• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Do LED lights flicker? (ac or dc current)

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Does anyone know if HQI (Hydrargyrum Quartz Iodide) have significant amount of mercury, I know hydrargyrum is latin for mercery, so it has to have some mercery. Just wonderin ghow much.
 
The 4100K bulbs are a LOT brighter than the 7000K ones
Click to expand...

That's because the human eye is most sensitive to light around 5000k, so it's no surprise. Otherwise, the term 'daylight' or 'full spectrum' in regards to a fluorescent bulb is essentially false advertising because there is no such thing. Fluorescent bulbs have characteristic jagged spikes in spectral emission which makes them approximate certain kelvin temps when summed on a graph, but it's never quite as good as a true, full spectrum source like HID, incandescent, MH, etc for color reference. These type of light sources essentially 'spray' photons of a much wider range of light frequencies to fill in spectral gaps, but it also accounts for their inefficiency.

3500k CFLs are hard to find, but in my practice they are the most visually appealing of all the fluorescent types and most ideal for home use because they mimmick late afternoon sun. Gorgeous light if you can find them. 2700k is the most common but way too orange, and 5000-5500k sold as 'full spectrum' by home improvement stores too 'hard' and cold. They make great porch lights though because they keep undesireables away.

HID, halogen and Xenon are all pretty much grease off the same meatball and more marketing than anything else. Some minor differences in trace gaces filling the enclosure, but not much else. Xenon bulbs are just more likely to be designing for 12-24 volt range, and dyed blue for car use.

MH is making a comeback because newer electronic ballasts are increasing their efficiency, and at least in the 400watt and higher range they can fill a big space more efficiently than pretty much anything else. Bulbs are short lived, but when you're trying to light a 100,000^2 foot warehouse it pays for itself. Also, even though MH has a lot of industrial waste issues, CFL is almost just as bad given the ballasts used for them are basically garbage, and given typical low quality chinese production they almost never approach their 5000+ hour lifespan. T5 is actually the 'top dog' in terms of fluorescent technology right now, and in some cases comes close to matching MH in some industrial settings.

Obviously LED beats everything else in terms of efficiency, but the production scale still isn't in reach yet.
 
Originally posted by: ICRS
Does anyone know if HQI (Hydrargyrum Quartz Iodide) have significant amount of mercury, I know hydrargyrum is latin for mercery, so it has to have some mercery. Just wonderin ghow much.
Click to expand...

Metal halide (both quartz and ceramic) lamps contain significant amounts of mercury.

Quartz MH lamps contain between 10-20 mg of mercury per bulb (for common sizes: 70 - 250W). Very high power lamps (e.g. 2000 W) contain more (up to 0.5 gram for certain types of light), and the smaller low power lights less. This photo of one of my iwasaki MH lamps actually shows the mercury droplets in the bulb.

In terms of mercury content per light output, they're similar to older CFLs - but they typically come in higher powers and intentisities, so contain correspondingly more mercury. At the same time CFLs tend to contain significantly more mercury than full-sized fluorescent tubes (especially the latest generation in high efficiency T5 tubes). [A lot is due to the operating gas pressure of the lamp - full sized tubes are low pressure, low temperature lamps. CFLs are medium pressure, and operate at a medium temperature necessitating a short warm up time. MH are high pressure, and operate at about 900 C, requiring about 5 minutes for full warm up] Progressively more mercury is required to sustain the pressures and temperatures required.

In the case of CFLs though, they are somewhat overfilled with mercury (or at least have been) because eventually the mercury becomes unavailable and the lamp fails. More mercury also improves warmup time. However, there's now a move to reduce mercury content in CFLs, and the manufacturers are now cutting down to the bare minimum - which means that the latest generation have up to 75% less mercury than previous CFLs.

There seems to be somewhat less scope for mercury reduction in MH, although this remains a path for development. Although HP sodium lamps (which are similar to MH in their mercury content) are now being released in mercury free versions (at the expense of the already poor color rendering).
 
