0
\$\begingroup\$

schematic

simulate this circuit – Schematic created using CircuitLab

I'm having trouble understanding from the datasheets what exactly happens when I put a circuit as follows:

The two 20W LED don't have a specific current description but they are 12V so I assume they take between 1,5 to 1,7 Ampere to run. My idea is to use a resistors and potentiometer to make a current regulator that allows currents up to 2.9A to under drive the LEDs and hopefully make them cool enough to last.

So I have multiple questions:

  1. What is the voltage drop across the TIP41c that I need to take into account when it passes continuously 2.9A? Is the back of the TIP41c connected to one of the pins? Is it ok to connect it to ground?
  2. Am I ok with the limitations of the LM317?
  3. I plan to use 3W resistors for R2 and R4. What is the minimal Wattage rating for the potentiometer I should use?
  4. Would 1K potentiometer be enough for dimming the LEDs fully? If not, what rating should I use for a good linear dimming using a 1 turn Potentiometer?
  5. Am I overlooking anything in design?
  6. I didn't understand the hfe part of the TIP41c datasheet. \$h_{fe}=\frac{i_c}{i_b}\$? How do I use that to know if R1 is good or not?

Many thanks

Second revision of my question:

I tried to learn from the first go and found a video that I want to adapt:

https://www.youtube.com/watch?v=soygy1rnLcs

Here is the circuit,

schematic

simulate this circuit

Now there are many things I'm not sure about:

  1. Am I right to use the cap to reference to ground? What should its rating be?
  2. am I forgetting resistors that need to be put to protect the transistors?
  3. I still want current to be in the range of a few mA to 2.9A for all the LEDs. Is this a good way to achieve that (and I just need to find the right resistor values) or did I get the video wrong, and what am I doing wrong?
  4. If I'm using 20W LEDs that are rated at 12V like that and I'll underdrive them since some Voltage will go to the circuit I'm trying to build here, is it really a bad idea to limit the total current of all 4 LEDs and assume they are the same and will use the same current? I know I'm using 1 opamp in the LM358, is it possible and better if I check on the individual nodes N1 and N2? How would I do that?

Again, thank you

Ok let's forget the LEDs for a sec. I'll try to copy what is in the video and fit it as closely as I can to just have the regulator:

schematic

simulate this circuit

If I start with setting the current I still have the problem using a \$R_{shunt}=0.1 \Omega\$ so that at the maximum current of 2.9A only 0.29W will be lost as heat, and if anything goes wrong it'll be 4A that run through it which is still ok for a 3W resistor. Is there a transistor (Q3) that can handle 1.5A and will work with that small \$V_{be,sat} = 0.19V\$?

Thank you

\$\endgroup\$
12
  • 3
    \$\begingroup\$ LM317 is a regulator, not a transistor. Based on the 317, it would have to be a series regulator circuit with an external transistor (the TIP41) for more power capacity, and I'm not even sure if it is right, looks like a shunt regulator circuit here. look up both of those... LM317 and TL431, look in in the "applications" section of the respective datasheets \$\endgroup\$ – Pete W Jan 13 at 16:50
  • 3
    \$\begingroup\$ Also the two LED strings in parallel is not a good idea at all \$\endgroup\$ – Pete W Jan 13 at 17:12
  • 2
    \$\begingroup\$ You can't put LEDs in parallel like that and assume that they will share the current equally. If they are designed to be driven with a constant voltage (12V, as you said) then the amount of current they draw could vary quite a bit. \$\endgroup\$ – Elliot Alderson Jan 13 at 17:12
  • 3
    \$\begingroup\$ A 1k potentiometer in series with 430 milliohm won't work. You will barely touch the knob and the current will fall very low. \$\endgroup\$ – Elliot Alderson Jan 13 at 17:14
  • 2
    \$\begingroup\$ Did you actually mean a BC317 BJT? \$\endgroup\$ – Andy aka Jan 13 at 17:29
0
\$\begingroup\$

you might be trying to do this, in principle. However, not quite suitable for power level, just showing the shunt regulator configuration.

enter image description here

Next, here is how to do constant current with the 317 regulator, again in principle, probably not good for 3A directly, although the TO-220 packaged 317 has some power capability

enter image description here

\$\endgroup\$
2
  • 1
    \$\begingroup\$ too much voltage drop \$\endgroup\$ – Tony Stewart EE75 Jan 13 at 18:58
  • 1
    \$\begingroup\$ You can always boost the current output of one of these regulator-based current sources with a pass transistor, which seems to be what the original question is asking about, but it's not going to help the voltage drop problem in those resistors. \$\endgroup\$ – Hearth Jan 14 at 16:53

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.