white led

[audioguru]

I don't use power supplies without a transformer.
My mains voltage is half the voltage of yours so I don't know high voltage parts.

[pier]

Hi Ag

      I am still searching , I've asked a shop owner for the same and he said he will let me know in 2 days . By the way AG hows 13003 for the small one ?

[audioguru]

I am in Canada. I use American parts. I have never seen parts from ST Micro.

The ST13003 is big but will work fine as the smaller high voltage transistor.

[pier]

Hi AG

        The circuit works fine now, but no brightness . Before with high wattage resistor it was really bright . But now it is less . I forgot to tell you AG that I am using 49 white leds . What could be the reason ?

Thanks
Pier.

[audioguru]

Post the DC voltages at various parts of the operating circuit without electrocuting yourself. The voltage across the 27 ohm resistor should be about 0.65V then the LED current is 0.65/27= 24ma which is a little less than their max allowed continuous current of 30mA.

Check which pin is which on the transistors from their pictures in their datasheets.

The circuit should work nearly the same with 1 LED or with up to about 87 white LEDs in series.

[pier]

Hi AG
        I apologise for electrocuting myself :-X . I measured the voltages at different parts of the circuit :

across 27 ohms resistor : 0 volts (no voltage)
at the base of 13003 with respect to gnd : 14.25 volts
across the 10k resistor : 0.59 volts
at the collector of 13005 with respect to gnd : 152.22 volts

All this voltages was measured when the input voltage was 300 volts DC.
No components are heating up .

Thanks
pier

[audioguru]

Hi Pier,
The base-emitter of the ST13003 is supposed to be a forward biased diode with a max voltage of 1V. You measure 13.7V so I think you connected its base and emitter pins reversed, causing the base-emitter to be reverse-biased and acting like a 13.7V zener diode.

[pier]

Hi AG
        Ur great!!!!Thanks for pointing out my mistake . The collector and emitter was in reverse . Now it is very bright and i have a little doubt will it blow off !!! Where do u measure its current AG . I am in front of it now .

Thanks
Pier

[audioguru]

Hi Pier,
Good, you fixed it!

The LED current flows through the 27 ohms resistor. Ohm's Law says that the current is the voltage across the resistor divided by its resistance value.

If the 27 ohm resistor has 0.65V across it then the current in the LEDs is 24mA.
The absolute max continuous current in most ordinary LEDs is 30mA.

[pier]

Hi AG

        There is no voltage across 27 ohms resistor when i measured it ? what could be the reason ?

Thanks
Pier

[audioguru]

There is no voltage across 27 ohms resistor when i measured it ? what could be the reason ?

Maybe the pins on the ST13005 transistor are also connected backwards.
This is what the voltages should be:

[pier]

Hi AG
        I am such a fool, I forgot to connect the 3904 in reverse when i corrected the 13003 properly . Now its fine . The voltage across the 27ohms is 0.56 volts. I dont understand AG how are u able to find faults without seeing the circuit I connected . Thanks a lot AG . as i said is 0.56volts across 27e ok  ?

Thanks
Pier

[audioguru]

I think 0.56V across the 27 ohms resistor is fine.Then the LED current is 20.7mA.

You could replace the 27 ohms resistor with 22 ohms then the current would be 25.5mA and the LEDs would be a little brighter.

[pier]

Hi AG
        I connected 22 ohms . Now the voltage across is 0.58 . How about reducing the resistance little more ?

Thanks
Pier.

[audioguru]

I don't know how good or how poor is the current regulation. When the rectified voltage is as high as 366VDC then you don't want the current to be higher than the max allowed for the LEDs which is 30mA.

[pier]

Hi AG
        Its right as u said . The voltage across the 22ohms resistor is 0.6volts when the mains voltage reaches 345Vdc . The 13005 is getting very hot and i 've put a bigger heatsink now . I will put it under testing and let u know the result after 1 or 2 days . Thanks a lot for ur co-opeation in helping my project AG . :)

Thanks
Pier

[audioguru]

The current regulation is pretty good. From 300V to 345V is an increase in voltage of 15% but the current increased from 26.4ma to only 27.3mA which is an increase of only 3.4%.

EDIT:
The current regultion is even better than that because the actual circuit voltage is nearly doubled.

[pier]

Hi AG
        The circuit was on test for 3 days continiously and working fine . People are asking me how much power it consumes from the mains . Bulbs consume 60Watts as written on it when we purchase it here . Do we have to connect an Ampere meter in series with the live mains i.e.230/250 volts AC to measure its current and calculate its power ?

Thanks
Pier.

[audioguru]

Hi Pier,
Since you have a full-wave bridge rectifier and 60uF of smoothing capacitors then the DC is without much ripple and is pure DC. You measured the current (calculated with Ohm's Law by the voltage across the 22 ohms resistor) at two voltages so just multiply the voltage times the current to calculate the total power:
300VDC x 26.4mA= 7.9W.
345VDC x 27.3mA= 9.4W.

You could measure the voltage across the LEDs and use it to multiply the current to calculate how much power is used only by the LEDs.

[pier]

Hi AG
        I made a board of 72 X 24 mm in size and the same size for the heatsink with a thickness of 3mm . Would this be a sufficient heatsink for a continious operation . I want to isolate the transistor from the heat sink and screw it up (Heat sink)on the roof of my room .

Thanks
Pier .

[audioguru]

We purchase extruded aluminum heatsinks for power transistors. The heatsinks have many fins for a high surface area. They are rated in their amount of degrees C temperature rise per Watt of power dissipated in the transistor.

[pier]

Hi AG
        Now suppose i have only 21 volts dc at the o/p of the power supply and i have to connect led's in series . Assuming the forward voltage is 3.5volts, can i connect 6 in series so that the total forward voltage reaches 21 volts dc ? I dont know the exactly how to calculate the number of led's to connect in the circuit . Can u please help me in this ?

Regards
Pier.

[audioguru]

Hi Pier,
you cannot be certain the forward voltage of each LED is 3.5V since each one is different. They might be 3.9V then 6 need 23.4V.
also, the transistors in the current regulator need at least 1V to work. So use 24V or 25V.

You don't need a high voltage darlington NPN power transistor and you don't need two resistors in series to increase their voltage rating with a low supply voltage. This circuit should work:

[pier]

Hi AG
        Thanks for the circuit. Actually I am not building the circuit for 6 led's but i wanted to know the calculations for deciding the number of led's used for certain voltage available because i read somewhere in the web that u can use leds in series provided the voltage across the led string should not exceed 0.9 X Vcc volts.So As i asked in the topic before as per my assumation i took 3.5volts per led and voltage available is only 21 volts , 6 led's take the entire voltage(21 volts), so i was confused if it was right to do so ????The other circuit u gave me for 49 leds is still working excellent on my roof top :D :D . 

Regards
Pier.

[pier]

Hi AG
        I removed the led light from the roof top to put a better heat sink and when i touched the track side i got a good shock, The capacitors were still charged . Should i put and resistor across the capacitor bank so that they get discharged as soon as it is put off ? If so which would u suggest the value to be and its wattage ?

thanks & regards.