Do switch mode power supplies flicker in time with mains?

[William Gothberg]

On Thu, 20 Dec 2018 03:54:10 -0000, Clare Snyder <clare@snyder.on.ca> wrote:

On Wed, 19 Dec 2018 11:23:00 -0000, "William Gothberg" <"William
Gothberg"@internet.co.is> wrote:

Do switch mode power supplies flicker in time with mains? Specifically LED power supplies in commercially available domestic lamps. By in time, I don't mean at the same 50/60Hz, but anchored to it. I.e. if you have several such lamps each with their own built in supply, will they all flicker in time, using the mains frequency to keep them in time, or will they be random, making the room overall not flicker due to them all being random? And is there any way I can test this? I tried taking photos of them, but my camera only goes as fast as 1/2000th of a second, which shows all the lights at the same brightness each time, I suspect the flicker is above 2000Hz.
Leds (at least white ones) on a switch mode supply will not flicker
because the persistance of the phosphor is longer than the period of
the switching frequency which is more than 100kHz - typically 2 mhz.
The answer to the second part of the question is no, the switching is
not syncronized to the mains frequency on MOST switch mode power
supplies.

I'm definitely getting 100Hz flicker from it, I timed it using a slow camera shot (1/10th of a second) while moving the LED across the camera's field of vision. There were exactly 10 bright spots, although they were only 8% brighter than the dim spots. The LEDs don't go off completely. It's enough of a flicker for me to see with my eyes if I scan past the light, and I can detect anomalies when watching something rotating, like a drill chuck. I've got an oscilloscope on order, then I'll be able to check the signal to the LEDs (and in other parts of the supply) accurately.

 

[Daniel60]

William Gothberg wrote on 20/12/2018 11:34 PM:

On Thu, 20 Dec 2018 07:09:59 -0000, Daniel60
daniel47@eternal-september.org> wrote:

Mark Lloyd wrote on 20/12/2018 3:21 AM:
On 12/19/18 6:01 AM, William Gothberg wrote:

[snip]

They probably are fairly crude.  I know they flicker, for example if I
use my cordless drill, the chuck appears to spin the wrong way under
the LED lighting.
I remember seeing that with a washing machine (under fluorescent
lights). As the tub was slowing down, the row of holes around the tub
would appear to reverse direction. Same thing with (spoked) wagon wheels
in movies.
... and, in real life, the Mag wheels of some cars seem to be spinning
backwards, dependant on the speed at which the car is travelling!!

In real life?  I assume you mean under streetlighting.  That effect
can't occur with a steady lightsource such as the sun.

No, I'm sure I've noticed it in sunlight, as well.

And my 'real life' was to distinguish from the mentioned "wagon wheels
in movies"!
--
Daniel

 

[NY]

"Daniel60" <daniel47@eternal-september.org> wrote in message
news:q0d2fs$vof$1@dont-email.me...

They probably are fairly crude. I know they flicker, for example if I
use my cordless drill, the chuck appears to spin the wrong way under
the LED lighting.
I remember seeing that with a washing machine (under fluorescent
lights). As the tub was slowing down, the row of holes around the tub
would appear to reverse direction. Same thing with (spoked) wagon
wheels
in movies.
... and, in real life, the Mag wheels of some cars seem to be spinning
backwards, dependant on the speed at which the car is travelling!!

That's not what I'd call flicker, which is when an object is illuminated
sufficiently infrequently that you can see each pulse of light, even when
there's no motion in the scene. CRT TVs, especially European 625/25 as
opposed to US 525/30, were prone to flicker, especially if you saw one out
of the corner of your eye which seems to be more sensitive to flicker. CRT
computer monitors on 50 Hz progressive flicker more than on 60 Hz, and 72 or
90 Hz are better still. Interlaced scans are smoother for the same refresh
rate, but suffer from "twitter" where alternate lines flicker even though
the picture as a whole looks constant.

What you are describing is a stroboscopic effect of viewing a moving
contrasty object by intermittent light. That can happen at any intermittent
rate, even those which are far faster than the eye perceive as flicker. If
you let your eyes pan across an LED, you will see several images of it for a
fraction of a second - it's noticeable with some car rear lights or some
traffic lights, if there is relative motion (you are driving past the car or
traffic light).

In real life? I assume you mean under streetlighting. That effect can't
occur with a steady lightsource such as the sun.

No, I'm sure I've noticed it in sunlight, as well.

If you are seeing stroboscopic effects under DC light or sunlight, which are
constant not intermittent light level, then there's something very odd going
on.

And my 'real life' was to distinguish from the mentioned "wagon wheels in
movies"!

My dad's old record player had two sets of black and white stripes round the
edge of the turntable, so you could set the correct speed stroboscopically
under 50 Hz or 60 Hz mains light. What was surprising was that it worked
even with filament lights which have a long thermal inertia, so the light
doesn't change instantly from on to off, as with a LED or fluorescent tube
(*), but decays gradually. Despite this, the stripes were still fairly sharp
and not blurred. Of course, any setting of the turntable speed is only as
accurate as the mains frequency at the time, which can vary by up to +/- 0.5
Hz (http://mainsfrequency.uk/fm-home). I think the strobe markings were
calibrated for 33 1/3 rpm, with the assumption that the "gearing" (sprockets
or friction wheels) for the other speeds was exact, so if you calibrated at
one speed, it would be correct by definition at the other speeds.



(*) OK, with a fluorescent you see two images: a bluish one caused by the
very rapid on-to-off transition of the mercury discharge and a yellowish one
caused by the more gradual decay of the phosphor, though this decay is still
a lot quicker than the light from a filament bulb.

 

[The Natural Philosopher]

On 31/12/2018 15:01, NY wrote:

If you are seeing stroboscopic effects under DC light or sunlight, which
are constant not intermittent light level, then there's something very
odd going on.

something ELSE going on.

leaves in the breeze casting dappled shade...

in a moving car running past an avenue of trees...


--
I would rather have questions that cannot be answered...
....than to have answers that cannot be questioned

Richard Feynman

 

[NY]

"The Natural Philosopher" <tnp@invalid.invalid> wrote in message
news:q0dbrc$ndc$2@dont-email.me...

On 31/12/2018 15:01, NY wrote:
If you are seeing stroboscopic effects under DC light or sunlight, which
are constant not intermittent light level, then there's something very
odd going on.

something ELSE going on.

leaves in the breeze casting dappled shade...

in a moving car running past an avenue of trees...

True, although I was assuming that the OP really did mean constant sunlight,
not sunlight modified by periodic variations due to shadows.

Even in bright sunlight I'm sure I could easily "freeze" a rotating object
if it was illuminated by direct sunlight coming through the spinning blades
of a desk fan



The stroboscopic freezing of spinning objects can be very serious. My
grandpa was a model engineer in his spare time and he had a lathe. The room
was illuminated by fluorescent tubes, but he made sure that the light which
he shone on the work used a tungsten bulb with a nice long time constant to
avoid the chuck appearing to be stationary or slow-moving even though it was
spinning fast, so he didn't instinctively touch the tool to change it,
thinking that it had stopped. I wonder what precautions are used nowadays
with the increased use of (pulsed) LED lights. Maybe two banks of LEDs which
are 180 degrees out of phase so the light level is constant even though all
the LEDs are flashing quickly with a variable mark-space ratio to control
the required brightness. Or else maybe a random element added to the
flashing rate.

I have an LED desk lamp and the coarse brightness control is by switching on
one or both banks of LEDs (fine adjustment is by variable M:S ratio), and
with only one bank lit there is a lot more strobing (*) than with both
banks, so I bet the two banks are out of phase.

(*) For example if you move your finger rapidly from side to side.

 

[William Gothberg]

On Mon, 31 Dec 2018 12:36:09 -0000, Daniel60 <daniel47@eternal-september.org> wrote:

William Gothberg wrote on 20/12/2018 11:34 PM:
On Thu, 20 Dec 2018 07:09:59 -0000, Daniel60
daniel47@eternal-september.org> wrote:

Mark Lloyd wrote on 20/12/2018 3:21 AM:
On 12/19/18 6:01 AM, William Gothberg wrote:

[snip]

They probably are fairly crude. I know they flicker, for example if I
use my cordless drill, the chuck appears to spin the wrong way under
the LED lighting.
I remember seeing that with a washing machine (under fluorescent
lights). As the tub was slowing down, the row of holes around the tub
would appear to reverse direction. Same thing with (spoked) wagon wheels
in movies.
... and, in real life, the Mag wheels of some cars seem to be spinning
backwards, dependant on the speed at which the car is travelling!!

In real life? I assume you mean under streetlighting. That effect
can't occur with a steady lightsource such as the sun.

No, I'm sure I've noticed it in sunlight, as well.

And my 'real life' was to distinguish from the mentioned "wagon wheels
in movies"!

There is no way you could observe reverse rotation with the naked eye and a constant light source such as the sun. You need a light source that illuminates intermittently, the timing so that the spokes have moved quite a distance, so you assume they moved backwards.