Maker Pro
Maker Pro

Fluorescents and migraines??

K

krw

Jan 1, 1970
0
[email protected]>,
alt.engineering.electrical, [email protected] says...
SNIP


I have used CFLs for outdoor lantern fixtures for the last 10 years,
in a moderately cold climate. 10F min. and I'm on only my third set of
lamps.

Try them at -20F. No thanks.
(Earlier poster was boasting of experience with a dozen lamps. One of
my clients had 900 CFL retrofits for 5 years. That's expericence!)

They are on a timer so 30 sec to warm up is a non-event. Color is
85CRI 3000K but is slightly different than the neighbors incandescent
and MH.

I save even more electricity than using CFLs. I only light an area
where someone is. 30sec (don't believe it) warm-up is
unacceptable.
If it's all about cost CFL is the only way to go!

No, if it's all about cost leave the damned light off. There are
things in this life more important than a few cents.
 
V

Victor Roberts

Jan 1, 1970
0
Victor Roberts wrote:
[snip]
I think the confusion arises because people assume that
there is no modulation of high frequency output. These are
people who don't understand how an electronic ballast really
works. They have never thought about the fact that the high
frequency oscillator is powered by an imperfect DC power
supply.

and also in another post:
You are assuming an ALL or NOTHING situation. All CFL
ballasts I have seen have a DC storage capacitor and
therefore smooth the DC link voltage to some extent. They
just do not have a large enough capacitor to completely
eliminate the 120 Hz ripple. In fact, many CFL ballasts
have about 50% ripple on the DC link.

Let me see if I can use the above to demonstrate mathematically that there
exists at least one kind of CFL which does not flicker.

Count the original oscillator in both the above cases as ONE, and the video FPS
rate as TWO. Vic, above, effectively says that we always have _some_ ripple even
in the best possible cases and hence some light flicker. Therefore, when we
video any (even so slightly) flickering light source, we effectively have two
coupled oscillators, so the effects can be analyzed mathematically.

It turns out that the slower oscillator (which in this case is the video)
"captures" the behavior of the fastest one:

http://ioannis.virtualcomposer2000.com/math/video.html

For those who wish to forgo with the math, the above simply means that:

"If there's light oscillation, then it shows on video. Hence, if nothing shows
on video, there's no light oscillation".

If, without loss of generality we try the above analysis for some ridiculously
high value of light oscillation, say 67361 Hz or 67.3 kHz, the resulting
Diophantine equation gives:

k=33680+67361*n
j=40+80*n.

Let's pick j, which is easier. The first "flickering" ripple on the video should
occur at 2*(40+80)/80 secs = 3 secs. The second flickering ripple on the video
should occur at 2*(40+80*2)/80=5 secs. This means that if the light was
oscillating at 67.361kHz, the video would have shown flickering ripples with a
time amplitude of 2 seconds.

Since the video of the CFL is much longer than 2-3 seconds, it follows that a
high light oscillating frequency at least up to 67.361 kHz is not likely.

Adjust the equation per your preferences, adding mayo and jalapeno peppers. I
suppose I could solve the problem backwards and try to find what is the maximum
oscillation frequency whose video ripple amplitude is larger than a minute worth
of video, but it's a little late now and I need to pass to the other dimension
for some rest.

(Sorry about having the videos on QuickTime format, but that's my camera's
capture mode and as I said, I am a little tired right now. For those who don't
know about it, QuickTime is a free download from Apple).

Anyone who disagrees with the above conclusion, please raise your hand... ;o)

I raise my hand in objection.

Instead of your complicated arrangement I just connect an
optical detector to my oscilloscope and point it at a CFL.
If the trace on the oscilloscope shows modulation, then the
CFL light output is modulated.

I've attached a trace of a CFL with small amount of 120
modulation of the light output. The zero level is about 1
division from the bottom of the screen, at the arrow marker.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.
 
V

Victor Roberts

Jan 1, 1970
0
Victor Roberts wrote:
[snip]
I think the confusion arises because people assume that
there is no modulation of high frequency output. These are
people who don't understand how an electronic ballast really
works. They have never thought about the fact that the high
frequency oscillator is powered by an imperfect DC power
supply.

and also in another post:
You are assuming an ALL or NOTHING situation. All CFL
ballasts I have seen have a DC storage capacitor and
therefore smooth the DC link voltage to some extent. They
just do not have a large enough capacitor to completely
eliminate the 120 Hz ripple. In fact, many CFL ballasts
have about 50% ripple on the DC link.

