Taking the Smoke and Mirrors Out of AC LEDS, Part 2

Here's how Seoul Semiconductor's Acriche 2 AC LED technology works for high-bay lighting applications.

These are the questions we're addressing in this two-part series:

1. What ways are there to power an LED (a DC device) from 120 VAC?


2. Why choose one way over the other? What are the functional pros and cons?


3. What's the basic idea behind the Acriche AC LED?


4. What's different about original Acriche?


5. What is Seoul's new and improved AC-LED approach using a floating IC driver?


6. Can we make a cheap, energy-efficient, 200-600 Watt commercial high-bay luminaire using AC LED technology and eliminating the power supply completely?

Part 1 covered the first four questions, and today we'll deal with the final two.

The Acriche 2 AC LED with floating driver IC

Circuits following the pattern of Figures 1 and 2 in part 1 of this series, and the resulting LED voltage/current wave shapes, reflect all Seoul AC LED approaches for the last 7-8 years.

In a circuit of about 36 LEDs, nothing happens until line voltage reaches about 72-75 volts, at which point the current barely starts to conduct at a sub-milliamp level. As the line voltage increases to its peak at about 165 V, the current is then at a maximum depending on the type of LED and series resistor value. Let's assume each LED is 3.2 V at 500 mA. That means all the LED V_fs equal 36 X 3.2 or 115 V. The resistor must make up the difference, so we might put in a 10-Ohm resistor rated at five Watts to drop any excess voltage.

This all works just fine if LEDs always have exactly the same spec (this never happens) and the line voltage never varies one iota from 120 VAC (this never happens either). But with some trial and error, we can find values where everything can work acceptably over line voltage variations, variations from LED to LED, and variations over operating temperatures.

Figure 5

We pointed out earlier that the sine wave distortion in Figures 1 and 2 in part 1 of this series creates power factor and THD problems. If we can get the LED to conduct much earlier, we can get closer to a true sine wave. If we theoretically had only one LED, it would start conducting as soon as we got to about two volts.

Unfortunately, when line voltage got to 165, that single LED, unless we had a 25-Watt power resistor in series, would go to dozens of amps and blow up. So Figure 5 is what the Seoul (or Texas Instruments) floating regulator IC does. First, diode string 1-12 starts to conduct at only about 12 V (good). Sine wave voltage increases and current increases, but before it is excessive, we switch in the second set, 13-24, and current drops back down a bit and starts rising again.

Figure 6

When current again gets too high as line voltage goes even higher, we switch in 25-36. Similarly, progressively, we switch the LEDs out when the line voltage sine wave is on the way down.

So instead of a very narrow and distorted sine wave we constantly switch LEDs and out so they don't have too little or too much current. A sawtooth current wave form results per Figure 6. Rather than actual mechanical switches, "solid state" switch (i.e. MOSFETs) are employed, driven by a simple IC which sends line voltage amplitude and sends turn single to the appropriate MOSFET LED-bypass switch. Patent application 20110279043 explains the concept. This "domino" type switching method has been used in a myriad of series switching circuits for decades.

Figure 7

This circuit obviously results in an improved THD, but it's still not comparable to a conventional LED driver. Figure 7 shows the output LED current in a conventional design using a high-power factor/low-THD power supply. With Figures 5 and 6, there is substantial low-frequency LED ripple current, most severely at 120 Hz but also at the switching frequency, which will be higher than 240 Hz but lower than 960 Hz.

High bay?

The table below compares how the requirements for high-bay lighting are met by LED solutions with a standard, constant-current DC power supply versus those using an AC LED approach, with or without a floating/switching driver.

Requirement	Standard approach with constant-current power supply	AC LED with or without floating/switching driver
1 — Operate 90-277 VAC	yes	no
2 — Use any conceivable combination of series or parallel LED arrangements -- surface mount or COB, and any conceivable LED COB array from any vendor, regardless of specific operating voltage range	yes	no
3 — Constant efficacy & light over full voltage range	yes	no
4 — Fully functional in near-brownout conditions	yes	no
5 — Safety-isolation UL compatibility with all LEDs circuitry	yes	no
6 — Compatible with analog dimming such as with daylight harvesting/occupancy sensing	yes	Difficult
7 — Low output ripple, absence of stroboscopic effects in presence of moving machinery	yes	no
8 — Adjust LED power to match final thermal operating environment or desired specs	yes	no
9 — Automatically scale back power if sudden thermal anomaly	yes	no
10 — Meet stringent THD requirements	yes	no
11 — Compatible with wireless controls	yes	no
12 — Compatible with integral battery backup for emergency lighting mode	yes	no
13 — Short-circuit protection (auto recovery)	yes	no
14 — Luminaire manufacturing cost (material, labor, overhead) for 100-600 Watt units	—	As much as 10-15% less depending on power level

I hope this contribution clarifies how suitable AC LED approaches are to the demanding requirements of high-power, high-bay LED lighting.

Related posts:

• Taking the Smoke and Mirrors Out of AC LEDS, Part 1


• AC Drivers for LEDs