NavList:

   

Reply

> The obvious question that comes to mind first is: "If the brain does not
have
> enough time to tell that the light is off, why does it have enough time to
tell
> it's on?"



Quite right...the answer is the threshold of quanta. As long as you can
cross it, then any added (within these paremeters) is just wasted. So the
answer is that once you get the system to see the light as "on", you can
shut it off for a very brief time and it will not notice the 'missing'
light...as long as the parameters are correct.



In other words, the argument could be reversed to explain why a
> flickering LED should look darker than a continuous one,



With the correct paremeters, flickering CAN make it look dimmer. But I don't
want a dimmer light.


BUT, if you turn it on for at least 100ms and then off for not more than
100ms (optimum), it will look almost as bright as if it were on the entire
time...provided the other paremetrs of min peak intensity ect are also met.
Then you can fool around with the different values to get different effects.



. At any rate, I am not
> convinced that the exploitation of this phenomenon makes any physical
sense.



Well, it has been used for quite some time...if you flick your eyes at leds,
you will see that many are indeed pulsed. In fact ON semi conductor as well
as several others mention it in their engineering application notes as a way
to conserve battery power...


>
> 1) The trick works only as long as there is only one light source, or as
long as
> all light sources are driven by the same driver.



Yes, to a large extent.



Otherwise the intensity of
> light would be averaged before it reaches the eye. This would seem to
limit the
> applicability.



It is true it is not as noticable in illumination applications...I designed
this driver mostly for visibility apps. However, it is also superior
electrically in many respects, so although it is not my fiirst choice for
large cabin lights (a DC-DC converter of very efficient design coupled with
a control IC is my first choice for illumination with larger lights), for
smaller bunks lights and task lights, it works very well, and for nav lights
it is ideal.



> 2) As you say yourself, the test person may well have the _impression_
that the
> light is brighter, but that does by no means imply that he sees better.



True...but there is some gain in these apps, just not as much.



inventing the perpetuum mobile.



The human eye is not liniar. So it is possible to take advantage of this a
bit and not violate any physics!



 So, what is the advantage
> of having a cabin light "LOOK" brighter? You won't be able to read better
in
> this kind of light.



One advantage of the driver (besides 'brightness enhancement', which as I've
already said is not the best app for this type of driver in illumination
uses) is 'loss-less' dimming so you can have a task light only as bright as
needed and not waste any power not needed at that time, and it also is a
very efficient regulator, as the other
thing I do with this driver as a regulating function is to trade off peak
intensity for average intensity, within the previous parameters, so that the
light stays almost at
the same intensity from 10 vdc to 14.5vdc. It does this without wastiing any
power, as it just adjusts pulse width as a function of voltage input. I
might mention that all the lights will shut off the leds when the voltage is
too high so as to protect them, and there is a tremendous amount of
overvoltage and transient protection built into the circuit...it can easily
handel anything it may be likely to encounter except a direct lightning
hit. -Ken

   

Reply