In the below circuit, the circuit maintains a constant current on LED. My question is after the LED there is a voltage of 18V, still, need to be dropped. Will the NPN Transistor heat up while dropping the voltage? Also, the voltage source to led will vary from 12V to 20V. How can we modify the circuit? Or the existing circuit is fine?

enter image description here

  • 2
    \$\begingroup\$ The polarity of your left-hand transistor is wrong. The transistor is operating in reverse mode (which it will do, just not very well). Try connecting the emitter to ground. \$\endgroup\$ – WhatRoughBeast Mar 17 at 14:31
  • \$\begingroup\$ Right. What difference will it make by the way? \$\endgroup\$ – Embedded Geek Mar 17 at 14:37
  • \$\begingroup\$ The transistor will have less gain in the reversed configuration than in the normal. So in normal configuration the current regulation will be greater for supply voltage and load variations. \$\endgroup\$ – WhatRoughBeast Mar 17 at 14:42
  • 2
    \$\begingroup\$ It is good practice to number the components R1, R2, Q1, Q2, etc. so we can discuss them without confusion. \$\endgroup\$ – Transistor Mar 17 at 14:47
  • \$\begingroup\$ @Transistor You are right, but the tool didn't allow me to add the component name. Apologies. \$\endgroup\$ – Embedded Geek Mar 17 at 14:48

Yes, it will work- it's a switched (not switching) crude linear regulator, but Q2 is operating in reverse mode and will regulate somewhat more poorly. It would be worse if your LED current was less since it has to sink about 3mA to throttle the base current of Q1, so that will consume about 1mA of base current in reverse active mode.

The transistor Q1 is dropping about 17.2V at 18.2mA so it will dissipate (ignoring base current dissipation, which is negligible) 313mW which is quite a bit for a TO-92 transistor. Assuming a 50°C maximum ambient and 200°C/W J-A thermal resistance, the junction will be at 112°C which is acceptable but not great for good reliability. If the '20V' was actually 30V it would be much worse, of course.

You can reduce the dissipation in the transistor by adding a resistor in series with the LED (based on the minimum the 20V supply could go such that you want to maintain regulation). For example, if it has to regulate from 15V to 20V you could add a resistor dropping about 5V, so 5/18.2 = 270\$\Omega\$.

If the 20V isn't going to change much you hardly need a regulator at all, and just switching a resistor with a single transistor (which will then run cold) would be better and simpler (and more reliable). For example, a 1K 1/2-W resistor.

It's not a very good regulator because the "reference" is the Vbe of Q2 and it will vary with temperature. In some cases that could be considered a feature but I'll leave that discussion for another time.

Here is an LTspice simulation, however the temperature curves apply to the junction temperature of the transistor. As you can see, temperature is much more of an effect than the voltage (ignoring self-heating).

What will happen in fact is that there will be some self heating so the current will start off higher and tail off as Q1 heats Q2.

enter image description here

If Q2 is connected correctly, the LED current will be a bit lower because the base does not need so much current. My simulation (using the LTspice models) shows a bit lower current than yours so the dissipation is proportionally less. The major factor may be how optimistic their number is for reverse beta.

For a little post-prandial edit, I've added plots of Q1 dissipation in relation to the LED supply and temperature (blue lines).

enter image description here

  • \$\begingroup\$ Actually the temperature of Q1 will have little effect on the current, it's the Vf of the (if it's reversed-active mode) Q2 base-collector that dominates and provides the voltage reference. Q1 has almost nothing to do with the reference. \$\endgroup\$ – Jack Creasey Mar 17 at 23:54
  • \$\begingroup\$ To show that Q2 V(BE) has no impact on the output current at all you could change Q2 for an N-channel FET and this would still regulate (though the reverse-active mode should no be used). \$\endgroup\$ – Jack Creasey Mar 18 at 0:18
  • \$\begingroup\$ @JackCreasey That's why I was talking about the heat from Q1 reaching Q2 and thus changing the reference. Vgs(th) of a MOSFET is temperature sensitive too, and much more variable from unit to unit, so a BJT is the way to go. \$\endgroup\$ – Spehro Pefhany Mar 18 at 0:20
  • 1
    \$\begingroup\$ @mkeith Good question, I would guess it to be similar, though the Vf is higher because of the much higher base current so maybe it's worse, and maybe the resistive component and beta tempco cancel the diode part somewhat or maybe the add to it. The main difference I'm aware of is that Vce(sat) can be lower in reverse but that's of no consequence here since it never saturates. The current really is not all that repeatable because of the reverse beta so I would guess it's just an error. \$\endgroup\$ – Spehro Pefhany Mar 18 at 0:56
  • 1
    \$\begingroup\$ Yeah. I am hesitant to fully trust the simulation. But I am definitely not going to build the circuit to check it either. ;-) \$\endgroup\$ – mkeith Mar 18 at 1:03

The circut will work and the main transistor will dissipate around 300mW and have a temperature of 55°C.

P = (Vc-Ve)*Ie

300mW = (17V-0.8V)*0.018A

(You could use Ic, the diffrence will be very small since Ie=Ic+Ib and base current is insignificant.)

You should use a transistor able to dissipate more than 0.6W (ideally 1W), such as 2N3904.

