Parallel LEDs prohibition

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
Post Reply
Newz2000
Posts: 507
Joined: Wed May 18, 2005 1:01 am
Location: Des Moines, Iowa, USA
Contact:

Parallel LEDs prohibition

Post by Newz2000 » Tue Apr 18, 2006 6:52 am

I seem to remember it being considered bad form to run say 7 "identical" leds in parrallel with one resistor. Is this still true?

I say identical but what I mean is 7 LEDs from the same manufacturer, same part and theoretically the same specs.

I'm not trying to save cost but trying to save board space. Some times all will be on at once, sometimes only some of them will be on at one time.

JPKNHTP
Posts: 488
Joined: Wed Jun 29, 2005 1:01 am
Location: Missouri
Contact:

Post by JPKNHTP » Tue Apr 18, 2006 6:58 am

-JPKNHTP
-God Bless

User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Post by Chris Smith » Tue Apr 18, 2006 7:35 am

Yes, BUT,...if you measure the highest draw of all the LEDS and then run the over all power level at less than the highest draw value Led, then and only then is it safe to use one resistor.

But, the LEDs will still not be equal in brightness.

In other words, run the heaviest draw Led at say, 75% of its power rating, while running all the other Leds at even less, and you will be at least “safeâ€

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Post by MrAl » Tue Apr 18, 2006 7:59 am

Hi there,

I did a quite complete analysis on this on another site and
found out that with parallel LED's and only one resistor
one LED could be drawing much more current than the
others so it's not too practical to do this. By 'practical',
i mean that you want to get as much brightness out
of the LED's as you can with out risking any of them.
For more complete discussions on LED's, see CandlePower Forums.
The analysis appears in one of the sticky subjects in
"Flashlight Electronics" under "CPF University -- EE Course".


Take care,
Al
LEDs vs Bulbs, LEDs are winning.

Newz2000
Posts: 507
Joined: Wed May 18, 2005 1:01 am
Location: Des Moines, Iowa, USA
Contact:

Post by Newz2000 » Tue Apr 18, 2006 1:44 pm

I didn't mention this, but I guess it has bearing...

all of the LEDs are rated at I = 20ma and I'll be running them at 4ma.

JPKNHTP
Posts: 488
Joined: Wed Jun 29, 2005 1:01 am
Location: Missouri
Contact:

Post by JPKNHTP » Tue Apr 18, 2006 2:15 pm

-JPKNHTP
-God Bless

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Post by MrAl » Tue Apr 18, 2006 5:58 pm

Hello again,

4ma? that's all? Why so low?

I would think you could get away with paralleling them, but
using separate resistors always works better.
LEDs vs Bulbs, LEDs are winning.

Robert Reed
Posts: 2276
Joined: Wed Nov 24, 2004 1:01 am
Location: ASHTABULA,OHIO
Contact:

Post by Robert Reed » Tue Apr 18, 2006 8:40 pm

With all that current fluctuation from lighting anywhere from one led all the way to seven led's through one dropping resistor, your supply voltage is going to be all over the place as well as LED brightness. This is just too obvious-am I missing something here?

JPKNHTP
Posts: 488
Joined: Wed Jun 29, 2005 1:01 am
Location: Missouri
Contact:

Post by JPKNHTP » Tue Apr 18, 2006 11:23 pm

-JPKNHTP
-God Bless

Newz2000
Posts: 507
Joined: Wed May 18, 2005 1:01 am
Location: Des Moines, Iowa, USA
Contact:

Post by Newz2000 » Wed Apr 19, 2006 6:37 am

First, I made an error, it's only 5 LEDs, not 7.

I'm making an LED matrix. I made an error and set it up so that 5 LEDs would share a resistor. It was going to be tricky to redesign the board but I just bit the bullet and did it right.

The reason I'm only using 4ma is because a: I'm working on a very low powered circuit and b: a PIC is sinking all of the current, and it can handle at most 25ma per i/o pin.

Colinr
Posts: 33
Joined: Wed Oct 23, 2002 1:01 am
Location: UK
Contact:

Post by Colinr » Wed Apr 19, 2006 11:06 am

Assuming the supply is a fixed voltage then the resistor value sets the current through all of the LED's so if you set the resistor value for 4mA so with one led on then that led will get 4mA if 2 led's are on each will get 2mA (assuming identical LED's) and 1ma if 4 LED are on. which will cause the LED to dim as more of them are switched on.


you can use one resistor if you use some form of time division multiplexing only alowing one led oan at any time.

Colin

k7elp60
Posts: 226
Joined: Wed Dec 05, 2001 1:01 am
Location: Ivins, UT
Contact:

Post by k7elp60 » Wed Apr 19, 2006 9:07 pm

I have been putting LED's parallel for quite a number of years. One of the latest projects is a flashing name badge. The letter K has 9ea T1 LED's in parallel. It is working off a 3V lithium battery with a 2N3904 as a driver. The current limiting resistor is 43 ohms. Each LED consumes about 3-4 Ma. The same badge the letter P has 7 LED's in parallel with a current limiting resistor of 62 ohms.

I have built miniature Christmas light strings with red, yellow and green LED's all in parallel in a string of about 50 LED's The string is powered by a pair of C cells in series with a 3A schotky diode. The string was built about 10 years ago and still works today. I don't remember ever replacing a LED.

A number of years ago I wired a live horse with over 150 red LED's It was for a night parade. There was a string down each leg. One string on
each side of his head. There were two battery packs built just like the one described for the Christmas lights above, to power all the LED's.

So I say you can wire LED's in parallel.

User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Post by Chris Smith » Wed Apr 19, 2006 10:05 pm

At lower currents, ALL Leds are quite happy with ANY configuration below their specs.

However pushing Leds to their normal limits of currents or brightness in parallel with a single resistor, or past this level, is where all the problems arise.

A Space saver tire is quite happy doing 35mph on the way to the garage.

Try it at 100mph and see what happens.

JPKNHTP
Posts: 488
Joined: Wed Jun 29, 2005 1:01 am
Location: Missouri
Contact:

Post by JPKNHTP » Wed Apr 19, 2006 11:00 pm

-JPKNHTP
-God Bless

Engineer1138
Posts: 458
Joined: Thu Feb 05, 2004 1:01 am
Location: Minneapolis, MN
Contact:

Post by Engineer1138 » Thu Apr 20, 2006 12:47 pm

Matt:
Here's a configuration I used that you may want to try, since you have a PIC in the circuit.

I recently built a PC board to drive 16 LEDs (among other things) and I forgot the resistors! That's right: power drivers sending 12V straight to the LEDs. What saved my bacon was PWM. I calculated a duty cycle to provide approximately 1.4 volts to the LEDs and drove them that way.

Since you are using 5 LEDs you can PWM them at a variable rate that keeps current at a constant ratio of number of ON LEDs. e.g. for 1 LED on, use 20% duty cycle, for 2 use 40% and so on. This should even out the brightness as LEDs turn on and off and you won't need any resistors at all.

This works because the LEDs can handle the full power supply voltage, but heat will kill them and the PWM makes sure they never get hot.

Post Reply

Who is online

Users browsing this forum: CeaSaR and 28 guests