[Public-List] LED Bulbs
Kris Coward
kris at melon.org
Fri Mar 29 12:39:23 PDT 2019
Well the LEDs in the bulbs are themselves low-voltage dc components
(ideal voltage drop across a white LED tends to be a little under 4V),
so running a nominally 120VAC bulb on 12VDC really depends on how those
120VAC are converted to the 4VDC that each LED in the bulb actually
uses.
Most AC->DC converters will gladly take DC input and just pop it right
out the other side as only-slightly-lower-voltage DC. Many of them will
produce considerably larger drops on AC input. If the circuit designer
knows that the available input power is higher than the needed output
power by a large enough margin, they can use a (cheaper) half-wave
rectifier that basically only draws power when the polarity of the AC
wave matches the desired DC polarity (instead of flipping the other half
of the wave to draw power all the time). For an LED bulb drawing its
power from a 120VAC circuit, the margin is definitely there, making it
quite likely that a half-wave rectifier is being used.
Now when the power is coming out of the rectifier, the voltage is still
quite variable. This is normally smoothed out with capacitors, but it
can't be smoothed out all that much without starting to use rather large
capacitors.
As luck would have it however, the best brightness/longevity of an LED
is achieved at a constant supply current, rather than a constant supply
voltage, so the output of the rectifier is very likely going into a
constant-current driver of appropriate size to run the LEDs themselves.
Most such drivers accept a pretty wide range of input voltages as it is,
so it's not even remotely unreasonable to imagine the bulb incorporating
a driver that takes an input voltage range of, say, 24-108V.
In such a circumstance, it's then just a question of how the driver
responds to being underdriven (e.g. if it expects 24-108V input, but
only gets 12V), but most such devices, when underdriven, just yield less
output (possibly none at all), instead of failing (catastrophically).
It's probably worth remembering that your light bulbs should not fail in
a way that burns your house down when there's a brownout (and I'm
guessing that any bulbs that would fail catastrophically when
underdriven wouldn't get UL, CE, etc. marks). Similarly, an underdriven
LED is likeliest to just put out less light than one that's fully driven
(assuming it's getting fewer amps at the same voltage, rather than fewer
volts -- the latter would likely see it failing to light up at all),
which seems to have been the effect that you observed.
If it's providing enough light, it's almost certainly not going to burn
your boat down or anything -- though you might want to swap some
incandescents back in when you have a survey done ;)
-K
On Fri, Mar 29, 2019 at 09:28:42AM -0400, Mike Lehman via Public-List wrote:
> We have several light fixtures on the boat that take standard base bulbs
> (like those at home). Naturally on the boat, the voltage is 12V while at
> home it is 110V. I plan to change out the bulbs on the boat to LED (have
> already made this change to all of the bulbs at home). So as an experiment,
> I took one of the LED (Home) bulbs, that was purchased at IKEA, and tried
> it on the boat not expecting it to work...but it did! Not quite as bright
> as it is when connected to a 110 lamp, but bright enough to read by. I was
> very surprised.
>
> Can anyone shed some light (pun intended) on this topic? Why did this work?
> is it safe? It sure as hell is a lot cheaper than 'Marine' 12v bulbs.
>
>
> --
> Mike Lehman
> ~~~_/)_/)~~_/)~~~
> _______________________________________________
> These businesses support your Association:
> http://www.alberg30.org/store/A30supporters.html
> Please support them.
> _______________________________________________
> Public-List mailing list
> Public-List at lists.alberg30.org
> http://lists.alberg30.org/listinfo.cgi/public-list-alberg30.org
--
Kris Coward http://unripe.melon.org/
GPG Fingerprint: 2BF3 957D 310A FEEC 4733 830E 21A4 05C7 1FEB 12B3
More information about the Public-List
mailing list