The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Resistor math - more confusion Dynamics Processors (HW)
Old 27th December 2010
  #1
Lives for gear
 
datune's Avatar
Resistor math - more confusion

So I found a site, which provides a formula to calculate what resistor is needed for a couple of LED's in series.

I quote (http://www.brighthub.com/engineering...s/66726.aspx):

Quote:
LED Current Limit: Just like transistors, LEDs too are very sensitive to high currents. A resistor when placed in series with the LEDs regulates a proper flow of current through them. To calculate the value of a series LED resistor, the following formula may be used:
R = V –( N.VLED) / I
Here R = Series LED resistor, V = supply voltage, N = number of LEDs in series, V(LED) = forward voltage of the LED used, and I = current through the LEDs
So let me get this straight, if I have 12Volts, and I want to put 8 2V / 30mA LED's in series, I don't need a resistor?

According to his formula:

R = 12V - (8 * 2V) / 0.03A

To me this formula appears to be wrong either way, shouldn't it be:

(12V - (8 * 2V)) / 0.03A

Also, when using this formula to calcute a single LED's resistor, which would be (12V - 2V) / 0.03A = ~334

So I took a 330Ohm resistor, and added 4x 1Ohm resistors, for a total of 334, but if I measure the Voltage across the LED, I get 2.05Volt, and NOT 2Volt. I experimented for almost an hour, and it turns out I need to put in 484Ohm (1x 330 Ohm, 4x 1 Ohm and 1x 150 Ohm), to get a Voltage of 1.99V.

What am I missing here?
Old 27th December 2010
  #2
Lives for gear
 
brianroth's Avatar
 

Egads, my Friend!

Let me begin simply. Eons ago, 10 mA into any given LED made it illuminate, and I think that will work now.

I will also assume 2 Volts drop across the LED...an ancient number.

Soooooooo...let us assume a 12 VDC power supply....


Now we have to calculate 10 mA into the LED and series resistor. We are assuming the LED drops 2V.

OK we need to drop 10 VDC across the resistor at 10 mA.

R = E/I

R = 10/.01

R = 1000 Ohms

Bri
Old 27th December 2010
  #3
Lives for gear
 
Kenton's Avatar
 

OK - first thing, you wouldn't put the LED's in SERIES. Each LED wants to drop 2V and if you have 8 of them you'd need a 16V supply and no dropper resistor. With a 12v supply you could put 6 in series if you really want to do it that way... The downside being that if one dies, either it will go open circuit and all the LEDs will just go out - or it will go short circuit and probably kill more of the LEDs due to them all receiving excess current.

The way I would do it is to put them in parallel - if you want only use one resistor, it's going to drop 10v (12v supply - 2v across the LEDs) with each LED drawing 30mA you will have a current through the resistor of 8 x 0.03 = 0.24A

V=IR and you get a value for the dropper resistor of 40 ohms (39 being the nearest value).
However VxI would be 2.5 watts dissipation so I'd choose a 5W to avoid the possibility of the resistor burning up.

It might end up cheaper to have individual dropper resistors for each LED :
10V @ 0.03A = 330 ohms, 1/2W

Quote:
if I measure the Voltage across the LED, I get 2.05Volt, and NOT 2Volt
you're being too literal, the difference between 2v and 2.05v is a tiny error compared to the tolerance of the components you're using. A 2.5% error is quite acceptable in this case.
Old 27th December 2010
  #4
Lives for gear
 

Hi
I would vote for neither series nor parallel for the whole lot.
Reasons:
All series, you don't have enough supply voltage for 8 approx 2 Volt LEDs.
All parallel, you should have a resistor for each as the actual voltage / brightness is quite variable AND 8 X 10mA =80mA which at 10 (12-2 Volts) is 0.8 Watts wasted. Resistors would be 1000 Ohms each.
Better 2 'strings' of 4 (about 8 Volts needed) with 2 resistors of 4 Volts (12-8) at 10mA which is 400 Ohms (preferred value 390 Ohms) which will 'waste' only 0.04 watts each.
Note that as Brian commented or at least alluded to, you do not normally need to run LEDS at as much as 20mA and 10mA is usually sufficient or less if you are using high efficiency types.
Matt S
Old 27th December 2010
  #5
Gear Addict
 
jaddie's Avatar
 

An LED is not a linear device, so watch out for variations in forward voltage drop as you increase or decrease current. Also, different color LEDs, and different brightness LEDs have wildly differing forward drops.

If you're experimenting to learn Ohm's Law, stick to resistors only, and don't obsess about the exact values. Even 1% resistors vary within stated tolerance. Your experiments will prove a principle, but will rarely verify high precision values. You don't have all the variables under control in real life.

Here's a pretty good source for LED info. Deals with difference in drop for color, series/parallel, etc.:
L.E.D Basics; gaining an understanding of how to work with L.E.D.s

One word when working with modern hi-efficiency LEDs. They are surprisingly static sensitive. I've blown them with a static discharge, which came as a surprise after decades of using the standard ones without issue. Also, never try to run an LED without a current limit of some kind, be it a resistor, or current-limited supply, or your LED could become an SED (Smoke Emitting Diode).
Old 27th December 2010
  #6
Lives for gear
 
datune's Avatar
Thanks for all the replies, much appreciated!

I am still trying to process all information given, it takes time, but sooner or later it will stike me ;-)
Old 27th December 2010
  #7
Lives for gear
 
Old Goat's Avatar
 

Quote:
Originally Posted by datune View Post
I am still trying to process all information given, it takes time, but sooner or later it will stike me ;-)
We hope not...
Old 27th December 2010
  #8
Lives for gear
 
datune's Avatar
Haha, no of course not. Just to clarify (just in case), I ment the wisdom to grasp all this knowledge will strike me, not the electricity ;-)

It's quite a lot, and my biggest problem is being to curious. I can't take much for granted, so if I read an explanation about what Amperage is, I smell that there must be more to it, so I go google, and next thing you know I am reading about coulombs and what not *sigh* This is how I basically spiral downwards from one issue to the next one...and really, all I want to do is make some blinky lights in a cool case (I'm kidding, but I do love LED's and glowing buttons, and rotary encoders with leds around them, oooh yeah baby!)
Old 28th December 2010
  #9
Lives for gear
 
Old Goat's Avatar
 

Quote:
Originally Posted by datune View Post
Haha, no of course not. Just to clarify (just in case), I ment the wisdom to grasp all this knowledge will strike me, not the electricity ;-)

It's quite a lot, and my biggest problem is being to curious. I can't take much for granted, so if I read an explanation about what Amperage is, I smell that there must be more to it, so I go google, and next thing you know I am reading about coulombs and what not *sigh* This is how I basically spiral downwards from one issue to the next one...and really, all I want to do is make some blinky lights in a cool case (I'm kidding, but I do love LED's and glowing buttons, and rotary encoders with leds around them, oooh yeah baby!)
Get yourself a copy of Stan Gibiliso's Teach Yourself Electricity and Electronics, and go through all the sections and quizzes. Made me semi-literate. Semi because I'm getting old and things don't sink in quite so fast. Saved my bacon a few times at work, too. I'm the de facto electrician, because a: I'm not afraid of it, and b: I'm the one old enough to remember how 60's stuff works. Simply. Don't touch that!heh

BTW, is wiring a 440 fuse box backwards a capitol offense? Cuz I'd like to kill the guy who did it...
Old 28th December 2010
  #10
Lives for gear
 
LeeYoo's Avatar
 

Forget most of the BS from the others.

Your leds need 2volt each. (*Not all leds are 2 volts!).
You only have 12 volts. Not enough for 8 leds in series.
Best to use two series circuits of [4 leds + 1 resistor].
4 leds in series will use 8volt from the 12volt you have got.
The remaining 4 volts has to be "used up" in the series resistor.
Most leds work from a few miliamps (dim) to a maximum of 30 miliamps (deadly bright).
10-20 ma is common.
Using ohms law, you can now work out the resistance.
The resistor has to be between 200 and 400 ohms.
Maybe use 330 ohm.
stike
Old 28th December 2010
  #11
Lives for gear
 
datune's Avatar
Argh, it's not working out as I want to.

I went back to using only a single LED (for testing purposes). So the LED's specs are 2V with a max. of 30mA.

I figured, to be on the safe side, I'll compute with 20mA.

Code:
R = 12V - 2V / 0.02mA = 500 Ohm
So I took a 470Ohm resistor, connected everything, but when measuring the Voltage across the LED I still get 2.1Volt. Is this normal? How much tolerance should I account for? I also measured the amps, which was at 21mA, so that seems to be fine.

Attached Thumbnails
Resistor math - more confusion-imageqqs.jpg  
Old 28th December 2010
  #12
Lives for gear
 

Hi
That is fine, no problem.
The LED will have 2.00 volts across it but only at one specific amount of current which will be different for another LED even from the same batch will be a little different.
If you look at ther full spec sheet for your particular LED there would be a graph showing brightness, current and a host of other info.
Keep going, you'll get there!
Matt S
Old 28th December 2010
  #13
Lives for gear
 
datune's Avatar
Hi Matt,

thanks.

So I finished my "blueprint", for a switchboard I want to build (basically I want to be able to turn on some synths and some LED strips in the studio), which allows me to not only turn things on, but also to group certain switches, so I only have to turn on one switch, and one or more devices get turned on.

Anyway, please have a look at the video, and tell me that those voltage variations across the 5 LED's is normal? One even has 2.33V.

LED Specs: 2V at 30mA maximum
Resistor used: 470 Ohm
Power Supply: 12Volt / 1A

I realize there is a lot of stuff for me to still figure out, but I need to get past this LED issue first. I am a bit freaked out that even though according to the calcuations (using Ohm's law), there is a Voltage difference of up to .33V, I mean that is almost 0.5 Volt.

Here is a picture of my breadboard, everything works as expected. I have basically two groups available if I want to, depending on the position of the switch.



Old 28th December 2010
  #14
Gear Addict
 
jaddie's Avatar
 

Try plugging your actual measured values into Ohm's Law, you'll see it's working fine. Also, your resistor tolerance is 5% of 470, not 0% of 500, so when you do your calculations, try varying the R value within 5% too. You might pull the resistor and measure it, then calculate using the measured resistance. Lastly, LED (and other diode) forward voltage drops also have a "tolerance", which is unknown in this case, but can also vary slightly with temperature. Some diodes have enough of this temperature variance to be used as temperature sensors. LEDs should not, but they do have some.

What you've discovered is the difference between circuit simulation/calculation and real world circuits. Simulation without taking tolerances into consideration almost always results in a difference between calculated and actual circuit measured values.
Old 29th December 2010
  #15
Here for the gear
 

leds are not resistors

Hi:

Ohms law is for resistors (a variation on Ohm's law is used for caps and inductors)

The led is an active device. It will always have it's characteristic voltage, around but not exactly 2v regardless of the current flowing through it. At upper and lower extremes of current the voltage will vary. When I say "around but not exactly 2v", it could be that by chance an led may have 2v on it's leads as all led voltages vary. But how would you know if the voltage was exactly 2v unless you had the most accurate voltmeter in the universe? Any reading from 1.9 v to 2.2 is normal, and some additional variation would not alarm me.

Also, resistors vary from stated value, but no much these days.

Life of led is reduced by running it near it's max allowable. Still you will get at least 10 years of regular use (that is, not on 24/7)

Also, it is customary for each led to have it's own resistor, because the led voltages vary. stacking say, 4 leds, and then using 1 resistor can result in uneven brightness, if not at time of build, then a ways down the road. Cheaper leds may have more variance. Other problem with stacking is that if one goes out they all go out.

The most right way to stack requires use of a transistor in a constant-current configuration - then the same current is always dragged through the stack even if led voltages vary or one or more begins to age and fade. In this way brightness is maintained into old age. But as with any stack, if one dies they all go dark.
Old 29th December 2010
  #16
Lives for gear
 
LeeYoo's Avatar
 

Some things not quite right in the last post.

A LED has a specific voltage, depending on the manufacturing process, type, colour, etc.
Not every LED from the same batch has the exact same voltage drop.
This is why we don't usually put them in parrallel.

The voltage over the LED is depending on the current!
I measured over a green led 2.0, 2.1, 2.2 volt with currents of 10, 20 and 30ma.

