LED calculation question

Thread Starter

thermalRunaway3141

Joined Mar 11, 2023
4
HI,
I'm testing some High power SMD LEDs i received in the mail and getting some unexpected results. the LEDs in question have a forward voltage of 2.6 volts and I'm running 3 in series at 12 volts with a 30 ohms of resistance. my target current is 150mA. I calculated 140mA with the set up i described. when i plugged these values into my simulation software it reads 149.7mA. I'm only getting between 67mA and 75mA. I've tried several different power sources and have tested on 2 different multimeters. also, I've done two identical tests with LEDs from two separate manufactures and am at a complete loss. i have no way of explaining this. Any ideas?
 

ericgibbs

Joined Jan 29, 2010
17,454
hi 3141,
Welcome to AAC.
What colour are the LED's.?
E

Edit: what voltage do you measure across the individual LED's when running at 67mA.?
 

Thread Starter

thermalRunaway3141

Joined Mar 11, 2023
4
they are white (6000) 5730 packages.

11.87v, 8.62v, 5.43v. for a voltage drop of 3.25

thanks, you have brought me to the answer. I didn't expect the thermal characteristics to impact the performance that quickly though. I would think my multimeters response time would be marginally faster than it took for the LEDs to reach an adverse temperature. although they are from some no name Amazon manufacturers so quality wasn't really what I had expected.

I've got a set of 3 in the freezer I'm going to test with my multimeters Data hold function set to record maximum value and see what happens.
 

ericgibbs

Joined Jan 29, 2010
17,454
Hi,
White LED's have a typical forward drop of ~3.2V, so approx 9.6V for the 3 in series.
That would account for the low mA with 30R resistor.
E
 

MisterBill2

Joined Jan 23, 2018
14,756
For LEDs at that power level you MUST have them on an adequate heat sink for any operation other than brief pulses. Like mosy diode devices the forward voltage drop depends a great deal on the temperature.
At 3 volts and 150 mA the dissipation is 450 milliwatts and so they will heat quite rapidly.
 

ElectricSpidey

Joined Dec 2, 2017
2,436
Forget about the thermal issues for a second...2.6 volts is low even for cold white LEDs running at even 60 or so mA.

After a few seconds thermal management becomes critical.

Get your LEDs set up on the proper heatsinks...let them run a few minutes then measure the forward voltages again and recalculate the resistor.

Working with the higher power LEDs takes some experience and patience.
 

MisterBill2

Joined Jan 23, 2018
14,756
It IS possible that the LEDs are running far below rated output. It is also possible that they are rejects and do not meet their specifications.
 

Jon Chandler

Joined Jun 12, 2008
615
If you're measuring current with your DVM,
in series, there's a voltage drop across the meter. It's better to measure voltage drop across the series resistor and to calculate the current from that.
 

MrAl

Joined Jun 17, 2014
10,102
they are white (6000) 5730 packages.

11.87v, 8.62v, 5.43v. for a voltage drop of 3.25

thanks, you have brought me to the answer. I didn't expect the thermal characteristics to impact the performance that quickly though. I would think my multimeters response time would be marginally faster than it took for the LEDs to reach an adverse temperature. although they are from some no name Amazon manufacturers so quality wasn't really what I had expected.

I've got a set of 3 in the freezer I'm going to test with my multimeters Data hold function set to record maximum value and see what happens.
Hello there,

The 2.6 volt spec you quoted earlier sounds strange, i dont think i've ever heard of a white LED that runs that low at any current. As Eric pointed out, it would be more typically over 3 volts. Considering the current you measured was around 70ma and with 3.25 volts that's a mere 1/4 watt which even small parts can usually take without overheating. You might mention the size of the LED though too, that would help understand this better also, and what kind of heat sink is being used for each LED or just what they are soldered to.
 

MisterBill2

Joined Jan 23, 2018
14,756
HI,
I'm testing some High power SMD LEDs i received in the mail and getting some unexpected results. the LEDs in question have a forward voltage of 2.6 volts and I'm running 3 in series at 12 volts with a 30 ohms of resistance. my target current is 150mA. I calculated 140mA with the set up i described. when i plugged these values into my simulation software it reads 149.7mA. I'm only getting between 67mA and 75mA. I've tried several different power sources and have tested on 2 different multimeters. also, I've done two identical tests with LEDs from two separate manufactures and am at a complete loss. i have no way of explaining this. Any ideas?
My very first guess would be that there is additional resistance not accounted for in that test setup, or a power supply issue. Or possibly a measurement problem, or maybe even unfiltered AC, so that average current was measured. Possibly a poor connection, or, as has happened, a high resistance test lead.
 

