In about 1985 I was asked to design Switchmode halogen lamp transformers for
a company here. Running them on normal dimmers was a problem, so I made what
is now called a "reverse phase" dimmer for them using a power fet. There
where no PICs around then but it would have been an ideal combination. It
was a very simple circuit that had built-it short cct protection, also no
more expensive than a normal dimmer, and could be improved now with the use
of IGBT's.
    It improves EMI performance because you can put a cap (0.1uf) across the
Fet to slow down the rate of change of voltage without putting undue strain
on the switching device, with a triac the triac has to charge the cap almost
instantaneously with a corresponding high current.

George

----- Original Message -----
From: "Mike" <mike.ford@EARTHLINK.NET>
To: <PICLIST@MITVMA.MIT.EDU>
Sent: Saturday, November 29, 2003 8:14 AM
Subject: Re: [PIC:] 220V Lamps with PIC- best switch


> On 29 Nov 2003 at 16:16, Jinx wrote:
>
> Date sent:              Sat, 29 Nov 2003 16:16:00 +1300
> Send reply to:          pic microcontroller discussion list
> <PICLIST@MITVMA.MIT.EDU>
> From:                   Jinx <joecolquitt@CLEAR.NET.NZ>
> Subject:                Re: [PIC:] 220V Lamps with PIC- best switch
> To:                     PICLIST@MITVMA.MIT.EDU
>
> > > But actually, the Motorola Thyristor Device Data book discusses
> > > phase control using zero-crossing opto isolators. By switching
> > > complete half-cycles, EMI is eliminated whereas a PWM solution
> > > using those PVI optos would introduce EMI. My original comment:
>
> Actually, there's still going to be harmonic content in a half wave, but
not as much as
> in a square wave. Complete cycles would be better. If you skip every other
cycle, you
> effectively have a 50% DC. DC is usually used in the context of a square
wave, but in
> this case the square wave can be thought of as the on/off time or switch
cycle/block
> cycle much as duty cycle is used to characterize RF power amplifier duty.
When
> controlling AC phase, delay angle and conduction angle are the terms used.
>
> <snip>
> >
> > As for EMI ;
> >
> > As I understand it, phase control requires that the triac must be
> > turned on after the z-c point so as to maintain its conduction.
>
> But only by a little. The point of z-c switching is to do it at the z-c
point because as
> little as 10V across the TRIAC at turn on time into a load of a few 100 W
will cause
> EMI and you've just pissed in the wind.
>
> Once a TRIAC is on, the gate voltage can be removed and the TRIAC will
remain on
> until the current drops below I sub h, the holding current - maybe around
100uA.
>
> >  It
> > then turns off when current through it falls below a certain value
> > at the next z-c point. So you're using the trailing half of the cycle.
> > If this is true then is it possible to turn the triac on at the leading
> > z-c point (-> no EMI) and somehow force it to stop conducting at
> > some point in the cycle while there's V and I across it ? And if so,
> > is there any EMI generated ?
>
> 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.
>
>
> BRs,
> Mike
>
> --
> 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