This all means that if we put LEDs in parrallel, we would have different brightness, because the led with the lowest specific voltage uses the most current.

If only a few LEDs are used in a piece of gear, or we deal with a low voltage, like 5 volt, a single LED/resistor is used.
We stack LEDs because of economics. Imagine LED strips for VU meters in a mixing desk, or a score board. 300 of them could use 12volt/6Amps.

Standard LEDs fail in time if you run them near 30ma. I have replaced many of them in commercial gear. But they will last decades at 10ma. Old LEDs are not as bright as new LEDs, not just because of age. The old ones never were that bright from the start, same as cheap LED packs. I have personally never seen a LED on 10ma go down in brightness, not even after years of use.

If you use the same LEDs in a stack, you will have little or no brightness variation. LEDs with the same part number are selected for brightness. You pay for that selecting. LED voltdrop does not change significantly over time, and so won't the brightness of the other LEDs in the stack.

LED brightness is depending on the current through them.
A resistor as a current regulator is fine if there is sufficient volt drop over the resistor, and if the supply is fairly constant. If we have a low overhead voltage, or a varying supply, an active current source is needed to keep the current/brightness constant.

Leo..
Old 30th December 2010
  #17
Gear Maniac
 

Another vote for putting 4 LEDs in series and running two such 'stacks', each with a resistor of about 180 ohms (nearest convenient value) for about 20mA current.

If you connect LEDs in parallel they will illuminate unevenly due to spread in their precise characteristics, and worse, the chances are that one will end up getting so much current it will blow, and the rest will go like dominoes. This is why they should always have their own resistor, one per LED or per stack of series LEDs.
Old 30th December 2010
  #18
Here for the gear
 

leds are contraptions and vary and fail variously

Leo - hello

assuming you were refering to my post, I stand by my statements - each statement is true, and the statements as a whole are also true.

Leds are not selected as they are commodity items. Tight control (not always the case) of manufacturing process substitues for selection.

stacking leds will cause problems over time as the leds degrade with age. Case in point is the leds in the Neve 8128 and V series consoles, Prism modules, etc. The leds are almost all used in stacks with a constant current circuit keeping current through the leds constant - an led is turned off by switch contacts shorting across the led leads - then no current flows in led, but through switch contacts. The constant current circuit adjusts to each switch opening or closing to keep same current flowing through on (unshorted) leds.

The leds on these consoles are notorious for failure, going dim or dark. when dim, the rest of the leds will still light to the desired brightness. when dark, that usually means the led has blown open, and then the whole stack goes dark.

SSL also used stacked leds w/constant current in the 5000 & 6000 series consoles, but the brand of led was better and problems were fewer.

The stacked/constant current circuit scheme was initially hatched as a means of preventing a little tick from being heard in the mix busses when leds switched on and off. In such circuits, the same current is always flowing even when all leds are off. In the Neve consoles this led to the console being too hot all the time, which then led to premature capacitor failure.

Stacked leds will eventually vary in brightness with age, even when a constant current circuit is used.

If "pro" audio gear is supposed to have greater reliability than "non pro" then stacking is not customer friendly. Thus the most judicious approach is one resistor per led. That approach does waste a little wattage, but the goal of "pro" gear should be reliability over energy efficiency.


as leds are available with good brightness in the 10ma range, wattage loss is rarely significant. In fact there are leds with good brightness at 2ma, but they cost a bit more.

cheers
Old 30th December 2010
  #19
Lives for gear
 
LeeYoo's Avatar
 

Hi RHSuperfly,

Thanks for the extra clarification.
I see what you mean with stacking LEDs. I have designed and build many VU meters with them. IMHO, failing LEDs is mostly the fault of the user. They want to squeeze too much out of the buggers, and then they fail, taking out the whole stack. When properly handeled (ESD), and properly used (current), LEDs last for decades.
Sure, the brightness can differ after many years, but I still don't see the difference in stacked LEDs with 20ma through them, or single LEDs with 20ma through them.
Leo..
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
rushton / Geekslutz forum
6
close/far / Geekslutz forum
0
Waltz Mastering / Geekslutz forum
4
davode / Geekslutz forum
3
PultecTweaker / So much gear, so little time
2

Forum Jump
Forum Jump