Light bulb power control/dimmer

[TT_Man]

Will using cycle dropping techniques work over a practical range, say 20% to
100%? Zero volt switching/+ phase angle is the norm I guess......

 

[John Fields]

Will using cycle dropping techniques work over a practical range, say 20% to
100%?

 

[Phil Allison]

"TT_Man"

Will using cycle dropping techniques work over a practical range, say 20%
to 100%? Zero volt switching/+ phase angle is the norm I guess......

** Cycle dropping causes a lamp to visibly flicker.

If that don't matter to you then go ahead.



........ Phil

 

[Joerg]

Will using cycle dropping techniques work over a practical range, say 20% to
100%? Zero volt switching/+ phase angle is the norm I guess......

John and Phil mentioned flicker. This is not only annoying but with some
bad luck it could make part of the filament resonate and then ... POOF.

 

[Tam/WB2TT]

John and Phil mentioned flicker. This is not only annoying but with some
bad luck it could make part of the filament resonate and then ... POOF.

I believe it can also cause epileptic seizures in some people.

Tam

 

[Joerg]

I believe it can also cause epileptic seizures in some people.

That would be really bad. Also, cycle skipping is very frowned upon by
utilities.

 

[TT_Man]

That would be really bad. Also, cycle skipping is very frowned upon by
utilities.

Thanks for the answers guys. I was wondering about the flicker and if it
would be a real issue.So it's in the bin for that idea and saved me a lot of
wasted time writing code.... on with the traditional solution.

 

[Tam/WB2TT]

Thanks for the answers guys. I was wondering about the flicker and if it
would be a real issue.So it's in the bin for that idea and saved me a lot
of wasted time writing code.... on with the traditional solution.

Remember, with cycle skipping the frequency changes. At 10 percent power the
fequency would be 6 Hertz.

Tam

 

[LVMarc]

Will using cycle dropping techniques work over a practical range, say 20% to
100%? Zero volt switching/+ phase angle is the norm I guess......

yes you can drop cycles or drop intra-cycle pieces parts dropping ..;-)
the change in perceived or actual power is not proportionate to
absolute phase angle but is more porportinal to number of complete
cycles "dropped" per 10 cycles.. any more and would see flicka..PLus
you could use traic (or two scrs) and double the cycles you have (per
0.142 seconds..flicka period)for control by cycle dropping..

BTW, this zero volt switching is needed where the load would be
sensitive to EMI like signal on the power. as you chopped intra-cycle
the harmonic content increases, fundamental decreases.. the system has
to handle and "filter" chopped AC... For a light bulb load, this is NOT
th case, as the electric power is turned into heat and that is not noise
or harmonic sensitive.. try it and see!


Happy designing

Marc

 

[Nobody]

Remember, with cycle skipping the frequency changes. At 10 percent power the
fequency would be 6 Hertz.

The flicker will be 12 Hertz (as opposed to 120Hz at full power). You
don't have to skip whole cycles; half-cycles will do. The polarity doesn't
matter for a resistive load.

 

[Tam/WB2TT]

The flicker will be 12 Hertz (as opposed to 120Hz at full power). You
don't have to skip whole cycles; half-cycles will do. The polarity doesn't
matter for a resistive load.

But you can't see the 120 Hz component; at least I can't - for an
incandescent lamp.

Tam

 

[Phil Allison]

"Nobody"

The flicker will be 12 Hertz (as opposed to 120Hz at full power). You
don't have to skip whole cycles; half-cycles will do. The polarity doesn't
matter for a resistive load.

** Skipping half cycles means you are loading the supply for half cycles -
this creates a DC component on the supply.

Very bad idea.



........ Phil

 

[Don Klipstein]

The flicker will be 12 Hertz (as opposed to 120Hz at full power). You
don't have to skip whole cycles; half-cycles will do. The polarity doesn't
matter for a resistive load.

In extreme cases, I have seen incandescents visibly flicker from
skipping every other half cycle. The main example is a 4 watt or 7 watt
120V incandescent nightlight with a diode. The thin filament warms and
cools fast enough to have significant 60 Hz flicker. 60 Hz flicker is
sometimes visible.

- Don Klipstein ([email protected])

 

[Spehro Pefhany]

On Sun, 18 Nov 2007 17:16:22 +0000 (UTC), the renowned

In extreme cases, I have seen incandescents visibly flicker from
skipping every other half cycle. The main example is a 4 watt or 7 watt
120V incandescent nightlight with a diode. The thin filament warms and
cools fast enough to have significant 60 Hz flicker. 60 Hz flicker is
sometimes visible.

- Don Klipstein ([email protected])

Yes, you can't practically dim an incandescent bulb by skipping
cycles. I tried it many years ago and the results were miserable.
Maybe if you had control over the bulb design and could add some
thermal mass to the filament, but I tried it with a huge incandescent
(kW, IIRC) and even that thick filament responded too fast. I used
some kind of clever (I thought) CMOS circuit (I forget the details)
using something like a rate multiplier to get a relatively fast cycle
time (eg. 50% would be a 30Hz cycle). OTOH, it worked a charm on
relatively fast-response heaters for tight temperature control, which
was the intended purpose.


Best regards,
Spehro Pefhany

 

[LVMarc]

On Sun, 18 Nov 2007 17:16:22 +0000 (UTC), the renowned



Yes, you can't practically dim an incandescent bulb by skipping
cycles. I tried it many years ago and the results were miserable.
Maybe if you had control over the bulb design and could add some
thermal mass to the filament, but I tried it with a huge incandescent
(kW, IIRC) and even that thick filament responded too fast. I used
some kind of clever (I thought) CMOS circuit (I forget the details)
using something like a rate multiplier to get a relatively fast cycle
time (eg. 50% would be a 30Hz cycle). OTOH, it worked a charm on
relatively fast-response heaters for tight temperature control, which
was the intended purpose.


Best regards,
Spehro Pefhany

The persistence of vision, the flicker frequencies for most folks is
about 8hz, so when the proportion f 1/2 cycle arpoaches that some folks
will begin to see it, and other will not, as you eat more cycles of
course the drive is more choppy and you cant average out the photons as
the other poster noted

But, you can conceptually make this dimmer and thats what the poster
wanted. The traditional and widely adopted intra-cycle modulation is
used everywhere.... Perhaps there is a space where you want one, maybe
one that can do both.. not sure?

OTOH, one poster suggested that , using 1/2 cycle non symmetry would
result in a net DC component and that it was bad. Not sure what would
dbe bad on a light bulb load, with a slight, very tiny mathemical sized
DC component.. PLease advise what sort of effects you were envisioning??

best regards,

marc Popek

 

[Jim Thompson]

The persistence of vision, the flicker frequencies for most folks is
about 8hz, so when the proportion f 1/2 cycle arpoaches that some folks
will begin to see it, and other will not, as you eat more cycles of
course the drive is more choppy and you cant average out the photons as
the other poster noted

But, you can conceptually make this dimmer and thats what the poster
wanted. The traditional and widely adopted intra-cycle modulation is
used everywhere.... Perhaps there is a space where you want one, maybe
one that can do both.. not sure?

OTOH, one poster suggested that , using 1/2 cycle non symmetry would
result in a net DC component and that it was bad. Not sure what would
dbe bad on a light bulb load, with a slight, very tiny mathemical sized
DC component.. PLease advise what sort of effects you were envisioning??

best regards,

marc Popek

I dimmed neon in a disco by skipping WHOLE cycles going into the HV
transformer (~1980).

...Jim Thompson

 

[John Larkin]

Yes, you can't practically dim an incandescent bulb by skipping
cycles. I tried it many years ago and the results were miserable.
Maybe if you had control over the bulb design and could add some
thermal mass to the filament, but I tried it with a huge incandescent
(kW, IIRC) and even that thick filament responded too fast. I used
some kind of clever (I thought) CMOS circuit (I forget the details)
using something like a rate multiplier to get a relatively fast cycle
time (eg. 50% would be a 30Hz cycle). OTOH, it worked a charm on
relatively fast-response heaters for tight temperature control, which
was the intended purpose.


Best regards,
Spehro Pefhany

You can send decent optical-link audio by modulating the current into
a regular light bulb or a flashlight bulb. I've heard that an auto
headlight works, too, but I haven't tried that one.

Hmmm, one could do a kind of integral-AC-cycle delta-sigma thing.

John

 

[Phil Allison]

"LVMarc"

The persistence of vision, the flicker frequencies for most folks is about
8hz,

** Complete bollocks.

50 Hz flicker is visible to most folk.

OTOH, one poster suggested that , using 1/2 cycle non symmetry would
result in a net DC component and that it was bad. Not sure what would dbe
bad on a light bulb load, with a slight, very tiny mathemical sized DC
component.. PLease advise what sort of effects you were envisioning??

** Go and find what I actually posted - you PITA jerk off.



....... Phil

 

[Don Klipstein]

The persistence of vision, the flicker frequencies for most folks is
about 8hz, so when the proportion f 1/2 cycle arpoaches that some folks
will begin to see it, and other will not, as you eat more cycles of
course the drive is more choppy and you cant average out the photons as
the other poster noted

I remember an electrical engineering lab day in college many years back.
One thing to be tried was to adjust the horizontal sweep rate of an
oscilloscope to the point where with no vertical signal the trace barely
did not visibly flicker. Maybe sweep was triggered externally, because
this was to determine sweep repetition rate for the trace to borderline
visibly flicker and not flciker.

Most of my classmates got frequencies in the 40's and low 50's of Hz for
this.

- Don Klipstein ([email protected])

 

[Michael A. Terrell]

I remember an electrical engineering lab day in college many years back.
One thing to be tried was to adjust the horizontal sweep rate of an
oscilloscope to the point where with no vertical signal the trace barely
did not visibly flicker. Maybe sweep was triggered externally, because
this was to determine sweep repetition rate for the trace to borderline
visibly flicker and not flciker.

Most of my classmates got frequencies in the 40's and low 50's of Hz for
this.

That experiment would be affected by the choice of phosphor on the
scope's CRT. I've had some with very fast phosphors, and some that took
over 10 seconds to fade.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida