I bought some RGB power leds, o.a. a Cree XML. The spec/description at the (Chinese) site mentioned forward voltage of 3.3-3.9V. But when driving the led at that voltage I measure much less current than expected. ~0.150mA @ 3.9V instead of ~1A for the Cree led, and something similar for other leds with a max of 700mA. I need to go above 5V to get the full max current.

Do I misunderstand the meaning of forward voltage? Or is it happening more often that leds are supplied with higher voltage drops (than described).

I planned to use a 5V power supply and a current regulator (mosfet, resistor and transistor feedback) which requires and extra voltage of ~1.2V. This does not fit if forward voltage is that high.

  • 2
    \$\begingroup\$ My answer to electronics.stackexchange.com/questions/353955/… discusses binning of LEDs to achieve consistency between samples. It won't explain the very high Vf but may help in some way. You should really test the LEDs using a current source rather than a voltage source. \$\endgroup\$ – Transistor Feb 19 '18 at 23:15
  • \$\begingroup\$ Watch out for high-powered LED's. The 10 mm dome type on a ceramic plate often has 2 LED's in series, and another pair parallel to the first. Turn-on voltage is about 6 to 7 volts. They can do the same with RGB LED's except they have six connections instead of two. At full power they MUST be mounted to a large heat-sink with thermal epoxy. \$\endgroup\$ – Sparky256 Feb 19 '18 at 23:21
  • 1
    \$\begingroup\$ How did you run these at the right voltage? With a variable power supply? Or did you use a current limiting resistor and measure the voltage across the LED? \$\endgroup\$ – jonk Feb 19 '18 at 23:31
  • 3
    \$\begingroup\$ If it's a chinese site then there's a decent chance the specs are simply wrong. There's also a good chance the part is not the part you think it is. \$\endgroup\$ – user253751 Feb 20 '18 at 1:04
  • \$\begingroup\$ I measured with voltage supply, but using limiting resistor, measuring voltage across the led and across the resistor (to get the current). I think I suffer from wrong specs at aliexpress. \$\endgroup\$ – pfp.meijers Feb 21 '18 at 6:54

The Cree spec shows 2.9 - 3.5 V for 700 mA

enter image description here

You should be able to engineer a current controller that works at 5 V without any problems.
Perhaps something simple like this, though it will have a thermal variation of current which can be reduce by putting the NPN close the LED.


simulate this circuit – Schematic created using CircuitLab

| improve this answer | |
  • \$\begingroup\$ Any suggestion for a current controller - must be very cheap (<$0.20) because I need to make many? \$\endgroup\$ – pfp.meijers Feb 21 '18 at 6:56
  • \$\begingroup\$ Added to the answer ....but <$0.20 seems a bit aggressive. \$\endgroup\$ – Jack Creasey Feb 21 '18 at 15:54
  • \$\begingroup\$ Yes it is ;-) Found a more-or-less similar schema on a blog. But that did not use a resistor R3. R3 will increase the voltage drop over R1, isn't it? \$\endgroup\$ – pfp.meijers Feb 22 '18 at 17:49
  • \$\begingroup\$ R3 has no real effect but is simply a safety limit on base current. The base current is only uA. \$\endgroup\$ – Jack Creasey Feb 22 '18 at 18:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.