LED's in parallel

[k7elp60]

There has been some discussion on LED’s in parallel. It is my sense that most experts recommend that LED’s not be connected in parallel that uses 1 current limit resistor for the x number of LED’s in parallel.
I would like to show the readers of this forum a couple of my projects that actually use high intensity LED’s in parallel with one current limit resistor for the parallel LED’s.

Both projects were built like name badges with the following dimensions 3.75” X 1.75” and each project consists of two circuit boards. The top board has the LED indicator and the transistor drivers, base resistors for the driver transistors and the current limit resistors. The bottom board has the electronics to flash the LED’s and the power supply. Both projects used 3V lithium coin cells for power. The first one I call the BOO badge as it was built for last Halloween used a CR2032, and the LED’s are multiplexed. The second project the LED’s just flash the different letters and are not multiplexed. Because I used blue LED’s the bottom board also has voltage pump that boots the voltage to 5V. The average current draw for the projects is as follows. Boo Badge 7mA, K7ELP badge 25mA. On the Boo Badge each letter of LED’s is split in 2 groups of LED’s in parallel with 1 current limiting resistor for each group, each current limiting resistor is 47 ohms. On the other badge each letter has all the LED’s in parallel with different values of current limiting resistors for each letter.

Both projects use CMOS IC’s

The schematic diagram of the BOO badge is also attached.

 

[audioguru]

LEDs are all different unless you test them and group them. If they were made at the same time then they are probably matched pretty well.

The LED with the lowest voltage will hog most of the current which will cause it to burn out, then the next one then the next one.

Maybe your very low current saved the LED with the lowest voltage from burning out or maybe you are lucky that they were all made at the same time.

 

[Ubergeek63]

It is the low current... He is sharing 25mA between 8 LEDs that can handle 25mA each, of course they are not burning out.

You could never get that to work if you actually needed the light. In fact a customer of mine just found that out the hard way driving 2 2A LEDs in parallel from a 2A source. Now I've had to add another 2A LED driver to an already dense PCB

 

[k7elp60]

There has been some discussion on LED’s in parallel. It is my sense that most experts recommend that LED’s not be connected in parallel that uses 1 current limit resistor for the x number of LED’s in parallel.
I would like to show the readers of this forum a couple of my projects that actually use high intensity LED’s in parallel with one current limit resistor for the parallel LED’s.

Both projects were built like name badges with the following dimensions 3.75” X 1.75” and each project consists of two circuit boards. The top board has the LED indicator and the transistor drivers, base resistors for the driver transistors and the current limit resistors. The bottom board has the electronics to flash the LED’s and the power supply. Both projects used 3V lithium coin cells for power. The first one I call the BOO badge as it was built for last Halloween used a CR2032, and the LED’s are multiplexed. The second project the LED’s just flash the different letters and are not multiplexed. Because I used blue LED’s the bottom board also has voltage pump that boots the voltage to 5V. The average current draw for the projects is as follows. Boo Badge 7mA, K7ELP badge 25mA. On the Boo Badge each letter of LED’s is split in 2 groups of LED’s in parallel with 1 current limiting resistor for each group, each current limiting resistor is 47 ohms. On the other badge each letter has all the LED’s in parallel with different values of current limiting resistors for each letter.

Both projects use CMOS IC’s

The schematic diagram of the BOO badge is also attached.

I decided to add some further comments to my original post. When LED’s are multiplexed they are actually flashed at a higher rate than the human eye can detect. It is my understanding that the human eye can see about 18 frames/sec. So if the multiplex frequency is 100 Hz then the eye sees the LED’s always on. An advantage of multiplexing is reduction in the average current consumption of the LED’s. I haven’t convinced my self that multiplexing reduces the brilliance as far as the eye is concerned.

This is a method I use to determine which LED’s in a batch to connect in parallel. I am aware of the Vf of the LED’s varying even at a specified current. I just checked the Vf of about 25 blue LED’s These LED’s I purchased in bulk from abcTronics.com.
The Vf a If of 5mA varied from 2.77V to 3.21V. What I did to group them was connect a series circuit of a 150W resistor and one LED to my adjustable power supply. Adjust the voltage until the current was 5mA. I chose the 5mA as these LED’s were very bright at that current. The specs of these LED’s specified a minimum of 2200MCD @ 20mA.

I measured and recorded Vf of each LED that I tried. I grouped the LED’s with the Vf variations of only 0.05V(50mV) The LED’s that I chose were that I tested were within .05V of 2.88V.
I then connected 10 LED’s in parallel, and using the single 150W current limiting resistor increased the voltage until the current drawn was 50mA. The Vf of the LED’s was 2.9V.

On the boo badge the LED’s are on for 2ms and off for 15ms, and the flash rate is pretty close to one second on, and one second off.
Both the badges are clearly visible in room with natural daylight and at a distance of 10 feet from the bades.
Cheers
Ned

 

[audioguru]

Cars that use LEDs for tail lights use PWM to dim them. I made some LED chaser projects that use PWM for dimming them.

PWM is exactly the same as multiplexing. The time that the LED is lighted is changed with the time the LED is turned off for each pulse.

On my chasers, when the pulse width is so small that I can't see the pulse on my 'scope but the dimmed LED still glows a little.

 

[Ubergeek63]

well that explains a couple things:

1: we were assuming you were running them on DC, and

2: you hand matched the diodes, ok for an experimenter who bought a bunch bulk but not practical otherwise.

You are correct though: there is not a direct correlation between PWM average current and perceived intensity. The eye is both a peak detector and an integrator-or more crudely you could think of a diode into a parallel RC.

 

[k7elp60]

I then connected 10 LED’s in parallel, and using the single 150W current limiting resistor increased the voltage until the current drawn was 50mA. The Vf of the LED’s was 2.9V.

Ned

This paragraph should read: I then connected 10 LED's in parallel, using the single 150 ohm current limiting resistor, increased the voltage until the current drawn was 50mA. The Vf of the LED's was 2.9V.
I originaly created the post in MS works and used the omega symbol, but when the post was pasted and posted on the thread the omega symbol changed to W.
Sorry for any confusion this has caused.

 

[Mikebits]

FYI, in the forum editor you can insert symbols for a few things like hm:
Just wish we had a few more like :lambda: and i: square root would be a nice one too

Hehe, doing pi in lower case gives you i: