Conversation

Notices

1. Embed this notice
   Ryan Castellucci :nonbinary_flag: (ryanc@infosec.exchange)'s status on Sunday, 26-Nov-2023 03:06:05 JST Ryan Castellucci :nonbinary_flag:

   Making LED bulbs work with triac dimmers is an abomination. The dimmer chops up the AC waveform, which was a clever hack for incandescents, but LEDs run on DC so the bulb has to have driver circuitry that not only rectifies and reduces the input voltage, but also figures out what the triac is doing and turn that into the appropriate voltages regardless and do whatever dimming it has to do that actually works with the LEDs.

   • Embed this notice
     Ryan Castellucci :nonbinary_flag: (ryanc@infosec.exchange)'s status on Sunday, 26-Nov-2023 04:02:19 JST Ryan Castellucci :nonbinary_flag:

     in reply to

     • Graham Sutherland / Polynomial

     @gsuberland It would not be difficult to inject a low frequency control signal onto the AC, but then everyone would have to standardize - or proprietary bulbs/dimmers I guess.

   • Embed this notice
     Graham Sutherland / Polynomial (gsuberland@chaos.social)'s status on Sunday, 26-Nov-2023 04:02:20 JST Graham Sutherland / Polynomial

     in reply to

     @ryanc yeah it's a huge pain in the ass.

   • Embed this notice
     David Schuetz (darthnull@infosec.exchange)'s status on Sunday, 26-Nov-2023 04:16:10 JST David Schuetz

     in reply to

     @ryanc I’ve wondered if, in the long term, we start to move to a secondary power standard for built-in lights that’s different from outlets.

     OTOH, we can’t even get X10/Insteon/Z-Wave/matter/whatever standardized.

   • Embed this notice
     Ryan Castellucci :nonbinary_flag: (ryanc@infosec.exchange)'s status on Sunday, 26-Nov-2023 04:29:01 JST Ryan Castellucci :nonbinary_flag:

     in reply to

     • Richard "mtfnpy" Harman (he/him)
     • David Schuetz

     @darthnull @xabean There's 0-10v, DALI, DMX, KNX, and a bunch of other things.

     Low voltage DC kind of has to be relative - the amperage gets pretty serious rather quickly, which requires big wires.

   • Embed this notice
     David Schuetz (darthnull@infosec.exchange)'s status on Sunday, 26-Nov-2023 04:29:02 JST David Schuetz

     in reply to

     • Richard "mtfnpy" Harman (he/him)

     @xabean @ryanc Yet Another.

     I’m just spitballing that maybe lighting circuits should move to low voltage DC, maybe directly dimmed but even better digitally controlled. Dunno if it would ever happen. Probably not in my lifetime.

   • Embed this notice
     Richard "mtfnpy" Harman (he/him) (xabean@infosec.exchange)'s status on Sunday, 26-Nov-2023 04:29:03 JST Richard "mtfnpy" Harman (he/him)

     in reply to

     • David Schuetz

     @darthnull @ryanc I *think* the lighting standard you're looking for is DMX?

   • Embed this notice
     Ryan Castellucci :nonbinary_flag: (ryanc@infosec.exchange)'s status on Sunday, 26-Nov-2023 05:08:31 JST Ryan Castellucci :nonbinary_flag:

     in reply to

     • John Ripley

     @jripley How would power negotiation work with the lights effectively on a string?

     48VDC would be my proposed voltage as well, though.

     Maybe run a dead simple control channel that's just manchester coded 16 bit channel id + 16 bit brightness level.

   • Embed this notice
     John Ripley (jripley@mastodon.social)'s status on Sunday, 26-Nov-2023 05:08:32 JST John Ripley

     in reply to

     @ryanc 48VDC with negotiated power and comms side channel (pretty much USBC) would be wonderful and relatively low current. But it would need all your lighting cables re-installed. Still thin enough that if you needed more than, say 3A, you could run multiple bundles as necessary.

   • Embed this notice
     Ryan Castellucci :nonbinary_flag: (ryanc@infosec.exchange)'s status on Sunday, 26-Nov-2023 05:09:07 JST Ryan Castellucci :nonbinary_flag:

     in reply to

     • John Ripley

     @jripley Use multiple channels for stuff that has color control.