BC337 + 1watt LED transistor won't turn completely off

Other Hardware General Electronics
1

I'm surely misunderstanding things or doing it totally wrong, after a night of forum hunting (many fora) i'm still a tad clueless (and i feel most comfortable asking things here)

I have a 4x1watt RGBW chip, for testing i'm only using the blue bead atm henche the picture says "blue led". (start small)

bc337-led-help
bc337-led-help1177×634 67.4 KB

when i turn on the power supply, even when i connect base to ground (through that 1k resistor) the led faintly lights up.

I first hooked up a 10k potentiometer to the base (from the same 3.3v supply that supplies the led) and i can control the brightness of the led with it. it won't fully turn off, it will go up to 150mA (which is fine, i dont need full brightness, less heat is good). When i hook up a 3.3v microcomputer (wemos d1 for testing, i have many so they are expandable) and put pwm on it (going from 0% duty cycle to 100% and back in a loop) it sometimes brightens up and dims as expected (though not going fully off) but it all seems intermittently working/not working/flickering.

To figure out where my logic is going wrong, things i think i understand about the bc337 bjt:

it needs 0.7 volts higher on the base then emitter. the emitter is connected to ground, 0v, so anything above 0.7 volt on the base of the bc337 should turn it on.

The amount of current going into the base controls the amount of current able to flow from collector to emitter. it has a minimal factor of 100, so 3mA should allow 300mA to flow.

Due to the 1 ohm resistor in front of the led, the current at 3.3volts is limited to 300ish mA, so just the bc337 turning fully on and off should be good enough to have the led working.

Many questions, many "not getting it", wish i've had electronics classes at school ages ago :frowning:
Is there anyone who can explain me what i'm missing/not getting, how to do it properly or that i'm just being stupid and should go about it a different way all together ?

2

Hi, @callthedutch
Rather than using paint to draw your schematic.
Can you please post a circuit diagram of your project, including power supplies, use a pen(cil) and paper and show an image of it would be fine.
Please do not use a Fritzy picture

What is your Base to Emitter voltage when you have the pot turned to OFF?

Can you note what voltage you have between gnd and collector of the BJT when you are full on and full off?

Thanks.. Tom.. :smiley: :+1: :coffee: :australia:

Also note pin names and voltages.

3

Is it a BC337 or a BC337-40 like shown in the picture?
Try shorting the base directly to the emitter, without the 1K resistor.
Does it still light dim?

4

circuit(1)

Hope i did this correct...

0.06 when off, 0.9 when on (according to my crappy multimeter)

1.3 when off, 0.3 when on

Is it a BC337 or a BC337-40 like shown in the picture?
Try shorting the base directly to the emitter, without the 1K resistor.
Does it still light dim?

It is a 40, apparently my new order is 25 (i'm using the 40 atm)
It stays lid dimly even when shorting base to emitter

Sorry it took a bit to reply, at work and forgot to bring the contraption with me, so i went to pick it up :slight_smile:

5

I would suspect a BAD transistor. Try another

1 Like
6

What is that, can you post a link to the specs please?

Most high power LEDs require a constant-current (CC) power supply, not constant voltage, otherwise they will go into thermal runaway and burn.

You can't use a transistor to apply PWM to the input or output of most CC power supplies. You may be able to buy a CC power supply with a voltage or PWM input. If not, keep looking!

7

So this was what i was making of it as well, but i felt that was the "easy way out thinking" and i must be misunderstanding things. So i switched it out, moved the test setup onto a breadboard and whatdayaknow, it turns off completely as expected (with the potentiometer, no resistor between pot and base pin). so now to test with resistor and pwm from a microcontroller.

isn't that why there is a resistor "in front" of the LED ?

i don't think i am "pwm'ing" the powersupply, am i ?

8

@callthedutch

One thing I should mention. The forward voltage (Vf) specification for the blue LED is 3.0 to 3.6V. That is the voltage at which it just starts to turn on. If you have the 3.6V one, it will never turn on with a 3.3V supply, you will 5V. Now you have to be careful about the current limiting resistance and wattage.

If you use 5V and you actually have the 3.0 Vf LED then a 1Ω resistor will only limit the current to 2A!

