On 29 Nov 2003 at 22:09, Jinx wrote:


> > Yes. Any fast edge will generate harmonics. Switching full cycles
> > reduces or eliminates EMI. Using the trailing half of the cycle was
> > practical because a DIAC (now we have SIDACs) could be used
> > to delay triggering and I know of no simple way of doing otherwise
> > and there's no point to it anyway.
>
> So if I understand correctly, if there's sufficient energy present to
> cause RFI at the switching point, it's irrelevant whether the triac
> is turned off (if that were possible) or on. The practical difference
> being that the triac "turns itself off" at the next z-c point, which
> keeps switching and s/w simpler

right, it turns off provided there's no gate signal to turn it back on.
>
> From what I can gather, series inductors, shielding and good layout
> are fairly effective at minimising RFI

Yup. Since you mention layout, minimize the area of your current loop.

Those EMI filters with the built in IEC connector might help.
>
> Just an academic question. Is it not possible using s/w and another
> device to rob the triac of through-current - at the same time both
> disconnecting the load and dumping the current for the remainder
> of that half-cycle into something like an inductor ?

I'm sure it is and I'm not going to try to picture what I'm not even sure =
you're getting
at, but to what effect? Dump current to an inductor and you're now driving=
 an
inductive load. Rob the TRIAC of through current and there's no current to=
 dump.
WTF?
>
> I ask because it would be kinder to all the devices if the switching
> on were to be done at low voltage rather than eg at peak to reduce
> stresses caused by inrush current.

Zero-cross switching solves that.

> To keep it simple, it seems that
> something other than a triac is needed for PWM if you wish to use
> the leading z-c and turn the juice off mid-wave

Never saw a TRIAC used for PWM. Probably since they're used for AC switchi=
ng.
When I talked about PWM, I was talking about MOSFETs. I probably spewed a =
few
different thoughts together. You can use the z-c opto to switch or not swi=
tch full
cycles of AC. It has to be done with a bilateral analog switch. That's 2 M=
OSFETs and
google is your friend. A single triac won't switch AC.

View in Courier font
                             Vac
                              +
                              |
                              |
                              |
                              |
           +-----------+   ||-+
    |      |           |   ||<-+
    V ~>  ---          +---||-+|
    - ~>   -           |      |
    |      |           |      |
           +------------------+
                       |      |
     PVI opto          |      |
                       +---||-+|
                           ||<-+
                           ||-+
                              |
                              |
                              |
                             .-.
                            (=3DX=3D) Light
                             '-'
                              |
                              |
                             =3D=3D=3D
                             GND

 Use z-c opto for full cycle P control or rig your own z-c detector to the=
 PVI trigger

created by Andy=B4s ASCII-Circuit v1.22.310103 Beta www.tech-chat.de

>
> --
> http://www.piclist.com hint: The list server can filter out subtopics
> (like ads or off topics) for you. See http://www.piclist.com/#topics

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics