Smoke detectors and dimmers

I've noticed in domestic smoke and heat detector's instructions it is stating that dimmers must not be on the same circuit.

Is this because of the noise that can be imposed?

Is it likely to result in an alarm not detecting a fire?

  • It is indeed due to conducted emissions from the dimmers, which slice the square mains cycle to give the dimming effect,

    Arguably it is an admission both that the filtering standards for permitted noise getting out of dimmers allow noise that is too high, and that the standards for filtering for immunity to  noise getting into smoke alarms are inadequate.

    Sadly there will be layouts in some buildings where being on another circuit in the same CU does not significantly protect against such effects and additional filtering would be a better solution. Bur a sentence in the instructions passing the buck is cheaper for the manufacturers.

    Without seeing the test results we do not know if the smoke alarms are blinded and do not sound, or are triggered to false alarm, in the presence of mains borne interference, but we can be pretty sure they do not perform to spec, or it would not have been mentioned at all.


  • So a code 2 or an FI then? I am beginning to think that guy Dave Cockburn was right all along, what’s not a potential danger these days!

  • What for ? It is probably impossible to show in a given installation with dimmers and alarms on the same or separated circuits if there will be a problem or not and the idea the splitting the circuits is going to help in all cases is as much a fiction as the idea that all shared circutis will have problems, It is a way for the makers to create enough confusion that it is not their problem.


  • Yes, apologies to the OP, I was being a tad tongue in cheek, I don’t know why as it really quite annoys me when respondents go off piste with silly remarks!

  • Some of the most useful stuff comes off this forum when things go 'off-piste' ! I'd not worry. 

    I'd hate to imagine someone failing an EICR because of a dimmer switch being added to a lighting circuit that also supplies a fire alarm, but equally I wonder how the makers imagine we are to assure ourselves that  all possible dimmer settings for any given alarm installation will be OK.

    As the power-line comms (homeplug etc) shows, waveforms with high frequency components (in their case wanted data, in the case of the dimmer, just interference) are happily propagated quite a distance around typical house wiring - the HF does not know it was only designed to carry  50Hz, and travels along the wires like a transmission line, albeit a rather kinky one that radiates like an antenna on the corners and where L and N split up for the light switch, and actually fuses and MCBs are by and large no great barrier, so all circuits in the building, and possibly in the neighbours buildings too,  are subjected to more or less of the same.

    The solution of course is filtering, but no-one wants to mention that as it is awkward to retrofit and unclear who has to pay for it.

    (and radio folk do not like power line comms or poorly suppressed dimmer switches as that interference can have an effect from a long way off )