Originally posted by: spikespiegal

3500k CFLs are hard to find, but in my practice they are the most visually appealing of all the fluorescent types and most ideal for home use because they mimmick late afternoon sun. Gorgeous light if you can find them. 2700k is the most common but way too orange, and 5000-5500k sold as 'full spectrum' by home improvement stores too 'hard' and cold. They make great porch lights though because they keep undesireables away.
Click to expand...
I guess everyone's got their own preference. Philips sells some T8 32W tubes that are rated 6500K. I love that light. Unfortunately, the closest I can get in a CFL is 5000-5500K. Oh well, close enough.


Obviously LED beats everything else in terms of efficiency, but the production scale still isn't in reach yet.
Click to expand...
I recall reading that LED lighting is actually about as efficient as halogen lights. As I understand it, LEDs are great for single-color applications, such as traffic lights. There, instead of filtering out all but one wavelength as is done with incandescent sources, only one wavelength is produced in the first place, thus providing a considerable energy savings. With white light output, I think it's that blue light hits a phosphor to produce something that's approximately white, but that this process isn't especially efficient.
Where they do.........*sigh*....."shine" is in terms of bulb life and shock resistance.

 
http://www.popularmechanics.co...provement/4215199.html
as for warmness... light is affected by the room color and interior, light bounces, so that may affect perfection of bulb color. i don't have a problem with the warmness of the current cf's. decent ones have an appealing warmness, cheap ones can sometimes have an off putting warmness. its just a vibe thing.


anyways on page 7 theres the sylvania daylight extra and thats 3500k if you want
 
Originally posted by: Jeff7
I recall reading that LED lighting is actually about as efficient as halogen lights. As I understand it, LEDs are great for single-color applications, such as traffic lights. There, instead of filtering out all but one wavelength as is done with incandescent sources, only one wavelength is produced in the first place, thus providing a considerable energy savings.
Click to expand...

That's pretty much right. The latest commercial white LEDs aren't quite that poor (that statement was true a couple of years ago though). Commercial LEDs are somewhat better than halogen, but inferior to CFL and HID and a good way off T5 fluorescent tubes. That said, there's some good work being done in labs to improve the efficiency, so it's likely to improve over the next few years (it's already made dramatic progress over the last few).

At the same time LEDs tend to have very poor color rendering and very high color temperatures (typically 10,000 K). There are 'warm white' and 'neutral white' LEDs available from high end manufacturers, but you pay a big price in terms of efficiency compared to 'cool white'. (But the 'warmer' LEDs tend to have much better color rendering, but still not as good as a CFL).

However, where colored light is required LEDs are the perfect choice e.g. signage and signalling; without the need for filtering, the efficiency is better than any other light source.
 
Originally posted by: Linux23
Are sulfur lights yellowish in color?
Click to expand...

No that would be sodium lamps. Sulfur (electrodeless) lamps have a very high quality, full spectrum output and are fully dimmable with no penalty in tint or lifetime. They are not very portable, however.

Text
 
Originally posted by: Rubycon
Originally posted by: Linux23
Are sulfur lights yellowish in color?
Click to expand...

No that would be sodium lamps. Sulfur (electrodeless) lamps have a very high quality, full spectrum output and are fully dimmable with no penalty in tint or lifetime. They are not very portable, however.

Text
Click to expand...

Cool. That sounds very similar to this type of lamp Panasonic LIFI
 
Just joined, great discussion! My observations:

1. LED Rope Light - I was SO disappointed, the flicker was SO obvious. I presume they have a small rectifier bridge to make it pulsing DC, -or- (shudder) they turn on some LEDs on the upward cycle, and others on the downward!!

1a. LED "strawberry" bulb shape Christmas tree string - totally white, big nicely shaped (like cut glass) "fruit" to break up the light - I do NOT observe flicker in these.

1b. Suspicion: The rope light is bad - flickers on half wave? while the strawberries - which do have a "lump" in their cord near the plug - are on a full wave bridge - especially since I can trace the wiring, and the strawberry string is just 25 LEDs in series. (I question if there is any current limiting in the "lump" or just a bridge).

2. Compact Fluorescents - In our house I find them being replaced by incandescents. BUT if they truly do NOT flicker, then it is the "color temperature" that is too glaring, so looking for a "warm" CF, but of course that means more $$

3. LED Lights. I have bought several different kinds. The easiest way to see if they flicker is to have a way to put a light bulb socket on a cord, either by having a prong-socket converter on an extension cord, or a trouble light kind of thing, etc. THEN just wave it rapidly, or even swing it -- perhaps in front of a mirror, or have another observe -- to see what kind of flicker is observable.

