Multi (wide) Voltage LED using LM317

,

This above picture is the typical application circuit of LM431, it shows us how to adjust VKA. If we replace R1 with LEDs, LEDs current will be adjusted by R2 and Vref. If  Vref and R2 is constant, LEDs current will be constant too

From  efficiency point, maybe only LM431 is not a good solution because input resistor will share more power dissipation.

Good solution. But the circuit is too complex. Especially the chopper is hard to adjust.

This is the circuit I would normally use to generate a constant current of about 25mA. The input voltage is anything up to 37volts. The required resistor is calculated (roughly) by 1.2 volts divided by the current required. Eg 1.2V/.025A= about 47 ohms Resistor power is  Voltage squared/Resistance (eg 1.2*1.2/47ohms = 30mW)  OR  Current*Current*Resistance The power disipitation in the LM317 is of course the voltage drop across it multiplied by the current flowing through it (eg same current as going through your LED, "current through a series circuit is the same")

NOTE:-47 ohm resistor used

LM317 (small TO-92 packages)  http://uk.farnell.com/jsp/search/browse.jsp;jsessionid=BBVYM0BUEM3HMCQLCIPZK0Q?N=551&Ntk=gensearch_001|gensearch_001&Ntt=lm317|to-92&Ntx=&displaytext=&_requestid=163594

your solution is only for small power LED driver.How about Power LED(350mA current or more ) for resistor will result in lower efficiency

In fact, your solution is a series regulator with discrete parts. The transistor will consume power like the resistance does in my solution. I agree that my solution is suitable for low power led only. For high power led application, I'd like to choose Malcolm's LM317 solution.

      Hello!

Attachments:

I don't think using a series regulator (LDO or some discrete parts) is a good idear for hight power LED(typical 350mA). we can count the power dissipation as Pd=350mA*(V_in-V_led).if we use a Multi (wide) Voltage,V_in-V_led will very large,that means we need a high power dissipation component and more heat-sink and more cost but low efficiency.by the way ,how can we desgn for the constant current control,it's another cost.

If we using switch regulator ,the issue above is not a problem.You will say what about the cost? infact,a switch regulator is expensive than LDO,but you can compare the total cost as the same function you need(CC control,dimming,open/short protect e.g.),what is the result you've got?

But in some special application we can use series regulator for power LED,like the fixed DC input.we can design the V_in-V_led very small for decrease Pd.maybe LM317 is a good choose

I don't think using a series regulator (LDO or some discrete parts) is a good idear for hight power LED(typical 350mA). we can count the power dissipation as Pd=350mA*(V_in-V_led).if we use a Multi (wide) Voltage,V_in-V_led will very large,that means we need a high power dissipation component and more heat-sink and more cost but low efficiency.by the way ,how can we desgn for the constant current control,it's another cost.

If we using switch regulator ,the issue above is not a problem.You will say what about the cost? infact,a switch regulator is expensive than LDO,but you can compare the total cost as the same function you need(CC control,dimming,open/short protect e.g.),what is the result you've got?

But in some special application we can use series regulator for power LED,like the fixed DC input.we can design the V_in-V_led very small for decrease Pd.maybe LM317 is a good choose