Let me see if I can use the above to demonstrate mathematically that there
exists at least one kind of CFL which does not flicker.

Count the original oscillator in both the above cases as ONE, and the video FPS
rate as TWO. Vic, above, effectively says that we always have _some_ ripple even
in the best possible cases and hence some light flicker. Therefore, when we
video any (even so slightly) flickering light source, we effectively have two
coupled oscillators, so the effects can be analyzed mathematically.

It turns out that the slower oscillator (which in this case is the video)
"captures" the behavior of the fastest one:

http://ioannis.virtualcomposer2000.com/math/video.html

For those who wish to forgo with the math, the above simply means that:

"If there's light oscillation, then it shows on video. Hence, if nothing shows
on video, there's no light oscillation".

If, without loss of generality we try the above analysis for some ridiculously
high value of light oscillation, say 67361 Hz or 67.3 kHz, the resulting
Diophantine equation gives:

k=33680+67361*n
j=40+80*n.

Let's pick j, which is easier. The first "flickering" ripple on the video should
occur at 2*(40+80)/80 secs = 3 secs. The second flickering ripple on the video
should occur at 2*(40+80*2)/80=5 secs. This means that if the light was
oscillating at 67.361kHz, the video would have shown flickering ripples with a
time amplitude of 2 seconds.

Since the video of the CFL is much longer than 2-3 seconds, it follows that a
high light oscillating frequency at least up to 67.361 kHz is not likely.

Adjust the equation per your preferences, adding mayo and jalapeno peppers. I
suppose I could solve the problem backwards and try to find what is the maximum
oscillation frequency whose video ripple amplitude is larger than a minute worth
of video, but it's a little late now and I need to pass to the other dimension
for some rest.

(Sorry about having the videos on QuickTime format, but that's my camera's
capture mode and as I said, I am a little tired right now. For those who don't
know about it, QuickTime is a free download from Apple).

Anyone who disagrees with the above conclusion, please raise your hand... ;o)

OK - I'll raise my hand.

Instead of using your approach I just connect an optical
detector with sufficiently short response time to my
oscilloscope and point it at the CFL. If the trace is nor
flat then the light output is modulated.

See
http://www.robertsresearchinc.com/Papers/CFL_Modulation.BMP
for an oscilloscope trace of the light output of a CFL with
a small amount of modulation.

The zero level is a bit below the first division on the
screen, where the small arrow pointer is positioned. The
average output is about 40 mV and the peak-to-peak ripple is
about 7 to 8 mV.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.
 
V

Victor Roberts

Jan 1, 1970
0
I raise my hand in objection.

Instead of your complicated arrangement I just connect an
optical detector to my oscilloscope and point it at a CFL.
If the trace on the oscilloscope shows modulation, then the
CFL light output is modulated.

I've attached a trace of a CFL with small amount of 120
modulation of the light output. The zero level is about 1
division from the bottom of the screen, at the arrow marker.

The attachment failed. I posted another message with a link
to a BMP of the scope trace.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.
 
| [email protected] wrote:
| [snip]
|
|> Don't believe what you read in Wikipedia, unless they get lucky and
|> have something correct (it happens often, but not in this case).
|>
|> The purpose of the ballast is...
|
| [snip ballast explanation for brevity]
|
| Do I look to you like I need a tutorial on what is a ballast? If that's the
| impression I gave you, either my exposition powers are weak or your
| comprehension abilities are not up to par.

I've looked at lots of people, and I've never found any real easy means to
determine from their appearance, mannerisms, hygiene, or other aspects,
whether then need a tutorial on what a ballast is, or not. That, and I
didn't even look at you at all, since this is the Internet. So it's just
a crapshoot. But you were looking in Wikipedia for information. That is
enough to make me worry.


|>> and later down:
|>>
|>> "Electronic ballasts do not produce light flicker, since the phosphor
|>> persistence is longer than a half cycle of the higher operation
|>> frequency.
|>
|> They do not _produce_ it. They may let it pass through by not
|> storing any energy to "cover" the zero-crossover time period.
|>
|> Magnetic ballasts do not _produce_ flicker either.
|
| Wikipedia is not an engineering manual. For the lay person their explanation is
| correct. There is no point in arguing insignificant minutae with me (or with
| Wikipedia). /Of course/ ballasts do not "produce" flicker (literally), since
| they are not the main power supply which drives the lamp. But from a
| non-technical standpoint, it's the inductive resistance of the ballast which
| allows the AC cycle to propagate FROM the AC source to the lamp almost
| unchanged, making it seem as "flicker", for whatever reason, whether it be
| insufficient attenuation of the AC signal, bad power factor, "flattening" of the
| AC signal, lack of capacitors or whatever have you.