5V-3V = 2V.
2V / 1Ω = 2A
Wattage = 2A x 2V = 4W!

9

it does shine fairly bright on 3.3 volt, but the power supply is adjustable so i could up the voltage to like 3.7v ? 4v ?

The idea behind the lowish voltage was to keep the resistor small (in watts) and heat down.

When i calculate for 4volts, 3.6volt forwarding on the LED, 300mA current i get this:

  • You will need 1 x 1.5 ohm 1/4 watt resistor.
  • The 1.5 ohm resistor is color coded: Brown, Green, Gold, Gold.
  • Each 1.5 ohm resistor consumes 135 milliwatt.
  • Total power consumed by the resistors is 135 milliwatt.
  • Total power consumed by the LEDs is 1080 milliwatt.
  • Total power consumed by the circuit is 1215 milliwatt.
  • Total current drawn by the circuit is 300 milliampere.

Now i don't have 1.5ohm resistors, or half ohm to make 1.5 ohm resistor but doing 2 ohm resistors produces this, which seems good enough (i don't need full brightness)

  • You will need 1 x 2 ohm 1/8 watt resistor.
  • Each 2 ohm resistor consumes 80 milliwatt.
  • Total power consumed by the resistors is 80 milliwatt.
  • Total power consumed by the LEDs is 720 milliwatt.
  • Total power consumed by the circuit is 800 milliwatt.
  • Total current drawn by the circuit is 200 milliampere.

i guess that could work ?

10

I would do 2ohms at 3.6 volts and step it up gradually until the current draw or brightness reach a max level and then decide if you like it there. Assuming you have two 1ohm resistors.

11

What if your LED is actually 3.0V, then what resistor do you use?

12

And what about the red and green. Will you have separate adjustable power sources for each of them?

13

having fun making me think ? :stuck_out_tongue: (i do, thanks man!)

I made the assumption the leds will work within a voltage range of 3.0 to 3.6volt. the plan was to power the blue green and white from 3.3 volt, and the red from a different source on 2.3.

I totally agree this is turning to look all sketchie and all, this was the best i could come up with using my very limited knowledge of electronics (i am more into software then hardware, it's a necessity for projects :slight_smile:

14

You'd be surprised how many "experienced" people get tripped-up with the simple LED current limiter calculation.

Are you using a solderless breadboard? They are not designed for high currents especially the cheap ones. Operation with currents over 200mA could become problematic.

So what is your ultimate goal, to learn electronics or just find an easy way to PWM your 1W LEDs?

1 Like
15

ok so the goal atm is to get this LED to be controllable from a microcomputer, so i can make a wearable light to go onto my girlfriends butterfly wings with fiber wires she made for an upcoming festival (it'll stick to a frame on her back with batteries and all) but for 1. i'm not used to just getting a "do this, get that" answer, most i know i've learned by myself, i just seem to bump into problems trying to figure out electronics somehow. and that is reason 2 i'm not asking for direct solutions, i want to stop bumping into these, i hope, stupid hurdles doing "simple" stuff like controlling a LED...

So i want to get this working, but i like to understand what it is doing and possibly why it is working so i can perhaps make a better plan next time around.

There is just so much info online, someone giving an answer, another one telling that is a wrong answer and they all like to use shorthand words like Vce (which i make up would be voltage from collector to emitter..but there are many more :stuck_out_tongue: ). and then there are people saying bc337 is fine, others swear you need a mosfet and so on... along with a small language barrier.. meh :slight_smile:

I don't mind the problem solving, but sometimes i feel like i have no clue about better ways :slight_smile:

16

Please draw us a diagram of the microcontroller and its setup.

I think what is happening is that the base has a capacitance value and isn't completely discharging which is intermittently keeping the current flowing. After the 1k resistor, put a 10k resistor to 0v ground.

17

You might want to consider this in order to get started.
It will work from one 3.7V Li-ion battery
You will need three, one for each color.

1 Like
18

With a high power led, a simple series resistor won't cut it. This type of led gets quite hot, even with an attached heatsink (which you also need). As it warms up, it's forward voltage changes. No fixed value resistor will be appropriate for all forward voltages. A constant current supply adjusts the output voltage automatically to maintain a steady current as the temperature and forward voltage changes.

19

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.