I found a novel gooseneck lamp at the hardware store that clamps on an edge and contains 24 Light Emitting Diodes in what resembles a stainless steel shower head.
I found a novel gooseneck lamp at the hardware store that clamps on an edge and contains 24 Light Emitting Diodes in what resembles a stainless steel shower head.
It provides very great light, but after a minute or so starts flickering like a fireplace and might cause someone to have an epileptic seizure.
Of course if I return it there could be an exchange or a refund, but I looked inside and got interested in possibly fixing it myself.
There is a wall wart power transformer at the end of the cable, and the power is distributed to the lights via about four tiny surface contact resistors, unless they’re capacitors…. I haven’t yet drawn the circuit, but I thought I spotted some cold solder points and after touching them up expected it to be fixed, but it flickered like a gambling table in Las Vegas.
What should I try?
I don’t know Carl, but I’m concerned about your safety.
If it’s defective, I don’t really
know what that means, but it might
not be safe.
I think most wall warts have thermal
fuses, but you never know. If something
should go wrong and the power supply
shorts…
I don’t know how anybody else feels
about this.
I love LED lights, but you will want to
be careful.
Bruce, Now testing 1020 CC again
Safety
I agree with Bruce on this. Exchange it. There is risk of shock and fire if you “repair” the unit, especially if you are not familiar with the circuitry.
That being said, is there a dimmer involved in the gooseneck lamp? I’ve been told dimmers don’t play nice with CFLs or LEDs.
Neil
Inventory
There’s a tiny wall wart AC to DC power brick, probably a switching power supply, an on/off switch in the cable, and the lamp.
One thing I might do which would be safe is to put a volt meter on the output from the wart.
The odd thing is that some lights don’t flicker, and the ones that do flicker at different rates, in groups, like… four flickering at one rate, six flickering at another rate…
If this was a disco lounge it would be great.
LED Theory
This might relate to your observed flicker. Our eyes perceive the brightness of a LED in relation to the peak light output and not the average output. A design trick used to make LEDs appear brighter is to pulse current through them at levels which produce high peak light output but if continuous would destroy the part. The duty cycle of the pulses determines the heating of the LED and by using a small duty cycle of high peak current the brightness appears high but the heating of the LED is kept within tolerance.
Maybe in your lamp the LEDs are pulsed in groups at different times to reduce flicker and individual LED average current while maintaining a perceived high brightness. Something might have gone wrong with the pulsing circuitry.
Neil
Unservicable
The pulsing circuitry would be in the little power block, which is sealed for the rest of eternity and can’t be inspected.
Back to the store.
Was (is) the flicker ~60 Cycles?
Many times they run a bunch of LED’s in series with a diode..
Four Different Happenings
There are 24 total LEDs.
6 of them don’t flicker;
6 0f them flicker slowly… off/on/off/on/off/on;
6 of them have a mid-flicker-rate… bu/bu/bu/bu/bu/bu;
6 of them flicker at a high rate, possibly 60Hz, sort of reminds me of the flicker of old analog cathode ray TV picture tubes. I think TVs were near 30Hz.
Looking at the lights to make these observations caused temporary “snow blindness”. I should use sun glasses.
LED light bulbs
LEDs are not 120 VAC friendly.
LEDs conduct current only in one direction. They turn on when the forward voltage is between 2 and 4 volts (threshold voltage, depending mostly on the color of the LED). “White” LEDs are most commonly produced by phosphor coating a blue LED. An array of LEDs can be connected in series to raise the overall threshold voltag to the sum of the threshold voltage of eah LED>
As the threshold voltage is passed from low to high, an LED transitions from a very high resistance to a very low resistance. Above the threshold, an LED would self destruct unless there is additional external resistance to limit the current. Added external resistance means less overall efficiency because the external resistance will dissipate power.
So, to maximize overall efficiency, the external limiting resistor must be small. To allow that, an electronic circuit is added to chop higher than threshold voltage into little slivers, once per 60 Hz cycle. That way a lower series resistance can be used while maintaining a safe average power to the LED. This is equivalent to the way “switching” power supplies operate.
The “switching” power supply inside an LED light bulb could be designed in an infinite number of ways. The simplest would be to chop the 60 Hz waveform, resulting in 60Hz flicker which is typically not perceived by most people. Another way would be to run an internal oscillator at some higher frequency and then chop that frequency into slivers. That would eliminate flicker, but would be more expensive. Guess which method wins in the market.
The electronic circuit sealed inside the bulb is the single most unreliable part of the bulb. Just like the electronic circuit sealed inside a CFL, the lifetime is dependent on heat. If the bulb is enclosed in a ceiling or wall can vs. a table lamp, it will fail sooner.
Carl describes a bulb with 24 LEDs. If each LED threshold is 4 volts, the total series drop is about 96 volts. With an appropriate, fairly small series resistor this scheme would survive 120 Vrms with high overall efficiency with a very simple electronic circuit, but flicker would be 60 Hz for all LEDs, which is not what Carl described.