I do NOT see any of the bulbs flickering. Just the LED Rope Light.

4. Power consumption. I bought a new "Kill-a-watt EZ", which lets you put in your power costs (mine is 9.4 cents/KWH "Delivered" -- seems to be 6.4 cents "raw" the rest going to "delivery", "Transmission", taxes, etc). When put on a 13W CF, it shows - 13 watts. and appropriate amps. When put on the LED Rope Light or the Strawberry light string, NOTHING shows up, meaning they are SO low power they are below the x.xx amperage scale. I am hoping if left on, they would still accumulate $$ just more slowly.

5. I'm not skillful enough - or maybe just not brave enough -- to try to implement an "invention" -- a box that you plug into the AC outlet, and converts the 50 or in my case 60 cycle to say 30KHz "AC". You then have a receptacle that you plug various LED lights into, and then adjust the frequency and duty cycle (aka PWM) for optimum, possibly even using PWM for dimming. I am hoping this allows me to use cheap rope lights, with a not too expensive hack to remove the flicker.

P.S. The rope light does not have a "bulge" in the cord, and seems to be made compatible with the bulb-type.

P.P.S. if you do buy bulb-type rope lights, may I suggest the slightly more expensive option of multiple 2 foot strands, rather than say one big 10 footer. Reason: My 10 footer eventually had burn out, and had to be tossed long before all bulbs had burned out. It burned out in "sections" implying that the bulbs were maybe 6V so they do 20 in series then another 20 in series etc.,


I just had an idea: I will buy a small self-contained (aka "surface mount") electrical box, and put some switches on it. I will embed it in an extension cord - i.e. cord---box----cord - then I will have 4 switches - that cut DIODES in and out on each leg of the AC, in each direction --|<|-- and --|>|-- so I can try seeing the effect on various LED devices.

Those that have full wave rectifiers, i.e. change AC into pulsing DC, should blink more obviously because they will light on only half the cycle. Those that simply string LEDs together, should show which ones are going which way (+ to - or - to +) etc.

P.S. PLEASE don't "block quote" this - it will waste space - just maybe say "re: Rope Light flicker" or whatever you wish to comment on. (sorry, it annoys me to skip over long quotes - which repeat what is right there on the page up a little) ;-)
 
Adding some more thoughts, I didn't want too big an item.

1. I bought some cheapie solar powered garden LED lights. They light up a little, but not much and not for long.

A friend bought some. His light up a lot and last a long time.

Umm, heh, he spent MUCH more - he has I think at least 8 if not 16 of those little solar cell strips on HIS, while I have 4.

HIS has 2 AA batteries, mine has one.

You get what you pay for. I could maybe make one good one out of every 2, haha.

2. I LOVE LED flashlights - just because it is neat technology. I PARTICULARLY love the single AAA ones, because "you can't run a 3-3.3V LED off of 1.5V" -- thus it has a small boost circuit, how cool! I took one apart and it was very nice - tiny coil, transistor, resistors, etc.

3. I have built a "joule thief" in various variation - this is the small 1-transistor, 1-home made coil, 1-resistor thing that lets a blue or white LED (meaning around 3V) run off of - well -- I've had a battery under a half a volt - that still ran fine. It is an oscillator that charges a coil, the charge of which tells the transistor to shut off, and the collapsing magnetic field makes a "spike" that is sufficiently powerful to run the LED - wish I had a good oscilloscope to find (1) the frequency (could be 200Hz or 200MHz, I have no idea); and (2) how high the voltage spike is.

I made a slight variation in which I "pull down" the base of the transistor with a photodiode, so it lights up only in the dark, haha -- after all don't most "jewel theives" work in the dark?

3. LEDs: If you are doing anything on your own from scratch, be aware that -- to me -- for INEXPLICABLE reasons, they always do a "lens" on the LED, when in fact if you sawed or sanded the lens off, it would be a better "bulb", in that it diffuses the light. I bought some high power 10mm LEDs to work on some lighting projects... TBD.

4. I like playing with "PICs", those little computers on a chip, in that you could take a little $8 "desk lamp" (which has 3 x AAA in the base and a power switch) and runs 14 LEDs, and change the power switch to a pushbutton and control on, off, brightness, and auto-shut-off with just say $2 more of components.
 
The experiment of seeing a 7-segment stop flickering at 30Hz was great!

I just wanted to add - and sorry if someone else did, haven't read the entire forum yet - that the edge of your vision is much more aware of flicker than the center.

I assume this is because the edges are sensitive to motion, i.e. notice an animal or foe coming into your field of view.

If you notice something flickering - say at 60 or 120Hz, it will stop typically if you look right at it, it will not flicker.
 
Help! I can't stop replying!!! haha

Speaking of moving a LED array around, I have a bunch of "blinkies" from 2Dkits.com, and especially the RGB, but even the bi-color, and somewhat the single color - make very interesting patterns in the dark when waved around.

As a matter of fact, some day, I plan to figure how to make my digital camera "stay open" (used to be called "Bulb") and then, in a very dark room, wave the blinkies around and see what sort of image I can capture.

There are some web sites which deal with "art" or "drawing" using lights in a dark room and an open camera.
 
Originally posted by: eldorado99
Originally posted by: bobsmith1492
Originally posted by: eldorado99
LEDs run on DC, so no.
Click to expand...

LEDs CAN run on AC but they'll just be on somewhat <50% of the time (half-wave rectified or none) so they would indeed flicker. Full-wave rectified they'd be on more of the time but still pulsating. Add a cap to help filter and there would be no flicker.

It depends on how they're hooked up. If it's a commercial LED bulb for lighting it shouldn't flicker.

Also most modern fluorescent bulbs have electronic ballasts so there is no 60Hz flicker to see (it's more like 30KHz).
Click to expand...

You know more than I do, sir.
Click to expand...

And he doesn't pretend to know more ...
 
Originally posted by: eldorado99
LEDs run on DC, so no.
Click to expand...

LED's are primarily DC devices, but there are several reasons they are often run in an AC mode.

Within their ratings, they output more light and dissipate more power with greater current. A pulse width modulated square wave can be used to obtain greater brightness by pulsing them at a higher current than by operating them at a lower steady state current.

This technique can be used to obtain either greater brightness from a given LED or equal brightess using less power and typically lengthening the life of the LED. It can also be used to vary the apparent brightness of an LED.

2. When many LED's are part of the same display, the same technique can be used to save power by multiplexing the LED's so that only one or several of many devices are on at any given moment. If the pulses are sufficiently close together, the eye integrates the overall image and the whole display appears to be on at the same time.

It's the same reason you see a picture on your monitor when in fact, only one pixel is illuminated at any given instant.
 
I am working on -- well, at this point - TALKING ABOUT - a project to make a box.

The box will prevent flicker of 110V LED products, such as ROPE LIGHTS, Christmas BULBS etc.

A power cord will go into an "outlet box" (I've done this before - you get a surface mount outlet box, and just put an outlet plate cover on it with a dual outlet. I have done this where the one outlet position is a switch; another the first position was occupied by a "dropping load" light blub (for a UV EPROM eraser).

Anyway, the circuit will convert the AC to DC, then using MOSFET transistors, convert the DC back to AC, but at a high frequency.

I'll breadboard it first, with a couple of "pots" ("Knobs to turn") so I can change (1) the duty cycle; and (2) the frequency.

With this I'll be able to see at what freq the flicker goes away, and at what combination makes it seem the brightest. Then I can read the pots, and substitute fixed resistors, or even leave them for further tweaking.

I'm figuring I could make it perhaps 100W, and fused, but use regular outlets; so if I put something inappropriate in it, it shouldn't be too bad....

Thoughts?
 
Wow, I thought this was an old thread.



Flickering is subjective. Some people will complain about a 60Hz monitor flickering. Others have no idea that monitors even flicker.

You could probably solve the flickering with a few capacitors, or at least reduce it.

LEDs also can use pulse width modulation, which I think is for brightness control. I've seen quite a few LED brake lights on cars which flicker, and as far as I know, they have access to an easy source of DC.
I have no idea what a car's alternator puts out, but the battery (and its charger of course) are DC.

 
You must log in or register to reply here.

TRENDING THREADS

Back
Top