100Hz flicker in lighting is a problem?

[james_s]

Yes, and also you aren't normally sitting very close to a CRT TV, computers on the other hand, 60Hz refresh on a larger SVGA monitor was horrid, some people didn't seem to notice at all but I couldn't stand it. Bump it up to 70-75Hz and it almost vanished to my eyes. On smaller monitors and those that displayed light text on a black background the flicker was not noticeable.

The first time I saw a PAL-50 TV I was blown away at how flickery it looked, it was almost like looking into a strobe light. I'm pretty sure it's one of those things where if you see it all the time you get used to it and your brain tunes out the flicker.

Personally I find that when I'm tired I'm much more sensitive to flicker. I have an older TDS scope with a sequential color display and normally it looks fine to me but if I try to use it when I'm very tired the flicker is pronounced and gives me a headache. I still love that display due to the novelty though, full color out of a B&W CRT with no shadow mask, it's neat, but that's off topic.

[NiHaoMike]

120Hz and 144Hz monitors that have pulsing backlights have selection of what duty cycle to pulse at. I run mine at 5% for the least motion blur. For general lighting, I see no reason to use such a low duty cycle. Or any real reason to use low frequency PWM dimming in the first place. Even cheap capacitive ballasted LEDs are not that difficult to get flicker free.

I like to know about the one who claims to be able to see flicker at 30kHz. In PC gaming, 144Hz is a popular choice for high end monitors and anything more is considered far into diminishing returns.

[tooki]

Yes, and also you aren't normally sitting very close to a CRT TV, computers on the other hand, 60Hz refresh on a larger SVGA monitor was horrid, some people didn't seem to notice at all but I couldn't stand it. Bump it up to 70-75Hz and it almost vanished to my eyes. On smaller monitors and those that displayed light text on a black background the flicker was not noticeable.

Yes!! I also assumed that the phosphor persistence on computer displays was shorter than on TVs, given the difference in effect.

The first time I saw a PAL-50 TV I was blown away at how flickery it looked, it was almost like looking into a strobe light. I'm pretty sure it's one of those things where if you see it all the time you get used to it and your brain tunes out the flicker.

Same here, as an American who moved to Europe! I found 50Hz TVs to be flickery even head-on. It's no wonder that 100Hz TVs were common in Europe if you could afford one, whereas I'm not even sure if 120Hz TVs were ever made.

Personally I find that when I'm tired I'm much more sensitive to flicker. I have an older TDS scope with a sequential color display and normally it looks fine to me but if I try to use it when I'm very tired the flicker is pronounced and gives me a headache. I still love that display due to the novelty though, full color out of a B&W CRT with no shadow mask, it's neat, but that's off topic.

Any info you can share on this??

I like to know about the one who claims to be able to see flicker at 30kHz. In PC gaming, 144Hz is a popular choice for high end monitors and anything more is considered far into diminishing returns.

You don't need it for a display, but that's not the topic of this thread. Again, google the "flicker fusion threshold" to learn about low duty cycle and saccading. Until you have, don't dismiss flicker even at high frequency.

[Tomorokoshi]

I like to know about the one who claims to be able to see flicker at 30kHz. In PC gaming, 144Hz is a popular choice for high end monitors and anything more is considered far into diminishing returns.

Let't calculate.

You sweep your eyes at a rate of 90 degrees in 0.25 seconds with a 30 KHz screen at a distance of 2 meters. The radial swept distance is 5.23 meters, for a rate of 32.86 meters / second. Call it 30 meters / second.

At 30 KHz, the step distance per flicker cycle is (30 meters / second) / (30000 cycles / second) = 0.001 meters / cycle, or 1 mm / cycle at a distance of 2 meters.

Seems within reason enough.

[Marco]

120Hz LED is a totally different beast. It has zero persistence

The language of LED tvs has never been very precise, but generally 120 Hz means motion compensated with capture and hold (so full persistence between frames).

240 Hz used to generally mean 120 Hz motion compensated with the backlight only on for a short time.

[ocset]

This is taking your ignorance to a whole 'nother level. How the f**k are you working as an engineer at a lighting manufacturer when you demonstrably know nothing about electronics or lighting?!?

Thanks, OK, i was workigng at a lighting manufac.......my contract expired two weeks ago....i have now been placed on a max 6 month contract, and my pay reduced by 25%, and i am only now here until  i handover to the next engineer.