The resistance in the inductor has nothing to do with it. Maybe you meant to
say "impedance" instead of "resistance".

Using the correct terms can, in quite many cases, be very crucial. If you want
to be an engineer (even if you just want to play an arm-chair engineer on the
Internet), learn to use the correct terms. I had to.

If you want to simplify things, or even over-simplify things, that's one
thing. But when things just get wrong, that gives me cause for concern.
I've had to deal with (in my field, computer software) people who too often
take something that was a simplification as being something of detail and
ended up with an entirely wrong understanding. I not saying you have
misunderstood these technical things; I'm concerned more about what others
might read from Wikipedia or your quotations.


| What really matters here is what Wiki says later, which you conveniently did not
| address:
|
| "Electronic ballasts do not produce light flicker, since the phosphor
| persistence is longer than a half cycle of the higher operation frequency."
|
| As far as I am concerned, THAT's the crucial point which proves there's no
| flicker.

Still, that is incorrect.

That statement from Wikipedia suggests that it is the persistence of the
phosphor that makes the electronic ballast not produce flicker. That is
not just wrong, it's even silly.

Phosphor persistence can help reduce flicker. In very extreme cases it could
even eliminate it, in theory (but expect the light to continue to glow for a
very long time after you turn it off). This effect would be the same whether
the lamp current was being limited by an inductive ballast, capacitive ballast,
resistive ballast (no one would use such a beast, but one could be made), or an
electronic solid-state ballast.

Try this statement on for size:

"Magnetic ballasts do not produce light flicker, since the phosphor
persistence is longer than a half cycle of the higher operation frequency."

Is that any more or less "correct" in the context of Wikipedian engineering
than what you quoted from there?


| Of course, if you say that you see flicker, there's no way for me to convince
| you otherwise. If I claim that yesterday I saw a grand pink elephant nest with
| green eggs and ham sitting at the center of a primordial black hole, there's no
| way for you to prove me wrong either.

There are a lot of things I see in a lot of aspects of the world that a lot of
people try to convince me is not really there. THEY don't see it, so as far
as they are concerned, it really isn't there. I hope you are not slipping into
that category.

I wouldn't even try to prove you wrong on your sighting, even though I am
quite certain that black holes have no center.


|>> The non-visible 100?120 Hz flicker from fluorescent tubes powered by
|>> magnetic ballasts is associated with headaches and eyestrain."
|>
|> Some people _can_ see it. Some people need to roll their eyes to see
|> that it is there. Some people can just see it directly. It seems
|> most people cannot see it either way.
|
| I am not talking about whether one can see it by rolling one's eyes back and
| forth. I can see flickering even on incandescent sources if I roll my eyes back
| and forth. The question is whether a lay person can perceive consciously 100-120
| Hz flicker without doing a circus act with one's eyeballs. THAT's the question.

And I can see the flicker directly. I just can't see very accurately how MUCH
there is, or how much of it is compensated by the phosphors. When I do scan
my eyes across, I can see things that give me more information. If the light
is literally on and off that tells me one thing. If the light descends into a
different color, that tells me another (phosphor persistence is usually variant
in color in fluorescent lighting ... the totally color you get is the average
over time).


| It's a question of whether the perceptual system "eye-brain" has the capacity to
| perceive this flicker on standing mode ON A CONSCIOUS LEVEL and whether this
| flicker can cause headaches.

I can see the flicker from most fluorescent lights even when only viewing the
reflection of it from broad surfaces.

However, I have found that this does NOT cause headaches for me. I cannot say
if it does or does not cause it for others. I used to THINK that the flicker
was the cause, basically because it had been suggested for decades.


| I am ready to agree that although the flicker itself may not be /visible/ on a
| CONSCIOUS (PERCEPTUAL) LEVEL, the brain may be able to pick it up
| subconsciously. That's a contention I am ready to argue about, as a potential
| source of migranes. The rest is irrelevant.

I certainly cannot just count the 120 pulses per second. I can see that it is
flickering, but I cannot see individual pulses happening. I do not see it as
going on and off. I see the sense of flicker. I see it in some LEDs but not
in others. The ones that are battery powered don't have the flicker. Yet they
can cause the headaches.


|>> How come nobody had headaches back then?
|>>
|>> Or did they?
|>
|> I did! I just misunderstood exactly why. Back then I thought it was
|> _because_ of the flicker. Now I understand it is because of the
|> spectrum.
|
| Huh? How can you be sure without knowing the EXACT cause of what bothers you in
| the spectrum?
|
| Or if you DO know the exact cause, what is it that bothers you in the spectrum?

The spectrum is not continuous. It has a large gap or two large gaps in it.
My eyes do not focus all colors at an equal distance. Glasses exacerbate
that problem (more so at the edges of the glass). As a result, the edges
where light is different, such as the edge of black text on a white page,
is not in perfect focus. With a smooth continuous spectrum, it will appear
to be slightly fuzzy, but tolerable. With a broken spectrum, there will
appear to be 2 or more distinct edges. In the latter case, my eyes are
constantly jumping back and forth trying to focus in one color or the other.
That constant refocusing creates stress, and eventually a headache. This
is what appears to be going on for me. I do not know if others have the
this kind of issue or not. I do not know if they can get headaches from
other things that don't affect me. I've learned that people are sufficiently
different to never make such assumptions (although I've met many people that
have not learned that for themselves).
 
| You are assuming an ALL or NOTHING situation. All CFL
| ballasts I have seen have a DC storage capacitor and
| therefore smooth the DC link voltage to some extent. They
| just do not have a large enough capacitor to completely
| eliminate the 120 Hz ripple. In fact, many CFL ballasts
| have about 50% ripple on the DC link.

That must be what I am seeing, then.
|
|
|>| and later down:
|>|
|>| "Electronic ballasts do not produce light flicker, since the phosphor
|>| persistence is longer than a half cycle of the higher operation frequency.
|>
|>They do not _produce_ it. They may let it pass through by not storing any
|>energy to "cover" the zero-crossover time period.
|
| Of course they "produce" it.

How is it that ballasts of any kind "produce" flicker?

Actually, I think I might have an example. I have seen many cars with
rear tail lights, red in color, that are flickering. If they are powered
via the battery, why would that be? I presume they are LED, so there is
no reason for a voltage boost that might involve an AC circuit or at least
some kind of pulsing DC. I am suspecting the pulses come from either a
current limiter meant to reduce energy loss through dissipation (e.g. not
using a resistor), or an intensity modulation (in some of the cards, the
flickering stops when the break lights come on, bringing the same LEDs up
to full intensity). So that may well be a case of a "ballast" producing
the flicker. In the case of an AC powered fluorescent light, maybe it is
"producing" a 30 kHz flicker?

That flicker on the highway is somewhat annoying _because_ I have to move
my eyes around a lot to constantly check many vehicles. It does NOT cause
headaches (although some of the drivers out there do).
 
| Instead of using your approach I just connect an optical
| detector with sufficiently short response time to my
| oscilloscope and point it at the CFL. If the trace is nor
| flat then the light output is modulated.
|
| See
| http://www.robertsresearchinc.com/Papers/CFL_Modulation.BMP
| for an oscilloscope trace of the light output of a CFL with
| a small amount of modulation.
|
| The zero level is a bit below the first division on the
| screen, where the small arrow pointer is positioned. The
| average output is about 40 mV and the peak-to-peak ripple is
| about 7 to 8 mV.

I believe that is still enough flicker for some to perceive the flicker as
being present. Maybe a lot fewer people than if it were 39 mV peak-to peak.
It could all be a matter of degree, and a matter of marketing. They won't
want to double the size of the capacitor just to satisfy 0.5% of the market.
So some people will have to depend on their stockpile of incandescent lamps
or use the black market (it will be there).
 
| Victor Roberts wrote:
| [snip]
|
|>> Anyone who disagrees with the above conclusion, please raise your
|>> hand... ;o)
|>
|> OK - I'll raise my hand.
|>
|> Instead of using your approach I just connect an optical
|> detector with sufficiently short response time to my
|> oscilloscope and point it at the CFL. If the trace is nor
|> flat then the light output is modulated.
|>
|> See
|> http://www.robertsresearchinc.com/Papers/CFL_Modulation.BMP
|> for an oscilloscope trace of the light output of a CFL with
|> a small amount of modulation.
|>
|> The zero level is a bit below the first division on the
|> screen, where the small arrow pointer is positioned. The
|> average output is about 40 mV and the peak-to-peak ripple is
|> about 7 to 8 mV.
|
| I see your objection and evidence, but I disagree that this modulation is
| "visible flickering".
|
| If the zero on your oscilloscope screen (as you say) is close to the lower
| arrow, then I count a modulation amplitude of something like 5, max 10%. If the
| zero is lower, the amplitude is less than 5%.
|
| Are you kidding me Vic? That's a practically flat signal.

Doesn't look flat to me. But maybe it is beyond your ability to sense
any flickering. Only you can tell us whether that is the case or not.


| Obviously there is a certain minimum modulation of the first oscillator in order
| for the video to display a /visible/ flickering effect on the coupled
| oscillator, but we could then argue ad nauseam what should this minimum
| modulation be, in order for the flicker to qualify as "visible".

I hope you understand that the level of light changes that gets sensed as
flicker is different for different people.


| I mean, if the CFL displayed a 2% amplitude modulation, does the corresponding
| flicker qualify as "visible"? What about 0.00006%?

The lower the percentage, the fewer people would see it. I don't know if
the effect would be linear, or just what point is needed to eliminate all
people seeing it. And the effect within a person could vary as people
might not see different levels the same way.


| Let me then correct the conclusion, by adding the word "visible", which I did
| not include in my previous post (but had it on my web page regardless):
|
| "If the CFL pictured above displayed any _visible_ flicker, then the video above
| should also have displayed flicker, as per the analysis above. But it does not.
| Hence this particular CFL does not display _visible_ flicker".

I put more trust in the optical sensor than I do in a video camera,
especially if the camera has CCD technology and/or artifacts processing.


| In my eyes a modulation of 5-10% certainly does not qualify as "visible flicker"
| and is a huge improvement over magnetic ballast flickering on old fluoescent
| lamps where there is an actual shutoff of the light.

Although I have not used the kind of setup Vic did to measure the actual
light levels and changes, I do suspect I can sense flicker in a lot lower
level than you can. I do sense some degree of difference: some lights do
have more or less than others. I reference a neon night light I have as
a case with extreme flicker. CFLs are not as bad at that.

I am gradually using more and more CFLs and am putting them in places I do
not already have fluorescent lights, with the exception of long term task
lighting areas (work bench, kitchen, etc) which will stay with incandescent
until some better spectrums come out.
 
V

Victor Roberts

Jan 1, 1970
0
Victor Roberts wrote:
[snip]
OK - I'll raise my hand.

Instead of using your approach I just connect an optical
detector with sufficiently short response time to my
oscilloscope and point it at the CFL. If the trace is nor
flat then the light output is modulated.

See
http://www.robertsresearchinc.com/Papers/CFL_Modulation.BMP
for an oscilloscope trace of the light output of a CFL with
a small amount of modulation.

The zero level is a bit below the first division on the
screen, where the small arrow pointer is positioned. The
average output is about 40 mV and the peak-to-peak ripple is
about 7 to 8 mV.

I see your objection and evidence, but I disagree that this modulation is
"visible flickering".

I don't remember ever using the word "visible" to describe
this flicker.
If the zero on your oscilloscope screen (as you say) is close to the lower
arrow, then I count a modulation amplitude of something like 5, max 10%. If the
zero is lower, the amplitude is less than 5%.

That is about right for this example.
Are you kidding me Vic? That's a practically flat signal.

But it's not constant, and other samples have more
modulation.
Obviously there is a certain minimum modulation of the first oscillator in order
for the video to display a /visible/ flickering effect on the coupled
oscillator, but we could then argue ad nauseam what should this minimum
modulation be, in order for the flicker to qualify as "visible".

I don't understand the "first oscillator" statement. CFL
ballasts normally use a single self-oscillating power stage.
If run from a DC power supply with a constant voltage output
and no ripple, the high frequency envelope they produce will
have no significant low frequency modulation.
I mean, if the CFL displayed a 2% amplitude modulation, does the corresponding
flicker qualify as "visible"? What about 0.00006%?

I really don't have to worry whether or not 2% would qualify
as flicker since I'm measuring more than that amount.
Let me then correct the conclusion, by adding the word "visible", which I did
not include in my previous post (but had it on my web page regardless):

"If the CFL pictured above displayed any _visible_ flicker, then the video above
should also have displayed flicker, as per the analysis above. But it does not.
Hence this particular CFL does not display _visible_ flicker".

I still think that even for your experiment you need to use
an instrument other than your eye to make the measurement.
Some people are more sensitive to periodic variations in
light intensity than others.
In my eyes a modulation of 5-10% certainly does not qualify as "visible flicker"
and is a huge improvement over magnetic ballast flickering on old fluoescent
lamps where there is an actual shutoff of the light.

I'll have to post some photos of CFLs with higher modulation
levels and also a the modulation generated by a linear lamp
operating from a 60 Hz ballast for comparison.
With your permission, may I add your oscilloscope pic on my web page along with
your objection citing your name?

Yes. But make sure you state it is only one example and
other lamps may have more or less of the 120 Hz modulation.


--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.
 
| Victor Roberts wrote:
|> On Thu, 17 Jan 2008 12:48:15 +0200, "I.N. Galidakis"
| [cut]
|>> I see your objection and evidence, but I disagree that this
|>> modulation is "visible flickering".
|>
|> I don't remember ever using the word "visible" to describe
|> this flicker.
|
| You didn't. But you raised your hand in disagreement for the conslusive
| statement of my analysis, which was:
|
| "If the CFL pictured above displayed any visible flicker, then the video above
| should also have displayed flicker, as per the analysis above. But it does not.
| Hence this particular CFL does not display visible flicker."

Append: ... that this video system is able to pick up.

Video systems vary, too. But, just because a video system does not pick it
up, that does not mean it is not there. I'm sure I can find people that can
say they do not see any flicker in my neon night light (that goes 100% off
at 120 zero crossings per second). Would such testaments mean no flicker
is there? Of course not. It would just mean I have found someone that is
unable to see the flicker or lies about it.
 
D

Don Kelly

Jan 1, 1970
0
----------------------------
Paul Hovnanian P.E. said:
Well, maybe if you hang the fixtures too low. ;-)

The problem with that is that they get into cold water when on the toilet.
Shrinkage deals with that but the transient is painful.

wishfully thinking :),
 
T

TKM

Jan 1, 1970
0
In alt.engineering.electrical Victor Roberts <[email protected]>
wrote:

| Instead of using your approach I just connect an optical
| detector with sufficiently short response time to my
| oscilloscope and point it at the CFL. If the trace is nor
| flat then the light output is modulated.
|
| See
| http://www.robertsresearchinc.com/Papers/CFL_Modulation.BMP
| for an oscilloscope trace of the light output of a CFL with
| a small amount of modulation.
|
| The zero level is a bit below the first division on the
| screen, where the small arrow pointer is positioned. The
| average output is about 40 mV and the peak-to-peak ripple is
| about 7 to 8 mV.

I believe that is still enough flicker for some to perceive the flicker as
being present. Maybe a lot fewer people than if it were 39 mV peak-to
peak.
It could all be a matter of degree, and a matter of marketing. They won't
want to double the size of the capacitor just to satisfy 0.5% of the
market.
So some people will have to depend on their stockpile of incandescent
lamps
or use the black market (it will be there).

--
-----------------------------------------------------------------------------
| Phil Howard KA9WGN | http://linuxhomepage.com/
http://ham.org/ |
| (first name) at ipal.net | http://phil.ipal.org/
http://ka9wgn.ham.org/ |
-----------------------------------------------------------------------------

But, as we've said here before, inandescent lamps on either 50 or 60 Hz do
flicker. Some people can see it and there's even enough flicker with
incandescent sources that we calibrate such things as turntable speed with
strobe meters. I've done it myself. So, while continuing to use
incandescent lamps may be a satisfactory answer, it doesn't seem like
flicker is the cause -- or at least the whole cause -- of the problem.

How about some new theories?

Terry McGowan



It would seem we really don't understand what it affecting the people, like
PH who
 
| But, as we've said here before, inandescent lamps on either 50 or 60 Hz do
| flicker. Some people can see it and there's even enough flicker with
| incandescent sources that we calibrate such things as turntable speed with
| strobe meters. I've done it myself. So, while continuing to use
| incandescent lamps may be a satisfactory answer, it doesn't seem like
| flicker is the cause -- or at least the whole cause -- of the problem.
|
| How about some new theories?

