3
\$\begingroup\$

schematic

simulate this circuit – Schematic created using CircuitLab

That's my circuit, although it's an RGB LED strip that I'm driving, rather than three individual LEDs. It works. I can PWM each channel and fade between colours as I need.

You may have noticed that the resistor value on the base of the transistor (you can find the data-sheet here) driving the red channel is lower than the others; a lot lower. This is the value I need to make each channel appear the same brightness.

My issue is that the transistor on the red channel is getting quite hot, too hot to touch for more than a few seconds.

I've put my multimeter in series with the collector of the transistor and the cathode of the LED to measure the current draw and it's reading 150 mA. I'm just a hobbyist, but from what I can tell, the transistor should be able to handle 600 mA from collector to emitter (at least that's how I've interpreted the data-sheet).

I suppose I have three questions:

  • Is the way that I've designed the circuitry reasonable/correct?
  • Is it normal/OK for transistors to get hot when operating within their limits?
  • As I've just read the transistor part number from the side of the component and then searched for the datasheet, is it possible that I've got a different spec'd device?
\$\endgroup\$
4
  • 1
    \$\begingroup\$ you can put a resistor in front of the red LED to pull some of the voltage drop away from the transistor. You might have to change the base resistor to a higher value after you do that. \$\endgroup\$
    – dandavis
    May 11, 2022 at 21:53
  • 2
    \$\begingroup\$ oh, and "too hot to touch for more than a few seconds" is plenty acceptable for a transistor; they can run hot enough to singe before they get problematic (electrically speaking). \$\endgroup\$
    – dandavis
    May 11, 2022 at 21:55
  • \$\begingroup\$ That's interesting, thanks for the comment. Hard to know all these small details when you don't do much of it! \$\endgroup\$
    – OdinX
    May 11, 2022 at 22:05
  • 1
    \$\begingroup\$ just keep with it, EE knowledge snowballs and build upon itself, so while the ramp-up is frustrating, before you realize it you'll be able to do all sorts of things you couldn't even describe before. \$\endgroup\$
    – dandavis
    May 11, 2022 at 22:08

1 Answer 1

Reset to default
1
\$\begingroup\$

600mA with a 2N5550? That's wildly optimistic. The saturation characteristics aren't specified for more than 50mA, and the current gain will be extremely low anyway.

enter image description here

That's in linear mode. In saturation mode, don't expect a current gain higher than 10. This means your transistor isn't getting enough base current, so it isn't fully on, and its Vce is much higher than it should, so it gets hot. There's no way to make this work with a 2N5550.

There are a lot of 50 cent MOSFETs that will do the job, with a RdsON below 50 mOhms, in convenient packages like SOT-23 or SO-8. TO-220 is overkill, but if you don't like SMDs, there are many in this package too.

Contrary to BJTs, MOSFETs don't have low current gain at low Vds: if there is enough voltage on the gate to turn it on, it turns into a resistor whose value is its RdsON. So a MOSFET can be driven at low frequencies, like your PWM, from a low current microcontroller GPIO without trouble.

Either get a MOSFET that will turn on fully with 3V3 on the gate, which is recognizable by a RdsON specification at this voltage or lower in the datasheet... or get one that will turn on with 5V on the gate (again, RdsON specified for Vgs=4.5V in the datasheet) and use a 74HCT logic gate to translate 3V3 logic levels to 5V and drive the gate.

Please do not use an IRF540, because this one won't turn on with 5V Vgs. Make sure to check it has a RdsON spec at the Vgs you'll use it at... and as usual, click "sort by price".

\$\endgroup\$
8
  • \$\begingroup\$ Hey, thanks for the answer. On the first page of the linked data-sheet, it says "Collector Current - Continuous" 600mAdc. What is that referring to if not how much current the collect can continuously handle? \$\endgroup\$
    – OdinX
    May 11, 2022 at 18:24
  • 1
    \$\begingroup\$ That's the maximum current before it is destroyed (without considering heating or safe operating area, which can destroy it at a much lower current). It doesn't mean the transistor will do any useful job at this current though, for example by providing current gain. For that, you have to look at the other specs, that show it'll work well up to 10mA in linear applications, with good constant hFe, and up to 50-100mA in switching applications, with the typical low hFe of bipolars used as switches. \$\endgroup\$
    – bobflux
    May 11, 2022 at 18:44
  • \$\begingroup\$ Right ok, thanks for clarifying. It sounds like I should be fine with the green and blue channels though, as they're only drawing about 30mA. So I just need a MOSFET for the red channel. \$\endgroup\$
    – OdinX
    May 11, 2022 at 18:48
  • \$\begingroup\$ @vimist is your red strip longer than the blue and green? No way that the red should be drawing 150 mA when the others are drawing 30 mA. Are you using pwm when measuring the current? Thats probably throwing your multimeter off. \$\endgroup\$
    – Passerby
    May 11, 2022 at 18:57
  • 1
    \$\begingroup\$ Maybe the red strip you got just has low efficiency so it needs high current. I got a strip like that and ended up not using it, at full power it couldn't match the green and blue. I ended up getting a more efficient one (read: less cheap) putting out a lot more light for the same current. \$\endgroup\$
    – bobflux
    May 11, 2022 at 19:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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