– Clip 1 Introduction – Clip 2 Course Description – Clip 3 Test Parameters – Clip 4 Test Equipment – Clip 5 Overview – Clip 6 Single LED test – Clip 7 LED array Transformer – Clip 8 Voltage increase LED to Transformer – Clip 9 Joule ringer – Clip 10 Optimising Joule Ringer – Clip 11 Calculation 1 – Clip 12 Summary – Clip 13 End of Part 6 

Clip 1

Clip 1

In this series I will investigate the power efficiency of driving LED’s. I focus on two principle. 1. the standard driving circuit via a transformer and 2. the Joule Ringer or Flyback transformer principle. I measure the light intensity and compare it to the power requirement. I will use 20 LED’s in an array.

Clip 2

Clip2

High Frequency Transformer and Flyback mode

– Measure LED power requirement to light luminosity levels
– Apply traditional high frequency transformer principle
– Apply flyback transformer principle
– Compare the power efficiency of both principles to light luminosity levels

Clip 3

Clip 3

Test Parameters

1. One Amplifier Module with 2N3055 Transistor
2. Single LED 3 Volt and 20mA, 14,000mCD
3. Array with 20 LED’s
4. Two resitor in series, 100KOhm and 1MOhm
5. Metglas/Nanocrystal core with 1:10 winding ratio

Clip 4

Clip 4

Test Equipment
Tools

– EA, PS 2000 B, 84 V 5 A, Power supply
– Agilent/Keysight InfiniVision 3000 Series oscilloscope DSOX3014A
– External 19 Inch Monitor for Oscilloscope
– Gossen Mastersix light meter
– Agilent 34450 A 5.5 digit Benchtop DMM

Clip 5

Clip 5

 In this series we focus on transformer efficiency and light emitting efficiency. The brightness is key. I use the Gossen Mastersix. This is a professional light meter use professional photographers around the world. I use special high brightness led with day light color. They are 100 mm in diameter and used as spotlights. We have also the Metglas/ Nanocrystal core here. I tested in this configuration other cores and non of them come close to this core. I therefore focus on this core only. I will use the core in transformer mode and also in flyback mode. The common description would be either an LED driver circuit or a  high voltage Flyback transformer, self oscillating as in a Joule Ringer. The power draw will be very low so that the power or current indication of the power supply is not reliable anymore but either showing zero or a high value. I use the 5.5 digit Agilent benchtop DMM.

Clip 6

Clip 6

For our first test we benchmark one LED. WE measure the brightness at a given power level and compare that power level with the brightness via the transformer. I use a 10 Ohm resistor so to reduce the maximum current draw at 3 Volt to 300mA, The LED itself has a voltage drop of 0.7 Volt. The LED is power rated at 20mA. I measure photo metric for Lux the brightness. I start and measure a current of 13.4mA. I measure 2000 LUX. The LED need to warm up and is increasing in brightness and is drawing more current as well. Our benchmark is now 14mA and 200 LUX brightness.

Clip 7

Clip 7

 The setup is very simple. WE have the primary coil connected to the power transformer. The secondary has 10 times more winding’s and will drive the LEDs. Ratio 1:10. The LED array required minimum 50 Volt but best would be around 70 Volt. There is no resistor in series connected. We set the voltage to 6 Volt on the power supply and the frequency is  260 kHz. WE draw 23 mA but we cannot see much light on the led array. I change frequency and observe the led’s

Clip 8

Clip 8

I changed now the voltage to 10 Volt and 7 kHz. Now it is nice illuminated. A strong light but not blinding at 0.76 mA. The Brightness is not more than 420 LUX. WE go up in the frequency to see if we find a better value. The current probe does not show more than 4-5 mA. At 30 KHz I am at 1 mA. The voltage spikes are now reaching over 90 volt and the voltage protection glob bulb is flashing. I lost one LED in the array. That did happen a couple of times and I had to replace them. the voltage spikes are very high and some led’s quite short after the test. Here my led came back to live. At 300 kHz there is now power in the led’s anymore. I go down in the frequency and light is coming up again. My ration between primary and secondary coil is not high enough. the only solution is to increase the voltage on the power supply. I increase to 10.8 volt and measure high brightness at 34 mA. That is result for the LED driver and move on now to the Joule Ringer circuit.

Clip 9

Clip 9

I use in the flyback configuration resistors. That allows you to compensate for coil variations. I also drive the low winding coil and use the high winding coils the driver coil. Resistance it over 1 MOhm. I also use the back EMF output and do not go direct to the collector with the lED array. Technically it does not matter because we use LED’s as load so the direct current from the power source is so low that it will not be able to provide any assistance to illuminate the led array. Here is is purely done via back EMF from the coil.

Clip 10

Clip 10

 I have now adjusted the power supply to 10 Volt.  I draw 10.9 mA. I can adjust with the resistor the frequency response of the coil and the brightness of the led array. Our frequency was much higher on the driver but here we are very low. The frequency depends on voltage and resistance. the lower the voltage the higher the frequency. The higher the resistance the higher the frequency. Let us measure the brightness at 10 Volt and 11 mA. It is not that bright so I try to increase it. I reach again the end of line with my configuration. I have therefore to increase the power supply voltage. At 11 Volt it draws 11.58 mA. Not much more. At 11.9 Volt and 11.6 mA I reach the benchmark value of 200 LUX. See all the details on the board.

Clip 11

Clip 11

We integrate our measurements into the formula

Single LED, DC powered with 10 Ohm resistor for 2000 LX at
3 Volt 14 mA = 0.042 Watt
Times 20 for the array = 0.84 Watt

LED array, Transformer mode at 34 KHz, 10.8 Volt and 34.319 mA = 0.37 Watt
Efficiency is 44% power requirement over traditional DC energizing

LED array in Flyback mode, at 588 Hz, 11.9 Volt and 11.612 mA = 0.138 Watt

Efficiency is 16% power requirement over traditional DC energizing

Clip 12

Clip 12

 Summary

This high efficient LED’s draw very little power and produce high luminosity with almost
no heating. The traditional energizing via DC is already very efficient. This can be
increased with an LED driver circuit and transformer. Where it starts to outperform the
driver is when we use it in Flyback mode or more common in a Joule Ringer circuit.
Here there are many options available where I use the opposite then usual driving the
primary type coil and get the secondary connected to the base of the transistor.
Of course of that I have to use very high resistance of over 1 MOhm to reduce the
voltage to a level the base of the transistor can handle. The traditional Joule Ringer
and Flyback transformer operates the opposite. I also use the diode isolated back
EMF output to the positive polarity of the LED array. The drawback is that I have to
increase the voltage from the power source to overcome the winding ratio deficiency.
The current displayed on the power supply does not show accurate readings anymore.
I will build new coils with this Metglas/ Nanocrystal core and will test again and measure
towards a Solar cell to harness this energy.

Clip 13

Clip 13

The End Part 6

High Frequency Transformer and Flyback mode

Clip 14