The temperature can be calculated this way:

Tj=P*Rthj-amb (thermal resistance)

For 2N3904:


This is hot to touch but will not damage the transistor, I suggest to put it away from the plastic case and capacitors.

  • 3
    \$\begingroup\$ FYI: 55 degrees C is not simply “hot to touch” it’s at the point of causing first degree burns in a few seconds. Not extremely hot for silicon, way too hot for skin. \$\endgroup\$ – Edgar Brown Mar 17 at 17:24
  • 2
    \$\begingroup\$ Small typo. You wrote 0.018 mA instead of 18 mA. \$\endgroup\$ – Mitu Raj Mar 17 at 19:58
  • \$\begingroup\$ Thanks for noting that typo Mitu. \$\endgroup\$ – thegamebusterpl Mar 18 at 16:47
  • \$\begingroup\$ @EdgarBrown Im still technically right tough!😂 \$\endgroup\$ – thegamebusterpl Mar 18 at 17:14

Yes, the NPN transistor will heat up by exactly the same amount as a variable resistor if it were used to set the same LED current.

The circuit shown should work reasonably well down to a few volts but don’t just trust me on this - get a free sim tool and prove it yourself.

  • \$\begingroup\$ Which free sim tool? Unable to find, any links? \$\endgroup\$ – Embedded Geek Mar 17 at 14:39
  • 2
    \$\begingroup\$ @EmbeddedGeek circuitjs is popular and easy, or you could use the more powerful LTspice. \$\endgroup\$ – Hearth Mar 17 at 14:50
  • 1
    \$\begingroup\$ or use falstad online simulator \$\endgroup\$ – Mitu Raj Mar 17 at 14:53
  • 1
    \$\begingroup\$ @MituRaj the Falstad simulator is circuitjs. \$\endgroup\$ – Hearth Mar 17 at 14:58
  • 2
    \$\begingroup\$ Nitpicking, but not exactly the same amount: it will get 0.5-1% extra heat from base current, and the 50 ohm resistor will reduce the heat a bit. \$\endgroup\$ – jpa Mar 17 at 20:02

Constant current for LED ...the voltage source to led will vary from 12V to 20V... the existing circuit is fine?

No, the circuit as shown DOES NOT WORK

I assume that the circuit should be as shown here:


simulate this circuit – Schematic created using CircuitLab

While you have not specified the devices used, I used a generic 2N3904 NPN here. The 2N3904, Q1 would dissipate about 288mW and it has a maximum rating of 625mW. At 25degC ambient you would expect the device chip to be at about 57degC and very hot to touch.

You could reduce the temperature of Q1 by adding a resistor for the LED, but all you are doing is spreading the dissipation between the resistor and Q1, you still have to dissipate 288mW.

A much better way to do this task is to use a single component such as the AL5809 constant current driver.

enter image description here

This is a single component, works from 2.5 - 60V and is cheap.

Update: It was suggested in comments that the circuit originally shown would work with the transistor in a reverse-active mode.
Maybe ….but the H(fe) would likely be in the single digit range (the 2N3904 is not exactly high gain) so trying to design the solution would be harder over the range of device parameters. See here.
Why anyone would suggest reverse active mode completely flummoxes me. There is no technical reason for choosing an unusual device operation mode for this type of simple application, so why even suggest it?
The feedback device (Q2 in my circuit) is never in Collector saturation, in addition it's never in Base/Emitter saturation, so there are no unusual configuration parameters.
The current does have a negative droop with temperature.

  • 1
    \$\begingroup\$ I understand I will be using, two resistors of 220 ohms each rated 1W in series with the LED, will dissipate much of the heat and much of the voltage, so Q1 won't be that much hot. Heat dissipation will spread between three components. Will that is fine? \$\endgroup\$ – Embedded Geek Mar 17 at 17:06
  • \$\begingroup\$ I think it actually would work as drawn, with Q2 being in reverse-active mode... I almost wonder if someone drawing this circuit was trying to use the "the BE junction of a 3904 makes a decent zener diode" thing, since a zener diode between base and ground would also make sense there... \$\endgroup\$ – Hearth Mar 17 at 17:09
  • 1
    \$\begingroup\$ @JackCreasey I don't think it makes sense either, I'm guessing as to what the person who drew the circuit may have been thinking. And with the transistor in reverse active mode, I just did some quick simulations and they seem to suggest that it still will work passably well as a current source. Of course, at that point, you're not getting much from having Q2 there and the circuit would work about equally well with it removed entirely, but it doesn't just stop being a current source. \$\endgroup\$ – Hearth Mar 17 at 17:28
  • 1
    \$\begingroup\$ @EmbeddedGeek I'd use just a single 0.25W resistor in series with the LED. If you want the current to be constant between 12V and 20V then you don't want to drop any more than about 10V across the resistor. So a series resistor of 520 Ohms would seem appropriate. \$\endgroup\$ – Jack Creasey Mar 17 at 17:28
  • 1
    \$\begingroup\$ @Hearth I would seriously question any simulation, it's very rare to see simulation characteristics defined for reverse operation. Much better to use the devices as defined for normal operation. \$\endgroup\$ – Jack Creasey Mar 17 at 17:29

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.