Re: Making an X-10 lamp module immune to dimming

[isw <isw@xxxxxxxxxxx>]

In article <MBK1k.37422$SV4.18879@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
 "Jeff Volp" <JeffVolp@xxxxxxx> wrote:

> Your are correct about transposing pins 6 & 8.  I caught that myself in an
> earlier post.
>
> I was thinking along the lines of a diode and cap too.  I have not looked at
> that signal myself.  However, if there is always a pulse there, a simple
> circuit may not be sufficient.  A pulse late in the cycle fed through a
> diode - capacitor network into the transistor would trigger the triac early
> in the next cycle.

Not if you get the time constant right.

> Or, does that pulse go away completely when the module is OFF?

According to my notes (several years old), there is always at least one
pulse there. The diode and capacitor, plus changes to the values of the
resistors, will integrate the pulse train into a single pulse that gets
wider as the dimming moves towards full on (the first pulse in the train
is always at the time the triac *should* fire; maybe all the others are
just "insurance" or maybe it's just the way that chip works. When the
unit is at "full off" that single pulse is very narrow, so an RC
integrator works pretty well (at least, it worked for me).

--
Side note

I took the optically isolated integrated pulse, and integrated it again
to a DC level, then fed it to a National LM-3914 Dot/Bar Display Driver.
I was looking for a cheap way to control a lot of off-on devices from a
single X-10 unit. I figured that by counting "brighten" pulses, I could
make whichever of the 3914 outputs hot that I wanted. Problem was, the
X-10 unit was not consistent from one time to the next, so everything
worked just fine except that say, 5 pulses, would not always make the
same output hot; sometimes the one "before" and sometimes the one
"after". Bummer.

--

So based on all that, I think that it is possible to get a single pulse
that gets wider as "brighten" commands are sent, even though there will
be some inaccuracy in just how wide that pulse is.

Then if you set up some sort of "thresholding" (like a Schmitt trigger
or an overdriven high-gain stage), the triac will either get a drive
signal, or it will not.

Isaac