Or old ones?

I used to think it was flicker that bothered me. But a few years ago I
came to the conclusion the problem is the spectrum of the light (and I
do not mean color balance).

I did a test just a couple days ago looking at some lights to see what I
could sense as flicker and compared that to viewing the same light in a
cylindrical mirror in motion that would spread the light in time across
my field of view where I can see the flicker as a row of dots (in the
extreme case ... fuzzy in non-extreme cases). What I found is that my
sense of flickering is not as strong with incandescent even though the
flicker is clearly there with the mirror test. But what I see in that
test is that the flicker in fluorescent lights changes color (it is more
reddish in the off cycles). Flicker in incandescent is more consistent
in color. So maybe my sense of flicker that works better with fluorecent
is really a sense of rapid color change.

Flicker at the 120 Hz level is not causing me headaches. What apparently
does cause them is a discontinuous spectrum where there emissions are at
specific wavelengths, and there is a gap between them. I've also used a
(near) monochromatic light source and found it works quite well. What I
think is the mechanism is that the eye jumps between the focus on one
wavelength and then the other, when sharp edges are present, such as the
black text on a white page. With a continuous spectrum, it is fuzzy, and
the eye settles in to the middle of the range.
 
J

Jeff Jonas

Jan 1, 1970
0
|> Or maybe we just need DC distributed in the home. But don't get any idea
|> that Edison was right ... he was selling pulsing DC.
| That's interesting. What do you mean by "pulsing DC" --- unregulated? The
| Smithsonian historical material indicates that the steam-powered generators
| were speed regulated and the load was just incandescent lamps initially, of
| course.
Edison's generators output DC by reversing the electrical connections every
half cycle. The end result is basically the same as a full wave rectifier
bridge. You get 2 pulses per cycle. I don't know what speed his generators
actually ran at.

Edison's generators were rather clever: they had a bus bar equilizer
so each shared the load equally (varied the field coil?).
If the generators were intentionally out of phase,
that would smooth out the combined output.

and a related story: many NYC buildings still depended on DC power
for elevators and ancient parts:

http://cityroom.blogs.nytimes.com/2007/11/14/off-goes-the-power-current-started-by-thomas-edison/

November 14, 2007
Off Goes the Power Current Started by Thomas Edison

By Jennifer 8. Lee
Con Edison's original power plant on Pearl Street.

Today, Con Edison will end 125 years of direct current electricity service
that began when Thomas Edison opened his Pearl Street power station
on Sept. 4, 1882. Con Ed will now only provide alternating current,
in a final, vestigial triumph by Nikola Tesla and George Westinghouse,
Mr. Edison's rivals who were the main proponents of
alternating current in the AC/DC debates of the turn of the 20th century.

The last snip of Con Ed's direct current system will take place at
10 East 40th Street, near the Mid-Manhattan Library.
That building, like the thousands of other direct current users
that have been transitioned over the last several years,
now has a converter installed on the premises that can take alternating
electricity from the Con Ed power grid and adapt it on premises.
Until now, Con Edison had been converting alternating to direct current
for the customers who needed it ... old buildings on the Upper East Side
and Upper West Side that used direct current for their elevators for example. ...
 
T

TKM

Jan 1, 1970
0
Jeff Jonas said:
Edison's generators were rather clever: they had a bus bar equilizer
so each shared the load equally (varied the field coil?).
If the generators were intentionally out of phase,
that would smooth out the combined output.

and a related story: many NYC buildings still depended on DC power
for elevators and ancient parts:

http://cityroom.blogs.nytimes.com/2007/11/14/off-goes-the-power-current-started-by-thomas-edison/

November 14, 2007
Off Goes the Power Current Started by Thomas Edison

By Jennifer 8. Lee
Con Edison's original power plant on Pearl Street.

Today, Con Edison will end 125 years of direct current electricity service
that began when Thomas Edison opened his Pearl Street power station
on Sept. 4, 1882. Con Ed will now only provide alternating current,
in a final, vestigial triumph by Nikola Tesla and George Westinghouse,
Mr. Edison's rivals who were the main proponents of
alternating current in the AC/DC debates of the turn of the 20th century.

The last snip of Con Ed's direct current system will take place at
10 East 40th Street, near the Mid-Manhattan Library.
That building, like the thousands of other direct current users
that have been transitioned over the last several years,
now has a converter installed on the premises that can take alternating
electricity from the Con Ed power grid and adapt it on premises.
Until now, Con Edison had been converting alternating to direct current
for the customers who needed it ... old buildings on the Upper East Side
and Upper West Side that used direct current for their elevators for
example. ...

Just like the newer cars, there could well be both AC and DC circuits in
houses and other buildings. In fact, that's already happening. Why the old
debate about which is "best"? Use what works most efficiently and
economically for the application.

Terry McGowan
 
|>|> Or maybe we just need DC distributed in the home. But don't get any idea
|>|> that Edison was right ... he was selling pulsing DC.
|
|>| That's interesting. What do you mean by "pulsing DC" --- unregulated? The
|>| Smithsonian historical material indicates that the steam-powered generators
|>| were speed regulated and the load was just incandescent lamps initially, of
|>| course.
|
|>Edison's generators output DC by reversing the electrical connections every
|>half cycle. The end result is basically the same as a full wave rectifier
|>bridge. You get 2 pulses per cycle. I don't know what speed his generators
|>actually ran at.
|
| Edison's generators were rather clever: they had a bus bar equilizer
| so each shared the load equally (varied the field coil?).
| If the generators were intentionally out of phase,
| that would smooth out the combined output.

2 or 3 alternators on the same rotor shaft, aligned at different angles,
could do that phase shift, I suppose.

Interesting concept ... 3 phase DC. I've seen circuits for AC to DC power
supplies that were designed to use 3 phase AC power to get smoother DC by
means of 3 full-wave bridges. I never worked out what kind of current each
of the 3 AC phase lines would have. The one I first saw was wired delta.
 
| [email protected] wrote:
|
|> Interesting concept ... 3 phase DC. I've seen circuits for AC to DC power
|> supplies that were designed to use 3 phase AC power to get smoother DC by
|> means of 3 full-wave bridges. I never worked out what kind of current each
|> of the 3 AC phase lines would have. The one I first saw was wired delta.
|
| Why stop at three? There are already DC/DC converters that use six
| "phases". The trade off is between circuit complexity and smoothing
| capacitors required. With IC components the former is trivial.

If by "six" you mean also to include the 180 degree opposing phases, that
would be what I include in "three" phases (six pulse). The result would be
a 300 or 360 Hz ripple for the capacitors to handle. If you get three
phase power in, that is perhaps the simple setup. But you could also get a
true six phases (twelve pulse) by having separate secondary windings on the
3 core transformer where 1 set are used for three of the phases, and
another set or two are used in pairs from separate cores to get cross phase
angles. You could also do this with a lot of separate transformers, but
doing it with just 3 cores seems the most economical. Then the ripple
would be 600 or 720 Hz and the capacitors could be even smaller.
 
| [email protected] wrote:
|>> [email protected] wrote:
|>>
|>>> Interesting concept ... 3 phase DC. I've seen circuits for AC to
|>>> DC power supplies that were designed to use 3 phase AC power to get
|>>> smoother DC by means of 3 full-wave bridges. I never worked out
|>>> what kind of current each of the 3 AC phase lines would have. The
|>>> one I first saw was wired delta.
|>>
|>> Why stop at three? There are already DC/DC converters that use six
|>> "phases". The trade off is between circuit complexity and smoothing
|>> capacitors required. With IC components the former is trivial.
|>
|> If by "six" you mean also to include the 180 degree opposing phases,
|> that would be what I include in "three" phases (six pulse). The
|> result would be a 300 or 360 Hz ripple for the capacitors to handle.
|> If you get three phase power in, that is perhaps the simple setup.
|> But you could also get a true six phases (twelve pulse) by having
|> separate secondary windings on the 3 core transformer where 1 set are
|> used for three of the phases, and another set or two are used in
|> pairs from separate cores to get cross phase angles. You could also
|> do this with a lot of separate transformers, but doing it with just 3
|> cores seems the most economical. Then the ripple would be 600 or 720
|> Hz and the capacitors could be even smaller.
|>
|
| 12 pulse converters are used routinely in variable frequency drives.

Right. But is it worth doing 12 pulse AC to DC conversion? Is the cost
of the extra transformer windings less than the savings in capacitors for
typical ripple reduction requirements?
 
Top