the cause.
Is the interconnect activating? (Are all the alarms in the house going off, or only the one?) If they're all going off, have you looked at all the detectors to determine which one was the originating detector? The Kidde 12060A you linked says that only the originating detector will have a flashing red LED, all others will be dark. 3am isn't the most fun time to be roaming around the house, but if you can check every detector, you can verify which one is initiating the alarm. If you have, for example, a CO detecting unit in the basement that's activating, that's going to change your next steps...
Are all the detectors in the house of compatible brands/models? Not all detectors are compatible. Kidde's compatibility chart is here; note that "Connection of Kidde products to a non-specified manufacturer's interconnect system, or connection with non-specified equipment from another manufacturer into an existing Kidde system could result in nuisance alarming, failure to alarm, or damage to one or all of the devices in the interconnect system." [emphasis added]
If you press the "Hush" button on the originating detector, does it reactivate after the 8 minute hush period, or does it remain deactivated once the hush period expires? Ongoing alarming would tend to rule out a wiring issue. I'm not sure if it's possible to hush at an alarm that's not the originating alarm, so you may have already done this.
Are all the detectors installed per the manufacturer's requirements? The Kidde ones require all the detectors to be on the same circuit. I presume your electricians would have checked this. I'd also expect that this would be more likely to cause damage or prevent the interconnect from working than cause nuisance alarms.
Assuming all the detectors are in fact compatible, if you disconnect the interconnect (red) wire at the location in question, but connect the power (black/white), does the detector still go off? That would isolate the problem to the detector itself, or the power to the detector, not the interconnect wiring. Coupled with RichardP's idea to disconnect the line(black) and leave the interconnect (red) connected, you could rule out some of the wiring.
posted by yuwtze at 8:30 PM on October 12, 2021 [2 favorites]
I doubt the problem is a transient power issue. Think about it: if you had a true power failure, would all smoke detectors on the interconnect start beeping with an alarm?
I am however very curious as to what the correct solution is because I have this issue at my house too, and so far the only “solution” that that works is leaving the originating detector location (a bedroom) disconnected—and therefore unprotected. I too have tried multiple models of compatible detectors, and the model and backup battery levels have been irrelevant to preventing the random 3am whole-house alarms.
posted by stopgap at 9:28 PM on October 12, 2021
I am however very curious as to what the correct solution is because I have this issue at my house too, and so far the only “solution” that that works is leaving the originating detector location (a bedroom) disconnected—and therefore unprotected. I too have tried multiple models of compatible detectors, and the model and backup battery levels have been irrelevant to preventing the random 3am whole-house alarms.
posted by stopgap at 9:28 PM on October 12, 2021
Can you just disconnect the signal wire?
If the single alarm in that room is still going off... maybe you have a CO leak? (If they're combo units)
I would personally feel fine with just a battery unit.
posted by flimflam at 10:14 PM on October 12, 2021
If the single alarm in that room is still going off... maybe you have a CO leak? (If they're combo units)
I would personally feel fine with just a battery unit.
posted by flimflam at 10:14 PM on October 12, 2021
Best answer: This sounds potentially like an EMC issue, i.e. electromagnetic interference from some other source that is being coupled onto your alarm wiring, and then being misinterpreted as an alarm signal.
This could be from a source outside your house - some nearby transmitter of some kind that goes off at 3am. That sort of schedule would not be unusual for some M2M/IoT monitoring & metering devices that use narrowband radios or the cellular network. Ever heard pulsing sounds on a radio/hifi just before a 2G GSM phone registers a call/text? Think that sort of thing, but happening on your alarm wiring.
In theory, all devices are supposed to be designed and tested to avoid creating such interference, and to avoid being affected by it. In practice, even if the testing is good, there are sometimes just combinations that don't get along.
I'm not sure what signalling is used in these hardwired alarm systems, but I'd expect it to be as simple as possible for reliability, and for the design to lean on the side of sounding an alarm if in any doubt. Those two factors mean it might be particularly susceptible to certain interfering signals.
posted by automatronic at 5:12 AM on October 13, 2021
This could be from a source outside your house - some nearby transmitter of some kind that goes off at 3am. That sort of schedule would not be unusual for some M2M/IoT monitoring & metering devices that use narrowband radios or the cellular network. Ever heard pulsing sounds on a radio/hifi just before a 2G GSM phone registers a call/text? Think that sort of thing, but happening on your alarm wiring.
In theory, all devices are supposed to be designed and tested to avoid creating such interference, and to avoid being affected by it. In practice, even if the testing is good, there are sometimes just combinations that don't get along.
I'm not sure what signalling is used in these hardwired alarm systems, but I'd expect it to be as simple as possible for reliability, and for the design to lean on the side of sounding an alarm if in any doubt. Those two factors mean it might be particularly susceptible to certain interfering signals.
posted by automatronic at 5:12 AM on October 13, 2021
This sounds potentially like an EMC issue, i.e. electromagnetic interference from some other source that is being coupled onto your alarm wiring, and then being misinterpreted as an alarm signal.
Definitely plausible. I have a smoke detector which is reliably triggered to activate by a nearby WiFi/4G hotspot, and will trigger even if the devices are separated by 6 feet (2 meters).
Is it possible you (or an upstairs neighbor) have some nearby devices?
posted by soylent00FF00 at 7:54 AM on October 13, 2021
Definitely plausible. I have a smoke detector which is reliably triggered to activate by a nearby WiFi/4G hotspot, and will trigger even if the devices are separated by 6 feet (2 meters).
Is it possible you (or an upstairs neighbor) have some nearby devices?
posted by soylent00FF00 at 7:54 AM on October 13, 2021
Wired smoke detectors don't function by monitoring a constant signal rather when put into an alarm state they put voltage on a wire to alert the other alarms. Loose connections prevent them from operating rather than cause them to operate. (Note that this is different from systems with a fire alarm panel like found in apartment buildings and commercial spaces which monitor a constant signal to warn against tampering.)
So it is likely something else is causing your problem.
Irregardless here regulations merely require interconnection rather than hardwiring. If that is the case in your area (or you don't mind breaking the technical rules) something like these battery operated wireless interconnect detectors can completely bypass your wiring while providing full protection.
posted by Mitheral at 10:10 AM on October 13, 2021
So it is likely something else is causing your problem.
Irregardless here regulations merely require interconnection rather than hardwiring. If that is the case in your area (or you don't mind breaking the technical rules) something like these battery operated wireless interconnect detectors can completely bypass your wiring while providing full protection.
posted by Mitheral at 10:10 AM on October 13, 2021
Response by poster: Thank you all; I'm marking all of the answers which point to things I'll be trying.
posted by aramaic at 11:15 AM on October 14, 2021
posted by aramaic at 11:15 AM on October 14, 2021
« Older 3rd Grade Homework: there has got to be a better... | Betabrand pants: Size clarification, please? Newer »
This thread is closed to new comments.
posted by RichardP at 5:04 PM on October 12, 2021