[NiHaoMike]

I wonder if his insistence of not using electrolytic capacitors at all is making it difficult to make the lights flicker free and within budget.

[Tomorokoshi]

I like to know about the one who claims to be able to see flicker at 30kHz. In PC gaming, 144Hz is a popular choice for high end monitors and anything more is considered far into diminishing returns.

Let't calculate.

You sweep your eyes at a rate of 90 degrees in 0.25 seconds with a 30 KHz screen at a distance of 2 meters. The radial swept distance is 5.23 meters, for a rate of 32.86 meters / second. Call it 30 meters / second.

At 30 KHz, the step distance per flicker cycle is (30 meters / second) / (30000 cycles / second) = 0.001 meters / cycle, or 1 mm / cycle at a distance of 2 meters.

Seems within reason enough.

My limit is 3 KHz with an LED driven by a 20% duty cycle square wave.

[james_s]

Any info you can share on this??

On what? The NuColor display, or my reaction to it? There is plenty of info out there on the display, it's used in all of the color TDS500, 600, and 700 series scopes from the 90s.

[Zero999]

It seems plausible that many people find flicker uncomfortable. There are some unfiltered LED lamps at my parents' house and I do notice the flicker. It's not too bad, when there are other filtered lamps on though.

Flicker can be dangerous around rotating machinery, as an object spinning at 6000rpm will appear stationary in a light flickering at 100Hz.

[Siwastaja]

Those who are bothered by 120Hz flicker must have had a really bad time back in the days of 60Hz TVs. There are also 240Hz monitors and TVs out there, but they're mostly for 3D glasses so that the effective framerate would be 120Hz after the glasses demultiplex it.

CRT phosphors do not blank out immediately. Actually, the raster scan happens for infinitesimally short time at each phosphor site, but the persistence of the phosphor makes it glow even when not exposed to the electron beam. If you turn off the lights in the room, then turn off the TV, you'll see that some persistence is there even for 10-15 minutes!

Basically, phosphors varied case-by-case. Some 60Hz CRTs can be as non-flickery as standard lightbulbs, while some flicker a lot. Computer monitors tended to use quicker phosphors, requiring refresh rates up to 85Hz to provide flicker-free image.

[tooki]

Any info you can share on this??

On what? The NuColor display, or my reaction to it? There is plenty of info out there on the display, it's used in all of the color TDS500, 600, and 700 series scopes from the 90s.

Info on the display. But with the name I can now google it. Thank you!!

[BrianHG]

Count me as a 100hz hater.  Even on 50hz European CRT.  In fact, I had super expensive progressive 120hz desktop screens and a 40 inch 90Hz computer monitor as a television with scan rate accelerator since the beginning of 2000 and never looked back.  Older on chip DLP projectors were impossible for me to look as including full color LED billboards which are configured only around 180hz.

My partner never saw a thing.  He was happy with 60hz computer screens.  So, you can say that some people just have faster response retinas in their eyes.

120hz lighting is borderline for me today, if I had to drive in 100hz lighting, I would throw up...

[JPortici]

FWIW i am sensitive to flicker. I get strong headaches
- From fluorescent lights, but that may also be correlated with the wwhite tint, glare..
- From LED headlight from certain auto manufacturer, mainly BMW and earlier Alfa Romeo. Seriously, sometimes i think i can see a hint of the sequence, like when you make a video and watch it back
- it if i see a 3d movie, plus i have a very low perception of 3d altough i don't know if it's correlated.
- from monochrome CRTs. Scopes, early monitors always gave me big headaches. see my previous thread -> https://www.eevblog.com/forum/chat/sickness-from-monochrome-crts/

one thing for sure, in some environments you mustn't underestimate stroboscopic effect like when dealing with machinery.

regarding street lights, yeah i can very easily see them flicker. LED lights were a blessing, but not because they don't flicker.. because they are much more effective! the target area is REALLY illuminated, the surroundings are not (which can be a good or bad thing)

[ocset]

FWIW i am sensitive to flicker. I get strong headaches
- From fluorescent lights, but that may also be correlated with the wwhite tint, glare..

Thanks JPortici, but i  feel i must tell you that the actual light output of  old magnetic ballast fluorescent lights does not flicker.....because of the luminescance of the fluorescent gas in the tube...it stays lit even when the current through it goes to zero.........so the fluorescents do not flicker.

[Siwastaja]

Fluorescents do flicker quite a bit - I have measured the fact myself.

An easy way to measure is to take a digital camera capable of very short exposure time (such as 1/4000 s), then randomly take a shitload of images to give statistically significant number of samples; then just plot the occurence of different luminances.

With fluorescents, you notice an interesting thing when you do this: the color temperature flickers radically as well.

[ocset]

Fluorescents do flicker quite a bit - I have measured the fact myself.

Thanks yes but the light doesnt go anywhere near zero...the current does, but the light doesnt....fluorescents are like "CFD <25% " in this site..
https://www.derlichtpeter.de/en/light-flicker/cfd/

So for fluorescents the flicker index (CFD)  is less than 25%...which is said to be "rarely perceptible to humans"

Fluorescents do flicker quite a bit - I have measured the fact myself.

Thanks,i searched the web and its not possible to find a light output waveform diagram for a fluorescent.

It is needed info, as per this page provided kindly by earlier contributor (thankyou maukka)
https://www.derlichtpeter.de/en/light-flicker/cfd/


....this page says that if CFD (flicker index) is >50% then it is illegal for indoor or outdoor work illumination....but i cant find the bit in DIN EN 12464 where it actually  says that its illegal...do you know where it says this?

The website just above says that the worst case of flicker (CFD>100%) IS "Hazardous according to DIN EN12464  and must be avoided" ...but then goes on to say that it's a desirable thing to have in discoteque's...............hmmmm...is this thing really hazardous then?......its ok in discoteques......how many discos have hazardous light warnings.

I am sensing that there is a lot of old cobblers going on here....aren't you?

regarding street lights, yeah i can very easily see them flicker. LED lights were a blessing, but not because they don't flicker..

Thnaks, but again the old sodium vapor streetlamps did not flicker...the luminescance carried on throughout the zero current intervals.

..............................
A survey was actually done in a busy shopping street in  Manchester in winter, .....streetlights using SMPS’s with outputs which were PWM dimmed at 100Hz and duty of 50% were used….at the end of the street, people were asked  (3000 people total) if they noticed the flicker in the streetlights……in virtually all cases…the people actually said they didn’t even notice the  fact that there were streetlights….there was no extra  illumination from shop front lighting in that particular street.

[helius]

I have an older TDS scope with a sequential color display and normally it looks fine to me but if I try to use it when I'm very tired the flicker is pronounced and gives me a headache. I still love that display due to the novelty though, full color out of a B&W CRT with no shadow mask, it's neat, but that's off topic.

Any info you can share on this??

Tektronix used a sequential-color system that was branded/licensed NuVision. The CRT was an ordinary monochrome tube with a single electron gun, but with a LCD filter stack in front of it. As the liquid crystal cells switched on and off, they steered the polarization of light through colored polarizers, one per color field. It was found on some TDS scopes and certain other products like virtual reality HMDs.
There was another wild color technology based on electron beam overpenetration. The phosphor on the tube face was applied in multiple layers, each layer a different color. The electron beam energy determines how far into the phosphor layer the electrons can penetrate, which enables multiple colors. This was found in some HP and Tek displays.

Part of what makes these techniques interesting is that they enable true vector displays in color.

[Gyro]

Tek Nuvision always gave me problems. It was a lovely image if you looked at it straight-on (as long as my eyes weren't flicking backwards and forwards too much) but if I was soldering something on the bench it played hell with my peripheral vision.

I don't know what multiplexing rate they used (presumably LCD shutter response time limited) but it certainly wasn't fast enough.

[NiHaoMike]

There are LCDs that implemented color with a RGB backlight rapidly cycling through the colors, but it never caught on due to the difficulty of making a LCD fast enough for that to work well. Without the color filters blocking 2/3 of the light, it even had a black and white mode that used only ambient light, competing with Eink displays.

[janoc]

I wouldn't call myself "sensitive" to the flicker but 50-100Hz flicker can be (and is) a problem in certain situations.

- Office workers. Flourescents are terrible and not recommended anywhere where people work with computers. The flicker is alone invisible for most but it can "beat" against the backlight of the screen (which typically runs at 60Hz) and you get 10Hz flicker in your peripheral vision which is really annoying. Lot of people get a headache from this without even realizing.

- Any place where things are moving. Stroboscopic effect is very real and such lighting could be literally deadly - e.g. on a factory floor with moving/spinning machinery or airports. In fact, the airport staffers are specifically trained to look for signs whether an engine is running other than a spinning fan/propeller because those could appear stationary to them thanks to the strobing of the lighting. The results of missing something like that are a bloody and greasy spot

When it comes to LEDs those are even worse at 100Hz than flouros, because the phosphor in the flouros has a persistence and won't go off completely between cycles. On the other hand, LEDs will, so the strobing is more visible.

Don't be a cheapskate and build a proper driver for your LEDs so that they don't strobe on 100Hz.

Re flourescents supposedly non flickering - connect a photodiode to your oscilloscope and point it at the nearest florescent. You will see how "non flickering" the lamp is first hand.

This is taking your ignorance to a whole 'nother level. How the f**k are you working as an engineer at a lighting manufacturer when you demonstrably know nothing about electronics or lighting?!?

Thanks, OK, i was workigng at a lighting manufac.......my contract expired two weeks ago....i have now been placed on a max 6 month contract, and my pay reduced by 25%, and i am only now here until  i handover to the next engineer.

Sorry to hear but given your history here, I am frankly surprised they kept you on board even this long. Some of the things you have presented here were quite shocking even for a non-EE software guy like me.

[StillTrying]

Thanks,i searched the web and its not possible to find a light output waveform diagram for a fluorescent.

There's the shape of a 9W CFL's light here, center image.
www.eevblog.com/forum/chat/20w-halogen-bulb-viewed-by-a-photodiode/msg1316515/#msg1316515

Unfortunately I'd not recorded the 0% light level for that one, so don't know the p-p depth of the flicker.

[Siwastaja]

Thanks yes but the light doesnt go anywhere near zero...the current does, but the light doesnt...

Yes, this is true. According to my tests, the light output drops to around 10-30% of the maximum in the valleys. Which is a lot less than incandescents at about 70-90%. So, more flicker. But not horrible, like some (improperly driven) LEDs, which not only go to 0%, but also stay there for significant portion of the cycle.

So for fluorescents the flicker index (CFD)  is less than 25%...which is said to be "rarely perceptible to humans"

Which seems true as well - the incidences indeed seem "rare". As said before, I knew one guy. A few have popped up here. Not usual, but not non-existent, either. "Rarely perceptible", indeed.

What does this mean? Whenever possible, we try to build our public spaces so that you can access them with wheelchair, even though it's "rare" to be in one.

So, if you can design a light which is good for everyone and not just "most of the people", I think you really should. You'll make a better, more universal product with good karma, not causing problems.

A good aim is to have at least the same performance to the old fluorescents, or preferably a bit better. Add some extra margin, because it's difficult to directly compare these two technologies which will result in different "shape" of waveform.

It's worth remembering that if you don't actively keep this in mind, you can end up with orders of magnitude worse with LEDs. And this is not OK.

OTOH, as I said in my first post to this topic, it makes no sense to aim for perfection here. Old mains-driven fluorescents were "fairly OK". Aim for a bit better (like half the "flicker index") and you'll be fine.

As a side note, when talking about "fluorescents", we mostly tend to mean the old style of direct drive from the mains through an inductor ballast. In case of electronic transformers, including CFLs, some are practically flicker-free, some flicker, depending on how much capacitance they use after rectifying the mains.

[james_s]

Tektronix used a sequential-color system that was branded/licensed NuVision. The CRT was an ordinary monochrome tube with a single electron gun, but with a LCD filter stack in front of it. As the liquid crystal cells switched on and off, they steered the polarization of light through colored polarizers, one per color field. It was found on some TDS scopes and certain other products like virtual reality HMDs.

I don't think it used filters, my recollection is that somehow controlling the degree to which the liquid crystals rotated could control the wavelength of light it lets pass. At any rate there are no visible cells or pixels, the filter appears as a single continuous window and allows a really unique looking display. It has the razor sharp monochrome CRT look but in full color. If I'm tired the flicker is distracting but otherwise it looks great and it's so unique that the cool factor makes up for any deficiencies.

[ocset]

There's the shape of a 9W CFL's light here, center image.
www.eevblog.com/forum/chat/20w-halogen-bulb-viewed-by-a-photodiode/msg1316515/#msg1316515

Thanks, may i ask, do you mean in  the Reply #7 ?