BobTPH

Joined Jun 5, 2013
6,542
I think it is much more likely that the LED forward voltage is not 2.6V. I have never seen a white LED with less than 3V. TS does not say where he got that figure. Perhaps a typo and it is 3.6V?
 

MisterBill2

Joined Jan 23, 2018
14,756
Some LEDs will emit some light at far less than the rated current, and less than the rated forward voltage. I own a string of 32 white LEDs that starts to illuminate at about 65 forward volts and gets brighter at about 120 forward volts from a bridge rectified mains. No obvious series resistor seen. So the TS may have some LEDs of that kind.
 

MrAl

Joined Jun 17, 2014
10,102
Some LEDs will emit some light at far less than the rated current, and less than the rated forward voltage. I own a string of 32 white LEDs that starts to illuminate at about 65 forward volts and gets brighter at about 120 forward volts from a bridge rectified mains. No obvious series resistor seen. So the TS may have some LEDs of that kind.
Is the 65v perfectly smooth DC or pulsating DC with ripple as is typical?
 

crutschow

Joined Mar 14, 2008
32,088
Like most junction diodes, an LED's forward drop has a significant negative temperature coefficient, so their voltage would decrease when they get hot, not increase.
Below is an example graph of that:

1678688090557.png
 

MisterBill2

Joined Jan 23, 2018
14,756
Is the 65v perfectly smooth DC or pulsating DC with ripple as is typical?
The 65 volts is neither ! It is the AC voltage reading from a variable voltage AC transformer as delivered to the full wave bridge rectifier feeding the string of LEDs. So it is an indication of the variable peak voltage supplied.
My point is that this is a string of white LEDs that lights at much less than the rated voltage. So there is at least one variation that illuminates at a lower voltage.
 

Audioguru again

Joined Oct 21, 2019
5,829
The output of the bridge rectifier is DC pulses of 100Hz or 120Hz with a peak voltage of 90V and drops to 0V between pulses. Your DC meter is measuring 2.6V average voltage for each LED. The average current and brightness will also be less than with filtered DC.
If you add a filter capacitor then the voltage and current will be too high and burn out the LEDs unless you recalculate how many LEDs and the current-limiting resistor.
 

Attachments

MisterBill2

Joined Jan 23, 2018
14,756
No matter what the actual peak value of the rectified but not filtered AC is, when the sine wave source is almost doubled, the peak value is also almost doubled.
My sole point in mentioning this was that at the lower voltage and lower current, the LEDs did illuminate,
Thus the LEDs that the TS was evaluating could have produced quite a bit of light at a much lower forward voltage.
 

MrAl

Joined Jun 17, 2014
10,102
The 65 volts is neither ! It is the AC voltage reading from a variable voltage AC transformer as delivered to the full wave bridge rectifier feeding the string of LEDs. So it is an indication of the variable peak voltage supplied.
My point is that this is a string of white LEDs that lights at much less than the rated voltage. So there is at least one variation that illuminates at a lower voltage.
My inquiry was to find out exactly.
If it is 65vrms then that's 91v peak, so peak LED voltage around 3v.
 

Thread Starter

thermalRunaway3141

Joined Mar 11, 2023
4
thanks for spitballing this with me everyone.

so its not a filtering issue. its an automotive application so im looking at 12.6 to 14.1 volts DC. specifically its a public safety project that has to function normally without the vehicle idling and just to be certain I used 3 series lithium polymer cells (12.6 volts DC) as a voltage source and got the same results.

crutschow i thought the same and thankyou for supplying the graph but these are showing a higher forward voltage under load than when tested out of the box. I could look at a data sheet for an equivalent part but these are "fake chips" from no name china manufacturer so unfortunately the numbers would be meaningless.

to those who didn't catch it before the LEDs in question are 5730 package 6000 kelvin smds. and yes there is a heat sink pad on the LED that in this case is connected to the anode. Ive designed a custom footprint and am etching PCBs by hand that are inlayed into aluminum. I'm forced to leave the fiberglass in the PCB as an insulator between the LEDs and the aluminum because of the polarity of the heat sink pads. the PCB should transfer enough heat energy (although it is thermally insulative) into the aluminum for the current design parameters to work.

no, its not a fluke manufacturing error, I have thousands of LEDs from two completely separate manufacturers and I'm pretty sure the sections of tape and real I have are from different spools. I also have LEDs from the same manufacturers in blue(FV 2.508), red (1.824) and amber (1.844). those values are tested and i should mention that all of these LED varieties are testing and operating normally at the rated 150mA with the same set up I've described, under the same conditions, with the same measurement equipment. only the white LEDs are preforming outside of my expectations.
 
Top