Where's the inductor?

I did read of mains LED drivers which first rectify, then switch in various series combinations of LED to follow the voltage. You need current limiting, and it's less efficient than an SMPS, but cheaper.

Here by the grace of google...

Cheers

--
Clive

Maybe. But the LED appears to be one of those single-assembly 32-volt types. Here's a pic:

--
 Thanks, 
    - Win

Well, I don't know anything about how these things are made, but couldn't that comprise, for example, ten rows of series LEDs?

I guess one way to tell is to sweep your eyes across one. 100Hz (or

120Hz for you) flicker should be easy to see as dot-dash patterns, but an SMPS will look like a continuous line.

Cheers

--
Clive

A photodiode and scope should tell the story. It's probably smart-ish PWM through a mosfet.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Chopping a high voltage CCR is gaining popularity:

I don't know why you think this is the same as the single voltage devices. I have some of those and this does not look like them. They are laid out with N rows of 10 LEDs with ~1 W/LED and ~10 W/string. Of course they assume the current will be shared equally by the N strings and the 10 LEDs in each string will have the same voltage. In reality these are not remotely matched and likely are seconds from the LED factory.

This module has 10 x 14 LEDs with an extra 8 LEDs scattered around the edges. The LEDs seem to be the same between different wattage units. The single voltage devices have different numbers of LED strings. What is different in these modules is the number of chip on one end. I suspect these are driver chips.

I wonder if they will come out with a version that had dimming capability? That's the one shortcoming that I'm trying to work around now.

--

Rick C

With battery-operated LED Christmas lights using little SMT LEDs encased in acrylic their "Smart Driver" is they have some guy bin the LEDs according to IV characteristics, then run them all in parallel off one or two CR2035s. One small-value series resistor for the whole string.

Some cheap LED things connect a battery to an LED with nothing between. LEDs can be run constant-voltage if the voltage is right, so an ac_line-bridge-mosfet-LED string should work if the fet is switched intelligently.

I have some small-screw-base LEDs, candelabra or something, where it's hard to imagine there's much if any electronics in the tiny screw base. I should dremel one open and see what's inside.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Hi,

I almost asked this same question after almost buying one of those LED's for testing for an in-house garden light, I haven't bought one yet, still using the "pink" (red/blue mix) LED's.

When trying to figure it out I saw this article:

"However, attempts have been made to develop "true" AC-LEDs at the assembly or packaged device level.. At the forefront of these developments are Lynk Labs, Seoul Semiconductor and III-N Technology.

The technology developed by Seoul Semiconductor and separately by III-N Technology uses the Christmas tree approach at the die level. The AC LED device is actually made up of two strings of series-connected die, connected in different directions; one string is illuminated during the positive half of the AC cycle, the other during the negative half. The strings are alternately energized and de-energized at the 50/60Hz frequency of the AC mains power source, and thus the LED always appears to be energized. The technology developed by Seoul and III-N specifically relates to LED devices designed for high-voltage 50/60 Hz mains AC power."

I don't know if that is what they are doing though.

cheers, Jamie

** The light output is bound to have lots of flicker.

The product is not magic - it's quite primitive.

.... Phil

I used a TL431 wired as a 2.5 volt "zener" in the cathode of each segment of a red common-cathode 7 segment display, anodes driven directly from 5 volt logic outputs. Works great; it's just the right intensity and I don't see any difference in illumination between segments or any problems over temperature.

Are these new-fangled LED lights compatible with triac light dimmers or modified square-wave inverters I wonder ?

boB

Den fredag den 17. marts 2017 kl. 13.54.17 UTC+1 skrev Winfield Hill:

These devices have one thing over the older LED only modules, they are actually designed to get rid of some heat! I have one of the LED only modules and the back plate is steel, not aluminum. I've measured the

too hot for LEDs. That was with the device mounted to a quarter inch thick aluminum plate over a foot in diameter. It takes several minutes for the plate next to the LED to heat up, but the LEDs get hot in seconds. Clearly the LED module is not designed to conduct heat out to the heat sink.

--

Rick C

Yikes, linear current regulators on heat sinks!

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Hi,

Here is a great video on these LED modules:

The IC used is the BP5132 apparently (mains dimmable linear LED driver):

The IC's are put in parallel to increase the current handling ability.

There is a bridge rectifier that feeds a string of 74 LED's, and the rectified voltage causes a lot of flicker which along with the extra dissipation in the BP5132 are the only drawbacks.

There are a lot of benefits to this type of circuit though, especially if you have a DC supply to get rid of the flicker and adjust the voltage to minimize the BP5132 losses. Maybe using an active PFC boost topology and putting a couple of these LED modules in series, driving 74*2 or more LED's in series.

In a follow-up video he added a ~10uF/450V cap across the LED's and reduced the apparent light flicker quite a bit.

cheers, Jamie

I'm a bit confused. With a full wave bridge, the flicker would be 120 Hz. How many people can see 120 Hz? What am I missing? Who would ever want to buy a lamp that has visible flicker?

--

Rick C

You can see that easily if the light source is moving, or if you are moving your eyes.

Next time you're in traffic behind a car in the dark, sweep your eyes up and down. If it's a newish car with LED lights, you'll probably see a line of red dashes when the tail lights are on, and a solid line when the brake lights are on. They use the same LEDs, but the tail lights are PWMed for lower intensity. You can see the effect with older mains frequency fluorescent lamps too.

(That works here in the UK, it's possible US vehicle lights are done differently)

I once diagnosed a fault with some equipment to the amazement of the technician using this. The reservoir capacitor in the non-switcher PSU had failed, and the power LED was flickering at 100Hz.

Cheers

--
Clive

add a bit of circuitry to to keep voltage just about what is needed