>
> In this case, the speed of the Triac is dwarfed by the speed
> of the lamp
> itself.
>
> Incandescent lamps take a non-trivial time to warm up enough
> to give off
> light.  How long they take depends on the filament design and bulb.
> This is one of the reasons that phase control dimming works,
> because you
> can get rid of half of each half wave of AC and not notice any
> flickering - it takes too long for the filament to cool down (and warm
> up) to lower the light level and bring it back up noticably.
>
> This is also why there's little reason to transmit light brightness
> updates more than 40 times per second (as is done in DMX-512).

A good point with typical lighting - also with industrial processes.
I did a system a few years ago for large commercial offset printing presses
that irradiates each sheet up to 78" wide as it is printed at 10,000 per
hour or faster. Even though that's only about 3 sheets per second I had to
both adjust lamp output and measure sheet temp and at 100 ms intervals. The
rationale for this control rate was that each single sheet was worth several
dollars to the customer. The irradiators have very low-density filaments,
designed specifically for fast response. With warm emitters I could actually
measure differences mid-sheet at lower control cycle frequencies, but
somewhere near 20/sec there was no perceivable difference. I believe this
was in part due to the fact that at that rate the change in output levels
was very small. At some point any digital process begins to appear
linear....

After all that, I'm still working on a PID algorithm that will keep up - so,
why bother to adjust any faster for even another reason.

On the other hand - think of the 7:1 or better inrush current while that
lamp is heating up from a cold start. Thyristor soft-starts make good sense
here.

C



>
> So, depending on the incandescent bulb, IMO, it could
> certianly take an
> incandescent bulb 40mS to achieve 80% brightness, regardless
> of whether
> you used a triac or any other switching device to turn it on.  Please
> note that I haven't measured this.  You can see the difference by
> looking at a car which has LED brake lights versus one with regular
> bulbs.  You can perceive that the LEDs appear instantaneously, whereas
> the bulbs definitely have a warm up time.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.