An LED that runs from 120V AC (o0r 240VAC) is an inciting item in the solid-state low voltage DC universe. No driver, no transformer. Just plug it into the house current and let ‘er rip! Available in 2W 65l (3000K) / 80l (6300K) and 4W 150l (3000K) to 195l (6300K) versions, these devices are fairly powerful and generate good efficacy (between 32 and 40 lumens per watt), not bad, considering there are no other power losses involved from voltage conversion and driver/conditioning. This is roughly the performance of a 50l/w 24VDC CC device operated from a balanced and high efficiency driver/power supply. Not bad.
Color rendering index is 70. At the low end of the good range, and is caused by a weakness in the red end of the spectral distribution. Further, the peak power is virtually right on top of the 575-600 nm of photopic vision. The result is that the 3000K light looks a bit “White” and less “warm” than one might expect.
To put these to the test, I retrofitted a pendant (in my office), which was originally illuminated with 4 – 75 W Halogen lamps (300 watts for the fixture). I used a total of 8 – 4W PCB Mounted Acriche LEDs mounted to 5W radial heat sinks (32 watts). One set of (4) LEDs are aimed up at the ceiling with customized diffuse parabolic reflectors, the other (4) were aimed down into the glass pyramids. Between the two is a metal plate, painted white. The LEDs were then simply wired into the original luminaires wiring, without even taking it down from the ceiling. The original lamps were housed on glass diffusers, within the glass pyramids and topped by sand etched glass plates. All the remains of this are the pyramids, so the retrofit suffers less diffusion loss.
In application, the lumens generated by the 8 LEDs totals 1,200. The original halogen lamps totaled 4,200. However, the amount of light generated in the space, and on the work surface is only diminished by 17%. Considering the pendant was most always dimmed, the amount of actual light being used has not changed.
The end result is a reduction of power use by 88%, with no change in usable light. On top of this, there is no heat, and the whiteness of the light is a welcome change from the dingy beige of the slightly dimmed halogens and all the glass they were forced to push their lumens through. The color is acceptable for general office lighting use, but not warm or good enough for a dining room or color sensitive application.
Now, let’s talk about alternating current and LEDs, and stroboscopic effect (flicker.) The human eye is able to see fluctuation in light emission below a threshold of around 80hz. That puts the 60hz AC sine wave within the range. Since incandescent lamps produce light from heated elements, this is not an issue. The slight cooling of the filament between cycles is small, so light appears to be continuous. LEDs have no such effect. To combat this, AC LED systems use two sets of LEDs, one to either side of the 60hz sine wave (half on and half off at any time). The result is an improved 120hz cycling. For most, this eliminates any visible flicker. As expected, the Acriche uses two reverse parallel strings of LEDs (to create the 120hz AC effect) and by using a lot of them, creates the high voltage drop.
But, does it flicker? No, and yes… At full brightness, when the sine wave is full, the appearance off any flicker is minimal. Using a simple spinning pin wheel test under an incandescent, then a DC LED product, then compared to the AC LED makes the difference readily apparent.
LEDs have a minimum voltage, below which they stop producing light (the Acriche produces just 50% of full light at 85 volts, then drops off rapidly from there.) With each cycle, the voltage drops past this point, causing the LED to shut off for a very small portion of the cycle. As the only source of light in a space, with rapid movement, those sensitive to flicker will likely see this subtle on-off cycling. Those who have not had an issue with flicker of HID, or CFL sources, will not see any more or less with the Acriche. However, when operated on dimmers that chop the sine wave off, the larger gaps in the wave form, coupled with the threshold voltage drop, flicker becomes much more noticeable. This is most visible when something moves quickly in the room, or when the head is turned quickly.
The dimmer effect cannot be readily solved using conventional dimmers. The solution will require more than sine wave chopping, or PWM control. This is an instant where Pulse Amplitude Modulation at a higher frequency would be a significant improvement both dimmed and undimmed. However, this defeats the cost driver behind this LED package.
We encounter a similar issue in sports lighting, where HID flicker is a serious problem in fast moving activities. The use of a 2 or 3 phase lighting system resolves this. If one were to utilize a three phase input (120/208 for example – which has three sine waves 120 degrees apart) to create a 120V x 120V two phase (center neutral) circuit to the luminaire, and connect half the LEDs on the A phase, and the other on the B phase, there would be 6 “ON” events per 60hz “cycle”, or 360 per second, vs. 120, virtually eliminating flicker, even if this were dimmed. Using all three phases would result in 540 on cycles per second, virtually invisible to human vision.
To resolve this in the test installation, the conventional dimmer was replaced with a wire wound 50W, 500K Ohm Rheostat. This simply reduces the voltage to the LEDs without effecting the shape of the sine wave. The result is no increase in flicker when reduced light levels are set. However, the measured loss through this device means that dimming does not reduce energy consumption. At 50% light level, the watts consumed remains roughly the same energy as the full brightness load. Still a significant saving over the halogen lamps.
At full brightness, the LEDs and heat sinks are just warm, measured at just 43 degrees C. Considering these devices are rated to operate at up to 85C (125C tj), it appears these will live a long life in this application.
Recommendation rating = 7 of 10. Very good lumen output, consistency between devices (color and lumens) were good, simplicity of application a real plus. Flicker of 60hz operation problematic for some, aggravated by dimming. This would be an outstanding source for outdoor lighting, and in retrofit accent lighting in spaces with higher ambient light levels, which will mask flicker issues. Would not recommend this to anyone who has serious issues with flicker, such as those who have issues with existing fluorescent or HID sources, or VDT monitors set at under 80hz. I would not recommend using these devices on conventional dimmers.
Actual product performance against manufacturer claims = 9 of 10. Product performs within expectations raised by manufacturer data. Only critical note is that the issue of flicker and dimming issues are not addressed in company literature or specifications. Strongly recommend that the use of A-B phase connection is roadway and outdoor lighting be included in product data, as this resolves flicker issues completely.