[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Dimmers



Original poster: Jim Lux <jimlux-at-earthlink-dot-net> 

Sort of...
They are phase control dimmers and work by chopping off the leading part of 
each half cycle.


Indeed, the output from them will have a peak voltage much the same as 
undimmed 120VAC (especially if you're using a meter that actually measures 
peak voltage and displays RMS with an internal scaling factor... meaning 
just about any meter other than a "true RMS"type), at least until you dim 
past the 90 degree phase mark.

The measured output (with a meter) of a phasecontrol dimmer will also vary 
a lot depending on what sort of load is attached, especially if the meter 
is a high input impedance (which most are).

The phase control circuitry (on an inexpensive dimmer) is done by a simple 
RC network. The dimmer is a two terminal device.  It works like this.. 
starting at the voltage/current zero crossing, the triac (i.e. switch) is 
off, so the full line voltage is across the dimmer.  As the sinewave starts 
up, the voltage increases. This voltage charges a small capacitor 
(typically around 0.1uF) through a fairly large resistor (so the current is 
small).  When the voltage on the cap gets high enough, the triac fires, and 
turns on (which essentially shorts the capacitor, by the way, but since 
there's a trivial amount of energy stored, it matters not).  The triac 
stays on until the current(!) goes to zero.

If you have a resistive load (light bulb), then the current goes to zero 
when the line voltage goes to zero, and everything starts over again for 
the next half cycle.

However, if you have an inductive load, it's a bit more complex.  First, 
let's look at the weirdness on turn on.  Assume everything is off. The line 
voltage starts to rise, but, because it's an inductor, the current flow 
lags a bit into the resistor capacitor combination in the dimmer.  It 
doesn't lag much, because, even though inductive, the impedance is pretty 
low compared to the resistance of the RC, however, it will turn on a bit 
"late".  When the C does charge up enough, the triac fires, and starts to 
conduct. Unfortunately, because there's an inductor in series, the triac 
doesn't necessarily get enough current through it to stay "on", and doesn't 
reliably stay on during the triggering process.  (this is why putting a 
resistive load in parallel helps.  When the triac fires, the resistive load 
provides enough current to keep it on while it's transitioning.)

Now, look at where stuff really gets screwed up.  The line voltage goes to 
zero (and normally, the triac would shut off).  However, the load is 
inductive, so the current keeps on flowing, as the voltage actually 
reverses across the dimmer.  This totally fouls up the phase control 
circuit made with a resistor and capacitor, so it's possible that a) the 
time constant elapses before the current stops flowing (so the triac never 
shut off) or b) some other weird interaction occurs because the voltage 
tending to charge the cap is of one polarity, while the current flow 
through the triac is of the opposite polarity. The triac is not just two 
SCRs back to back, and can trigger in any of the four quadrants.  There's 
also an issue that as the triac starts to turn off, the voltage rapidly 
starts to rise (because of the inductive "kick" from the load.. the 
inductor wants the current to keep going, and the voltage rises according 
to Ldi/dt).  If the voltage rises quickly enough, the triac will self 
trigger from dv/dt.

Industrial dimmers are made with two back to back scrs and trigger 
logic.  Essentially they are two complete half wave circuits in parallel, 
so, at least, you don't get the backwards trigger polarity problem or the 
dv/dt problem (you still have the dv/dt, but it's the "wrong" polarity to 
fire the SCR).


As a practical matter, though, the NST, when hooked up to a tesla coil (or 
a neon sign, for that matter) is not totally inductive.  It's a sort of odd 
hybrid with a dynamic impedance, depending on whether the spark gap fires, 
etc.  Therefore, it might work, and it might not.

ALso, bear in mind that theres a fairly significant resistive and inductive 
impedance in the output of the NST, so the voltage on the tank cap (which 
is what determines when the gap fires) rises somewhat behind the AC line 
voltage.  The gap might fire many degrees pat the peak of the AC line, and 
since the phase control dimmer works on AC line phase, there's another 
potential weird interaction.




At 05:27 PM 11/24/2003 -0700, you wrote:
>Original poster: "Jim Mitchell" <Wrenchead-at-starband-dot-net>
>Correct me if im wrong but dont triac dimmers just fire at varying rates to
>make somthing appear dimmer since the filament of a light bulb never does
>extinguish instantly, and that little resistor just changes the firing rate,
>right?  So if you put it on an NST which it probably wouldn't like much
>since its an inductive load, wouldn't you just get pulses out of the NST at
>the SAME VOLTAGE AND CURRENT?  As when I measure the output of a triac
>dimmer, it is always 120v, contrary to popular belife, these don't lower the
>line voltage and don't like inductive loads.
>
>Regards - Jim Mitchell
>----- Original Message -----
>From: "Tesla list" <tesla-at-pupman-dot-com>
>To: <tesla-at-pupman-dot-com>
>Sent: Monday, November 24, 2003 6:00 PM
>Subject: RE: Dimmers
>
>
> > Original poster: "Philip Brinkman" <peeceebee-at-mindspring-dot-com>
> >
> > I've been using an $8.00 lamp dimmer with attached cord on my Allanson
> > 15,000 volt 60ma
> > transformer for the past two months, It works great, it even has a built
>in
> > RF filter. It does get slightly warm when running for long periods... if I
> > have to buy a new one in a year or so..so what?. It has an attaced 6'
>cord,
> > on off switch, and lets me control the power from a safe distance. Plus it
> > has a plastic case and switches. Just plug and go!
> >
> >
> >  > [Original Message]
> >  > From: Tesla list <tesla-at-pupman-dot-com>
> >  > To: <tesla-at-pupman-dot-com>
> >  > Date: 11/23/2003 7:48:38 PM
> >  > Subject: Dimmers
> >  >
> >  > Original poster: davep <davep-at-quik-dot-com>
> >  >
> >  >
> >  > >I was about to reply with "No, a dimmer just won't work with an NST".
> >  > >This question has been asked often on this List and that's the usual
> >  > >response.
> >  >
> >  >          ...and its mostly right.
> >  >
> >  > >  But since I happened to have a dimmer and had never
> >  > >personally tried it, I thought I would.
> >  >
> >  > >I was surprised to find that it did work - mostly.
> >  >
> >  >          ...and for how long.
> >  >          May things will work for briefish periods of time that will
> >  >          work a lot less well for extended periods.
> >  >
> >  >          and each design of dimmers varies...
> >  >
> >  >          best
> >  >          dwp
> >  >
> >  >
> >
> >
> >