Hi, Matt.

 On the pulse train topic first things first. Quoting National
Semiconductor COP, Note 6: "It is possible to trigger a triac with a stea=
dy
state logic level. This is accomplished by allowing the triac gate to sin=
k
or source current during the desired on-time. When utilizing this method =
it
becomes easier to trigger the triac and leave it on for many cycles witho=
ut
having to execute code to retrigger. This approach is advantageous when t=
he
triac must be fired is for relatively long periods and conduction angle
firing is not desired, thus more time is available to accomplish auxiliar=
y
tasks."  This is something you may wish to do when you are implementing
ON/OFF control on the triac, i.e. M cycles ON , N cycles OFF. This is not
the usual case for lamp dimming where the angle firing approach is the on=
e
to go to achieve the goal of controlling the mean voltage to the lamp.=20
This leads us to other thing: In lots of power converters de period of
conduction of the thyristor depends on the power factor of the load,
meaning that the begining of the of the conduction might not be well
defined. This means that its usually needed to trigger the triac in a
continuos way but this raises the loses on the thyristor. For pure
resistive loads this is not an issue, but =BFwhere do you find a pure
resistive load?. I, as a rule of thumb, prefer to use a  train pulse to
avoid this problems. What I do is to make a logical operation AND between
the firing pulses from the uC with a 20 KHz square wave and voila! For so=
me
applications this might be overkill but if you begin to find weird proble=
ms
this could help a lot.

Hope I did=B4n=B4t confuse more!

Regards,=20

*Carlos Marcano*
-Guri, Venezuela-

P.S: Sorry for my english!

_______________________________________________
http://www.piclist.com
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist