Hi
Am I the only one to absolutely HATE those cheapo LED lights that strobe at 100hz? the vile things make me feel positively ill. How the heck do manufacturers get away with producing such rubbish?

No, you're not the only one. I hate them as well, I am in fact trying to stock up on incandesent Christmas lights, just in case... I also noticed that the LED backlighting in a colleague of mine's Volkswagen Jetta does the same, as well as the tail lights of some vehicles. The backlighting of the keyboard on my new MacBook Pro does the same, that is why I turned it off. The  most frustrating part is that the parts required to prevent that are fairly inexpensive, but the manufacturers go the cheapest way, regardless of quality. With incandescent and most discharge lamps you can get away with it because of the persistence of the phosphors or of the filament, but with LED, there is no persistence, so very noticeable flicker.

The flicker/"flicker free" is not as much matter of driver cost, but it's reliability (on mains lamp) and the result of the dimming requirements (vehicle lighting).

In vehicle's lights the LED brightness is controlled by PWM. By the duty ratio is controlled, if it is just a perimeter tail light (low duty ratio between ~10..25%), or a brake (or signal) light (50..100%). At the same time the brightness is adopted to the exterior light level, so at night it is 10/50% (tail/brake), at sunny daylight 25/100%. This behavior (adoption to the surrounding light level) is required by regulations, using all LED's is usually dictated by the need to reach the required brightness (for brake/signal) and the desire to prevent uneven and excessive wear (for tail light mode). And the only method available to control the brightness in such accurate way is the pulse width modulated flashing.
And the PWM is usually run at ~150..200Hz, what is a compromise allowing still quite slow edges (to prevent disturbance spreading across the vehicle network) and the flicker not being visible at least when the vehicle is still. Of course, when moving, the flicker may cause a stroboscopic pattern...

Similarly, the dashboard lights need to be regulated as well (to be visible during the day, but do not cause any glare at night), but as they are of way different colors and signalling different things, the only known method to keep the brightness somehow in sync is again the PWM modulation. So again the LED's are in fact flashing.


For the 100Hz flicker on general lighting lamps, it is indeed not as good as incandescent. Yes, it is technically feasible to make the LED drivers flicker free, but than involve to use the an electrolytic capacitor and that mean the life expectancy can not exceed about 10..15k hours. But as the flicker is not worse than magnetic ballasted fluorescent and/or HID, I don't think it is really any real issue except in special cases around fast rotating machinery.
But there the filter in the driver is not a solution either, because it may fail in such way, the lamp still lights, but does flicker and so can cause the stroboscopic effect problem, so pose a hazard without the user noticing that in time.

I can assure you that LED flicker is the worst of the worst. There is no persistence of phosphor unlike a fluorescent tube indeed the light output goes all the way to zero and back. Waving something bright under these lights will show that the on time is less than the mains half cycle. I recently stayed a night in an hotel where they had installed LED down lighters that strobed so appallingly that I had to spend my time with only the bathroom light (fluorescent) on. The bar and restaurant areas were the same and I could barely manage to eat my meal because of feeling so uncomfortable. DAMN these blasted el cheap LEDs! On the other hand I have several LEDs of good make ie Philips and gasp, Tesco, that do not strobe at all and are perfectly satisfactory. I also loath sodium street lights thanks to their strobing and if in a hotel with poor light exclusion in the rooms and the inevitable HID lights outside I just cannot sleep so I carry a roll of bin liners and some sticky tape to cover the window if necessary.

 Aaargh... etc

I also find strobing annoying, however the worst offender is actually the TV as the flicker is obvious, especially if I'm reading something while the TV is on. The flickering from HID lights doesn't really bother me unless the arc starts jumping around or there's a bit of rectification which seems to happen sometimes, especially after a voltage dip. Seems to be more of a problem with SOX.

The strobing LED vehicle lights are particularly distracting ..


BG

The LEDs of our traffic lights, flickering. I usually don't notices this flickering unless I uses my camera.
Why LED backlights for LCD displays aren't dimmed by the same mothed of fluorescent backlights (Dimming fluorescent backlit LCD displays, don't causes any flickering at all, unlike LED backlit LCD displays)

The LEDs of our traffic lights, flickering. I usually don't notices this flickering unless I uses my camera.

Again it came from the dimming (to be visible during sunny day, but don't cause glare at night), the easiest method win. Moreover with signals the relative brightness between different colors have to be kept under quite tight control, as well as the color itself, where both vary (different LED "chemistries" used for different colors behave differently) with the actual current. So by the PWM the current, so relative brightness and color effects is all the time the same on all dimming setting.
The dimming, better to say generally the driving style is not designed to look nice on cameras, but for direct view. Nobody care about cameras at all.

Why LED backlights for LCD displays aren't dimmed by the same mothed of fluorescent backlights (Dimming fluorescent backlit LCD displays, don't causes any flickering at all, unlike LED backlit LCD displays)

But mostly it is exactly the same method: PWM at few 100Hz switching ON/OFF the CCFL ballast.
The LED's copied the same concept, as the PWM signal usually originate from the main display chipset originally designed for the CCFL, so regardless of the actual dimming method, the ballast should be controlled by this PWM signal. So the simplest way is just directly use it.
The reason for PWM and not varying current is the color stability: Mainly the CCFL, but as well the LED's to some degree, color change with the current - what vary is the ratio between direct and phosphor converted light components. In CCFL the direct are the green, blue and some yellow mercury lines, whose are produced mainly at higher current densities, the rest come from the UV converted by the phosphor and this tend to lower it's efficiency at higher levels. The result is more pinkish (so off-white) color at lower brightness.
With LED's the direct radiation is blue, the green till red is converted and again, the phosphor efficiency drop at higher levels, so more current mean slightly higher CCT. But compare to CCFL, it stays closer to the blackbody color (assume it was initially there).

If fluorescent backlighting dimmed by PWM, so why it isn't flickers at the lowest setting? (When I wave my finger in front of my fluorescent backlit TEAC 7005L12 17" 4:3 LCD monitor, when its brightness is at the lowest settings, the shade of my finger on the screen still looks blurry and haven't a sharply defined edges.

What could differ is the frequency, or synchronization with the screen refresh (the refresh cause slight flicker of the LCD itself, what can create interference products with the backlight dimming flicker)
Why the frequency would be different, I don't know. CCFL dimming was operating in the range of 200..400Hz, but I do not see any reason for LED's to have it any different (the display chips have this frequency firmware-programmed, so it can generate PWM signals from few Hz till few kHz).

Maybe the features like dynamic contrast (for dark scene it dim down the backlight, instead of shutting off the LCD pixels; it allow way better picture with changing scenes) or so alter the dimming way more and so yield way lower levels, so duty ratio, than was common with older monitors using CCFL. Then the flicker would be associated with this feature, rather than the user "brightness" setting. But the fact is, this feature require very wide dimming range and that is not possible with CCFL.

The fact is, the LED can easily go to way lower duty ratios than the CCFL, so with CCFL the minimum brightness setting was around 25..30% (what does not flicker), but with LED's you may go easily down to 5..10%, what could become visible.
Other thing is, the rising and falling slopes of the light intensity are sharper with LED's, the resonant nature of the CCFL make them more smooth (what yield to the limit in the dimming range), so it yield not as sharp stroboscopic images.

No, you're not the only one. I hate them as well, I am in fact trying to stock up on incandesent Christmas lights, just in case... I also noticed that the LED backlighting in a colleague of mine's Volkswagen Jetta does the same, as well as the tail lights of some vehicles. The backlighting of the keyboard on my new MacBook Pro does the same, that is why I turned it off. The  most frustrating part is that the parts required to prevent that are fairly inexpensive, but the manufacturers go the cheapest way, regardless of quality. With incandescent and most discharge lamps you can get away with it because of the persistence of the phosphors or of the filament, but with LED, there is no persistence, so very noticeable flicker.

Have you ever tried following a modern volvo at night?   I didn't notice the strobing if you look at it straight on but if the car moves rapidly (goes over a bump or over a pot hole) I could defiantly notice a strobing.  Very disconcerting and I think dangerous as I was trying to see the strobing. 

What could differ is the frequency, or synchronization with the screen refresh (the refresh cause slight flicker of the LCD itself, what can create interference products with the backlight dimming flicker)
Why the frequency would be different, I don't know. CCFL dimming was operating in the range of 200..400Hz, but I do not see any reason for LED's to have it any different (the display chips have this frequency firmware-programmed, so it can generate PWM signals from few Hz till few kHz).

Maybe the features like dynamic contrast (for dark scene it dim down the backlight, instead of shutting off the LCD pixels; it allow way better picture with changing scenes) or so alter the dimming way more and so yield way lower levels, so duty ratio, than was common with older monitors using CCFL. Then the flicker would be associated with this feature, rather than the user "brightness" setting. But the fact is, this feature require very wide dimming range and that is not possible with CCFL.

The fact is, the LED can easily go to way lower duty ratios than the CCFL, so with CCFL the minimum brightness setting was around 25..30% (what does not flicker), but with LED's you may go easily down to 5..10%, what could become visible.
Other thing is, the rising and falling slopes of the light intensity are sharper with LED's, the resonant nature of the CCFL make them more smooth (what yield to the limit in the dimming range), so it yield not as sharp stroboscopic images.

I will try this in my Samsung monitors of my father and my mother home. The 0% brightness of my TEAC screen is actually way brighters than my parents Samsung monitors, and I actually work with it all the time, so it is expected that if it is dimmed by PWM, at that brightness wiht flicker would still be invisible.

Have you ever tried following a modern volvo at night?   I didn't notice the strobing if you look at it straight on but if the car moves rapidly (goes over a bump or over a pot hole) I could defiantly notice a strobing.  Very disconcerting and I think dangerous as I was trying to see the strobing. 

It is not only Volvo, but all cars with LED tail lights ysing dimming to alter between perimeter to brake light functionality. So >90% of all LED rear light assemblies. Yes, if you focus on that, you will be able to notice the flicker by one way or the other. But that is not the point of the light design. The point is to mark the corner of the vehicle at low visibility and to signal to the driver behind and to not cause too much glare as a secondary requirement. That's all...

When driving you should not focus on any tail light, but to the traffic condition in the first place.
If someone let his attention to focus on how some tail light operate instead of all the surrounding traffic, it is indeed very dangerous. But if someone become so easily distracted when driving, then he have to stop driving and work on his selfdiscipline first. Otherwise there are tons of way stronger distractions around, each way more potent to "steal" way too much of the drivers attention...

I remember in the UK when PEEK and Siemens cam out with their first generation of LED traffic lights, they had huge problems with Red Light cameras.  As the LEDs are being over driven/strobed to maximise brightness there was a theoretical instance when pictures of a motorist jumping a red light could have been taken by the camera but the LEDs could be in their "off/dim" state.  Therefore the motorist could have got off on a technicality hence a halogen signal was always sited in frame of the camera.  I hear this has been resolved with the introduction of CLS technology.  Instead of 100 or so LEDs being over driven in an aspect there are now 4 or 5 super bright LEDS that are being strobed offset to each other with optics allowing the signal to confirm to minimum light output legislation.

Well, if the camera can not make a picture of the LED light, it is failure of the camera. Nonworking camera does not pose a safety hazard, while the nonworking signal light does so. So it is quite stupid to impose unnecessary restrictions on traffic light design rather then to the camera system, each design have to be a compromise, so more restriction on one side mean other parts have to suffer. Here allowing the 100Hz flicker mean simple cabling and no filter (= troublesome electrolytics), what mean the related problems are easily avoided.
And e.g. the film/TV industry solved basically the same problem (HID flicker interfering with the frame rate) many years ago by just synchronizing the camera with the light. Don't tell me it is not possible to do the same, the traffic light flicker just with the mains, so the same mains could be used to synchronize the camera shutter.

But the LED traffic signals have way more severe problems related to bad weather: As they are energy efficient, they run very cold. And that mean, the snow and other mess stay on them, eventually blocking the light out, so making the light effectively not working. But compare to e.g. burned out incandescent, this situation can not be detected, so then treated (e.g. by complete signal shut down, so the traffic become controlled by signs or right hand rule, or whatever is in place in given legislation) by the controller.

From speaking to engineers the regulations for traffic lights here are very rigid.  There has to be a minimum number of lumens output, has to be a certain colour of RAG etc.  Also has to comply with industry wiring regulations, with the exception of halogen has to be resistant to phantom sun.  Another big problem is lamp monitoring, as I understand it each aspect has its own driver but how do you monitor if the LEDs are working as the driver board draws the same current and voltage regardless if the LEDs are working.   Traditionally with halogen this was done with affectively a circuit test back at the controller.