# Measurement of unknown LED

Discussion in 'General Electronics Chat' started by Avijit Palit, Aug 11, 2017.

1. ### Avijit Palit Thread Starter Member

Jul 19, 2017
81
1
I found this kind of LED from garbage.
I don't know its voltage or wattage rating.
And there is no printing over LED.
I connected it with smps (12 volt 0.5 amp) and its heats up.
Is there any way to measure this LED.

Jan 15, 2015
3,150
1,402
Start with a low voltage and while monitoring voltage and current slowly increase the voltage until the LED offers the intensity you expect and note the voltage and current. Starting with 12 volts and a LED of that type with no apparent current limiting resistor is not at all a good idea. You want an adjustable power supply or start with 1.5 volt batteries addig a battery in series each step, such as 3.0, 4.5, 6.0 ....

Ron

Tonyr1084 likes this.
3. ### DickCappels Moderator

Aug 21, 2008
4,764
1,419
It looks like a white LED and if you were lucky it was a reverse polarity diode that heated up.

The voltage is probably somewhere between 3 and 4.5 volts depending upon the quality. Except in very carefully controlled circumstances you should, as Reloadron just told you in post #2, always limit the current to an LED with a resistor or current source rather than driving it with a voltage. It looks like a nice big one that could take a lot of current, but if you limit the current to a few milliamps when you start out you should be able to see a glow if the LED is still any good. A 12V power supply and 330 to 470 ohms would be a good place to start.

IF there is a reverse polarity protection diode it will measure less than a volt when the LED is biased in reverse.

One has to wonder why it was in the garbage.

Avijit Palit and Tonyr1084 like this.
4. ### ScottWang Moderator

Aug 23, 2012
6,104
950
Vcc : 9-11V
I = 600mA
W = 10W
Led : 9 piece
3 in series and then 3 in parallel, so there are 9 Leds inside.
When you use this kind of led then you need to use a heat sink for the heat dissipation.

Avijit Palit likes this.
5. ### DickCappels Moderator

Aug 21, 2008
4,764
1,419
Wow Scott! How did you find out there are 3 sets of 3 parallel LEDs in the package??

6. ### ScottWang Moderator

Aug 23, 2012
6,104
950
I guessed that the Vf of led could be 3V~3.6V, and there are 3 leds in series in the picture, so it should be as 9V~10.8V, and that was matched the voltage from the labeled, they also labeled it as "三串三并" it means that it is a 3x3 led, you also can see the post #4.

Sinus23 and DickCappels like this.

Apr 5, 2008
18,117
3,446
ScottWang likes this.
8. ### ScottWang Moderator

Aug 23, 2012
6,104
950
According to my experienced that they like to exaggerate the values of spec, like as I have had a electronics load that the real wattages about 150 watts, but they claimed that it has had 400 watts.

Yes, you are right.
I didn't calculate it, just according to the labels.

Oct 15, 2009
5,353
1,181

10. ### Avijit Palit Thread Starter Member

Jul 19, 2017
81
1
But how do I identify? It is heats up on 8 volt as well as 12 volt or 14 volt.

11. ### Avijit Palit Thread Starter Member

Jul 19, 2017
81
1
I attached it with 6 volt 500 mA and among there 6 lights only one lights up. And also added 220 ohms to 12 volt power supply, it is heats up.

12. ### Avijit Palit Thread Starter Member

Jul 19, 2017
81
1
I attached a heat sink to this led, you can see above. Is the heat sink so small to power dissipation?

13. ### WBahn Moderator

Mar 31, 2012
22,614
6,729
As has already been mentioned a couple of times, start out slow and work your way up.

If you just have a fixed 12 V supply, but a resistor in series that would limit it to no more than about 100 mA. So you might start with a 100 Ω resistor, which will limit it to no more than 120 mA even if it was a dead short through the device. If the LEDs are good, then they will probably drop 7 to 9 volts at that low a current, which will mean that your current will only be about 30 mA, which is good because that will only be about 100 mW, in rough numbers, of power dissipation in the resistor. So connect it up and then measure the voltage across the resistor. From that, you can calculate the voltage across the LED array and the total current in it. Plot that on a piece of graph paper (or in a spreadsheet). If you have a bunch of 100 Ω resistors, then put one in parallel with the first one and measure the voltage across it. Calculate the LED voltage and current and plot it. Keep putting more resistors in parallel and taking measurements. That will let you map out the I-V characteristic for the device. At some point you might want to trade out to some lower valued resistors. If you are using 1/4 W resistors you want to keep the power in each down in the 100 mW range or so. Eventually, if it takes 11 V to get 600 mA in it, you will been something around 1.5 Ω and you will need about five or six such resistors.

Once you start getting significant current flowing in the LED it is going to get hot. 6 W is a LOT of heat -- try touching a 1/4 W resistor running at its max power and you will become a believer.

14. ### WBahn Moderator

Mar 31, 2012
22,614
6,729
Sounds like you are discovering why it was in the trash.

Sinus23 likes this.

Aug 23, 2012
6,104
950

16. ### Tonyr1084 Well-Known Member

Sep 24, 2015
2,116
459
Enjoyed the video. FIRST PROBLEM: The solder never bonded to the heat sink. SECOND PROBLEM: (a question I don't know the answer to) Can you solder to aluminum? (assuming the heatsink was aluminum). THIRD PROBLEM: The technician in the video discovered that the LED will get hot enough to melt lead (probably). FOURTH PROBLEM: The technician attempted to (and assuming he did) accomplish thermal bonding on four corners, four corners of very small contact via the solder will not provide sufficient thermal bonding between the LED and the heatsink. But I did enjoy the video.

17. ### dl324 AAC Fanatic!

Mar 30, 2015
6,611
1,519
The proper technique would be to use a thin layer of thermal paste to fill in all of the voids between the heatsink and the LED.

18. ### Tonyr1084 Well-Known Member

Sep 24, 2015
2,116
459
Yes. And likely to clamp the LED to the heat sink. Unless you use a thermal epoxy to bond it to the heat sink.

19. ### ian field AAC Fanatic!

Oct 27, 2012
6,141
1,110
That type of LED often has multiple dies in series, so no way a DMM diode check will tell you anything.

A fixed voltage supply will either do nothing or kill it.

Constant current, or take your chances with current limited is the only safe way - you have to use some judgement, but 50mA should at least make it glow - my guess of somewhere in the general direction of 3/4A seems to be supported by some other posts.

20. ### Avijit Palit Thread Starter Member

Jul 19, 2017
81
1
Dont panic. I just wanna make a emergency light by it.