# How hot is normal heat for a resistor before a 12V 20mA LED?

#### ledtheyled

Joined Jan 23, 2024
4
Hello!

I bought some chinese LED with pre build resistor from Amazon for my model train houses. Forward voltage 12v forward current is 20mA for the LED. https://amzn.eu/d/fLFy9r3

My power supply delivers 12V 2A DC.

1) I put some of the LED (6) parallel on an electronic board. This works great, but the resistors are getting really hot.

I was able to melt candle wax with it, so they have at least 65 degrees 149 °F.

Is that normal?

2) For an second test, I put two of the LED in series. The resistors heat is then OK. Hopfully this will not damage the Power supply , or LEDs...

#### WBahn

Joined Mar 31, 2012
30,077
12 V at 20 mA is 0.24 W. From the picture, I have a hard time believing that the in-line resistor is rated at 1/4 W -- which would really be too small as the rule of thumb is that you want a resistor that is rated for about double of the amount of power it is expected to actually be called on to dissipate. So it's not surprising that it is getting very hot. Enclosing the resistor in heat shrink effectively reduces the power rating because it can't dissipate heat well into the air, which is what the power rating is predicated upon. Looks like a typical crap Chinese product.

That looks like a yellow LED, which probably has a forward voltage drop of about 2.3 V, which puts about 9.7 V across the resistor at 12 V, so it is probably either a 470 Ω or a 510 Ω resistor.

Putting two of them in series should then result in something in the range of to 7 to 8 mA., which will reduce the dissipation of each resistor by about a factor of 7, which should let them run a LOT cooler.

There is no danger to the power supply by putting them in series -- it's actually easier on the supply. If the LEDs are adequately bright this way, go for it.

If you really need the brightness from running them at full voltage, keep that in mind wherever you put them in your model -- that they are going to run quite hot and, enclosed in a model structure, that will only get worse.

#### ledtheyled

Joined Jan 23, 2024
4
Ok thanks.

I think that i will wire than 2x LEDs with the build in resistor in series. Yes they are a Lot of cooler than

DC 12V - resistor - LED - resistor - LED -

#### ElectricSpidey

Joined Dec 2, 2017
2,786
Or you could just power them from a lower voltage supply, such as 9 volts or 6 volts.

#### dl324

Joined Mar 30, 2015
16,943
Welcome to AAC!
I was able to melt candle wax with it, so they have at least 65 degrees 149 °F.

Is that normal?
The resistor would be dissipating a couple tenths of a watt. Assuming they used a 1/4W resistor, it would be operating at about 80% of its rated power.

If that's the case, your perception of "really hot" isn't relevant. A Yaego 1/4W resistor is rated to dissipate 100% of its power rating at temperatures under 70C. Other brands will be similar because resistor specifications are fairly standard.

This is for Yaego 1/4W carbon film:

Metal film (1%) are the same.

Conservative designers will derate at room temp, but it's not really necessary.
2) For an second test, I put two of the LED in series. The resistors heat is then OK. Hopfully this will not damage the Power supply , or LEDs...
The only things that's important is the brightness you want. If they're bright enough with 2 in series, there's no reason for you to operate them at a higher current.

Operating at a lower current dissipates less power. If they're bright enough, everything and everyone are happier. Except those who profit from power waste...

#### Audioguru again

Joined Oct 21, 2019
6,707
Of course the little resistor gets too hot.
You buy cheap junk from Amazon and you get cheap junk improperly designed.
Amazon is not an electronics parts distributor.

#### N5SLI

Joined Jan 24, 2024
1
I would use a buck converter. I use them for power injection for led's on my home. converter will drop the 12vdc down to 5vdc.

#### ledtheyled

Joined Jan 23, 2024
4
Or you could just power them from a lower voltage supply, such as 9 volts or 6 volts.
Yes i gave that also a try. 9volt battery works great.

#### ledtheyled

Joined Jan 23, 2024
4
If they're bright enough with 2 in series, there's no reason for you to operate them at a higher current.

Operating at a lower current dissipates less power. If they're bright enough, everything and everyone are happier. Except those who profit from power waste...
Yes the brightness is OK then for my tiny scale z model houses. 2 LEDs in series are also better in my Situation, because i can mount than in different places in a house then.

One last question.

The resistor is build in before the LED (you can see that also on my candle picture) the series Installation would look like that with 2 LEDs

DC 12V -> resistor-LED -> resistor-LED ->

Would not then the first resistor in line consume the the whole power and get hot?

