Bizarre issue with Eico smoke detectors

[dom]

I've got mains-powered, battery-backed, interlinked Eico smoke detectors (hard-wire interlink, not RF).

1 test/hush/locate button, 1 heat detector, 2 monoxide detectors, 3 ionisation detectors.

All on their own RCBO circuit. All wiring thoroughly tested for insulation & continuity.

There's a way in which they will *consistently* mistrigger.

My laptop, running, if I unplug it briefly (say less than 5 minutes), then plug it back in - no effect. If I unplug it for some time, running, (say 30 minutes), then plug it back in - it will trigger or partially trigger the alarms.

(plug/unplug at the laptop - charger running continuously)

This happened a few times on the upper floor sockets circuit, one or more alarms sound briefly.

Have just done this again, this time with the laptop charger plugged into the kitchen sockets circuit. This time it fully triggered the alarms, more than one alarm sounded continuously.

Interesting thing - I tried the 'hush' button - no effect. Then I tried the 'locate' button - and it silenced all the alarms.

Suggests to me that somehow the laptop/laptop-power-supply has modulated the alarm signal onto the interlink wire between the detectors.

Anyone ever 'scoped the interlink signal to see what it looks like?

Other thoughts?

[Mathew Newton]

This may not be very helpful, but I always thought the interlink 'signal' on the Eico alarms were just a straightforward 240v AC switched feed.

I'm not sure what that thus belief is based on though - be it measurement, product literature or perhaps outright assumption (guessing)...

Mathew

[ARW]

I do not believe that any alarm of this sort of set up has a 230V "switched link". It's just not going to work if there is a powercut and you are relying on the batterys.

I did once post some links to show that the interconnect was a 9V dc signal. Dunno where those links are now.

[Mathew Newton]

Very good point. I can therefore surmise that my assumption was based on an assumption, and an incorrect one at that.

Apologies for the noise.

Mathew

[Graham.]

Have a listen in a MW radio in the same room as the laptop/charger.

The *cause* is probably the laptop and or charger but the *problem* is undoubtedly the alarms for not having sufficient EMC immunity.

Years ago I had some wired alarm PIRs they would trigger when a taxi or police car drve past with the radio keyed.

I had a National Band III transceiver for work in my car that would also trigger them.

[Brian Gaff]

I wonder what happens in a thunderstorm. Seems to me that the wiring is sensitive to transients. Brian

[dom]

Never had it trigger in a thunderstorm, but yes - it must be sensitivity to transients.

My guess would be each device has a high value resistor pulling up the inpu t to a logic gate, and that logic-input is also linked to the interlink wir e. Any device can then make a low impedance connection to ground on the int erlink wire to trigger the other alarms. If that's correct, adding a small capacitor between the interlink wire and ground at one place, might do the trick.