#### tonyStewart

Joined May 8, 2012
131
If you wire them as 4S from 12V, no resistor needed and no over-current drawing about 15 mA

#### ElectricSpidey

Joined Dec 2, 2017
2,786
Would not then the first resistor in line consume the the whole power and get hot?
If both resistors are the same value in a series circuit, then no, they will both develop the same voltage and have the same current so the power will be equal.

#### dl324

Joined Mar 30, 2015
16,943
DC 12V -> resistor-LED -> resistor-LED ->

Would not then the first resistor in line consume the the whole power and get hot?
No.

Since all 4 components are in series, each LED will drop about the same voltage and the remaining voltage will be divided equally between the two resistors.

#### ThePanMan

Joined Mar 13, 2020
794
9volt battery works great.
9V batteries don't last long enough to be of use. They will drop voltage fairly quickly and your LED's will grow dim in a short period of time. The math is out there, I just don't feel up to chasing the numbers for you.
Would not then the first resistor in line consume the the whole power and get hot?
If both resistors are the same value in a series circuit, then no, they will both develop the same voltage and have the same current so the power will be equal.
This is 100% true. The total resistance and the total forward voltage drops from the LED's all form a single circuit. The current will be the same in all points. As Spidey said, if the resistors are the same then their heat dissipation would be the same.

ME? I'd just opt for an extra resistor between all the LED's and ground. Again, without running the number, the single resistor would need to be large enough in wattage to dissipate all the current. Suppose you have 10 circuits running at 20mA. That would be 200mA total. In that case you'd want to use a 3 watt resistor minimum but would recommend the next bigger size, that being 5W. IF you have enough room to place such a beast in your project somewhere.

Better still, use a buck converter to drop the voltage. Lower voltage means less current. Less current means less heat in watts dissipated. Select a variable buck converter and you're in the money. You can dial up or down the voltage. The result will be having a variable brightness to your project.

NOTE: LED brightness is not a linear factor. Half the current will not result in the appearance of half the brightness. The human eye has the ability to adjust to light levels.

#### WBahn

Joined Mar 31, 2012
30,077
Yes the brightness is OK then for my tiny scale z model houses. 2 LEDs in series are also better in my Situation, because i can mount than in different places in a house then.

One last question.

The resistor is build in before the LED (you can see that also on my candle picture) the series Installation would look like that with 2 LEDs

DC 12V -> resistor-LED -> resistor-LED ->

Would not then the first resistor in line consume the the whole power and get hot?
The two resistors have the same current flowing through them and the heat generated is the square of the current times the value of the resistor. Both resistors will produce the same heat.

#### WBahn

Joined Mar 31, 2012
30,077
If you wire them as 4S from 12V, no resistor needed and no over-current drawing about 15 mA
What do you base this on?

How are you coming up with the notion that 3 V applied across one of those LEDs, with no resistor, will result in about 15 mA of current though it?

#### tonyStewart

Joined May 8, 2012
131
Most 5mm White LEDs of decent quality are 3.05 to 3.1V @ 20 mA and std. tolerance to 3.3V max So the current will be less than 20 mA @3V and down to 2 mA or less at 2.8V typical. It was a comfortable wild assed guess or a WAG as they would have said in Silicon Valley.
I have about 3k surplus white 5mm LEDs all between 3.0 and 3.2V >=16 Cd if anyone wants any.... 30deg.

#### WBahn

Joined Mar 31, 2012
30,077
Most 5mm White LEDs of decent quality are 3.05 to 3.1V @ 20 mA and std. tolerance to 3.3V max So the current will be less than 20 mA @3V and down to 2 mA or less at 2.8V typical. It was a comfortable wild assed guess or a WAG as they would have said in Silicon Valley.
I have about 3k surplus white 5mm LEDs all between 3.0 and 3.2V >=16 Cd if anyone wants any.... 30deg.
Except these aren't white LEDs.

They come from a package containing different colors and the color of the LED is indicated by the color of the non-black wire.

Even if they were, with four in series, you have less than 3 V across each, and so they would be expected to have very little current in them. You can't relay on having LEDs that just happen to have an abnormally low Vf.

#### tonyStewart

Joined May 8, 2012
131
My misteak for not translating the german LED page.

Yes They are supposed to get this hot in a poor design. 1/4W will raise temperature to 135'C above 25'C ambient to 160'C and about 25% more with heatshrink insulation in convection free air. The designer ignored this problem ( but it automatically shrinks the heatshrink LOL) and should have split the waste power into multiple \$0.001 resistors.

Two in series is a quick solution to reducing the 240 mW by 50%

The optimal solution is to cut the wires and resistors and drive everything from 3.3V with smaller resistors and consume 29 mA* 3.3V = 66 mW max including the LED then add heatshrink.

Dang verbotten cheap solutions are not always best.
Lyeteung not!

Last edited: