# [Solved][Follow-up question on page 2] Resistor placement - 30 LEDs

Can someone quickly tell me whether these two schematics are equivelent? I have a really space sensitive project and want to use as little components as possible. I feel they should be equivilant, but I am worried, becouse I see so many circuits which have multiple resistors in it, so I am uncertain.

Thanks

p.

Nope, they are not equivalent.

-- The Gadget Shield: accelerometer, RGB LED, IR transmit/receive, speaker, microphone, light sensor, potentiometer, pushbuttons

nope the second one will draw less current than the first and, unless the LEDs are exactly the same, they won't glow with the same brightness

ok. i get it. in the first circuit the resistors are in parallel, so the overall resistance is less than 330 Ohms (without getting into the math or anything) so the first circuit will draw more power.

right?

so in order for the circuits to be equivilant I would need to use higher value resistors in the first circuit, or a lower resistance in the second.

* ok, so another scenario:

I hook up one of these sensors + breakoutboard to my arduino: http://www.sparkfun.com/products/9453 The board includes an IR LED with an 100 Ohm resistor attached. As its only one sensors+board everything is ok.

If I however take say, 10 of these sensors, and hook them all up to the same powersource/ground, then I have a problem, becouse the overall current increases, as the resistors are now in parallel. Correct?

(concidentally there is no way of increasing resistance on these boards without decreasing the resolution of the sensor output --- stupid design --- but anyway thats not the point.)

fkeel: ok. i get it. in the first circuit the resistors are in parallel, so the overall resistance is less than 330 Ohms (without getting into the math or anything) so the first circuit will draw more power.

right?

Wrong, the first circuit has three independent series circuit, all they share is the power and ground. In each circuit, each led is limited by it's own resistor. So the three led/resistor series CIRCUITS are in parallel, not the three RESISTORS. To say it differently, if you changed the value of just one of the resistors in that first circuit only one led's brightness would change, the other two leds would remain the same brightness as before. Therefore they aren't acting as a single parallel circuit, but rather three independent circuits wired to a single power supply.

So in order for the circuits to be equivalent I would need to use higher value resistors in the first circuit, or a lower resistance in the second.

The two circuits are fundamentally different, regardless of the size of the resistor(s) used. However, yes, if you wanted all the leds in both circuits to have the same brightness you would either have to increase all three resistors size used in the first circuit or lower the one resistor size used in the second circuit.

* ok, so another scenario:

I hook up one of these sensors + breakoutboard to my arduino: http://www.sparkfun.com/products/9453 The board includes an IR LED with an 100 Ohm resistor attached. As its only one sensors+board everything is ok.

If I however take say, 10 of these sensors, and hook them all up to the same powersource/ground, then I have a problem, becouse the overall current increases, as the resistors are now in parallel. Correct?

No, The total current drawn from the power supply would increase only because there are 10 independent series sensor/resistor circuits all consuming current at the same time. Resistors only are considered being in parallel if both their leads are wired together (at both ends), two 10 ohm resistors wired in true parallel end up having a real resistance value of 5 ohms. Your first led circuit example has the three resistors wired together only at one end, so they are not acting as a single resistor, but rather as three independent current paths. Total circuit current is the total of each branch circuit.

(concidentally there is no way of increasing resistance on these boards without decreasing the resolution of the sensor output --- stupid design --- but anyway thats not the point.)

so in order for the circuits to be equivilant I would need to use higher value resistors in the first circuit, or a lower resistance in the second.

They still will not be equivalent. The first circuit will work properly even if the LEDs are not identical. The second circuit with a lower resistance may work acceptably but only if the LEDs have nearly the same characteristics, which is not likely.

Don

I use this website: ... to calculate LED resistor values. It has options for calculating the proper resistor for either of your option (1) or (2).

Which just goes to emphasize that you shouldn't blindly trust what you find on the web. Not recommended.

Don

If you want to minimise component count, place your three LEDs in series and use one resistor to feed the chain. Because LEDs are "current" driven devices such an arrangement is sound providing the rated currents of the LEDs are identical. You might see minor voltage variations across each LED but that's of no concern in such a circuit.

There are enough differences in LEDs that hooking them up in parallel, one will draw more current ending up burning it out faster than the others and once that happens, because your circuit's current limiting resistor was designed to supply current to 3 LEDs the amount of current draw on the remaining two LEDs will increase. This increase will quickly burn out your remaining LEDs in a snowball effect.

The only time it can be recommended to drive LEDs in parallel is if you have an LED Driver chip designed to drive LEDs in parallel, because they can adapt to a fault situation with one of the LEDs. The amount of board space this driver chip will take will render your concern of board space mute. Go with multiple resistors.

Most LED drivers however are designed to run multiple LEDs in series, the same as these good people are recommending you to do. It's easy to keep all the LEDS at say, 20mA running them in series with one resistor. The only draw back is you need a higher voltage power source to do this, which wouldn't work if you're running, say 2VDC forward voltage LEDS and want three coming from a 5VDC source.

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=4608X-1-151LF-ND

take a look at the datasheet on the bussed resistor array above. You may want to consider this type of component as an option.

*

I am building a more complex circuit, so I am trying to brake it down in little problems, to make sure I understand them. The LED example was just something I made up, to understand how I should be placing the resistors in my circuit.

I usually construct circuits on a more or less trial and error basis... I have used leds in parallel a lot... I am beginning to understand that "it works" is very relative...

*

Anyway thanks for everyones input. This has been an educational thread :-D

p.

I usually construct circuits on a more or less trial and error basis...

When I was first learning electronics I found I learned best by reading any and every schematic drawing I could lay my hands on, not easy in the pre-internet age. Not that I understood them all, but by studying them until I figured out why each and every component was there, did I start to better get a handle on things. Learning by building 'smoke bombs' can be a little frustrating and expensive way to go. :D

So with each months arrival of say Popular Electronics, I would study every project schematic regardless of if I thought I would ever build it or not. I bet I only built one project over every 100 I studied. I would also research the parts specified in the projects by looking them up in the Allied Electronics catalog, that way I slowly got to understand the specifications that every electronics component has and how that related back to the project the author had used them in.

Maybe not the fastest way to self teach, but it was a simpler slower time back so many decades ago. Now many have expectations of a internet-fast learning pace that sometimes leaves gaps and lack of fundementals that really are needed to master electronics. I suspect that reading a lot of code first as a learning method also might be better then just jumping in and trying to create programs from scratch.

Lefty

KE7GKP: There are probably others. It would be interesting to see where Arduinos and clones are advertised most.

Make Magazine has been pushing Arduinos like crazy, though they're a little more 'crafty' and all around DIY.

http://makezine.com/

http://www.makershed.com/SearchResults.asp?Cat=43

I am beginning to understand that "it works" is very relative

Yes there are several aspects to this:-

1) How repeatable is your design, if you were to make 1000 would they all still work? I know you are not going to make that many but if you put the design on the net maybe 1000 others will try.

2) How reliable is your design, are the components over stressed? How much understressed are the components, this is known as derating and as a rough rule you should ever use a component at more than 80% of it's maximum rating.

3) How stable is your design, will it work under a wide range of temperature, will it work for long periods without overheating? Are there any parasitic oscillations, these might not cause immediate problems but will cause long term ones.

4) How prone to interference is the design, it might work in your room but can you operate it anywhere?

There are also other things to consider but that will do for a start.

Hi Fkeel

You have had many responses explaining why the two circuits are not equivalent.

I am assuming that you would want to control each LED independantly.

If space is really at a premium, and your require several resisitors of the same value, investigate SIP (Serial Inline Package) Resistor components. They come in various packages and configurations. A 8 resisor SIP with common connection takes up about as much space as (2) 1/4W resistors. They are not that expensive.

Regards Bill

If I have understood everything correctly, I see a conflict when handling a large number of LEDs:

Parallel LEDs is dodgy, becouse if one fails, then all are bound to fail relativly soon.

LEDs in series requires fairly high voltages. Assuming one LED needs 1.4 Volts, I can power maximum 3 LEDs with 5 Volts.

Correct?

Ok, I would like to power 30 LEDs which require 1.4 Volts at 20 mA. I would like to power them with 5V.

My reasoning is that even if I where to use 12 volts, I would end up using parallel circuits for this. So I thought I might want to combine a parallel & series set up:

I have 10 parallel circuits which each have 3 LEDs in series. The LEDs in series each have a 47 Ohm resister. I am limiting the whole system with a 1.5 Ohm resistor, but I am not sure if I really need that one. (I got lazy and didnt want to make the drawing any uglyer by adding arrows over all the LEDs. They are all LEDs. No regular diodes involved)

This seems terribly complicatedâ€¦ Any comments on how I am going about it?

Another problem I have: I was initially planning on powering this with the arduino. However If I am right I will be pulling 600mA which is way to much, correct? Even if I used an external powersource instead of USB power, I still could not get the amount of amps out of a digital pin. What about the 5V out pin? Can that Pin handle 600mA? Is there any simple way of powering this circuit, except for using an external powersupply and transistors for turning it on and off?

Ok, thats it for now. I feel like I am confusing myself. If anyone can take a look at this and give me some advice it would really be apreceated.

Regards

P.

Is there any simple way of powering this circuit, except for using an external powersupply and transistors for turning it on and off?

No, you will require an external +5vdc power source rated at or above 600ma. You will have to use some form of external switching for the led array as it's current draw is too much for direct control from a output pin, however a simple switching transistor is easy to use. Your arrangement of leds in series/parallel is fine, however that lone 1.5 ohm resistor serves no useful purpose, lose it.

Keep in mind that different color LEDs have different forward voltage drop values and your circuit would not work with say blue leds.

Parallel LEDs is dodgy, becouse if one fails, then all are bound to fail relativly soon.

That is a false conclusion. There is nothing about one parallel led string failing (somehow?) that would effect the other strings at all.

PS: By my estimation (if using red leds) each parallel leg will draw around 10ma for a total of 100ma, how did you get your 600ma estimation?

Lefty

retrolefty: That is a false conclusion. There is nothing about one parallel led string failing (somehow?) that would effect the other strings at all.

dshay: There are enough differences in LEDs that hooking them up in parallel, one will draw more current ending up burning it out faster than the others and once that happens, because your circuit's current limiting resistor was designed to supply current to 3 LEDs the amount of current draw on the remaining two LEDs will increase. This increase will quickly burn out your remaining LEDs in a snowball effect.

The above is only true, if I have one resistor limiting the current for each parrallel LED. If I have a limiting resistor for each parallel circuit, then I dont run into this problem, correct?

*

Thanks for teaching random people on the internet electronics, Lefty :-)

EDIT: I am using IR LEDS. Each LED requires 20mA Hm. That means that I need 200mA and not 600, correct? (I was thinking 20x30 = 600 but it doesnt work like that, right?)

KE7GKP:

Parallel LEDs is dodgy, becouse if one fails, then all are bound to fail relativly soon.

That is a false conclusion. There is nothing about one parallel led string failing (somehow?) that would effect the other strings at all.

@Lefty, the original reference was to LEDs hard-wired in parallel with a single current limiting resistor. Since LEDs are not exactly perfect they will tend to draw slightly different currents. Under worst-case an LED could hog the current and if that causes it to burn-out open, then too much current will be presented to the remaining LEDs, causing them to fail prematurely, also. NOT a good circuit design practice.

I was just commenting on the latest drawing he posted where there was no sharing of resistor. Except for that strange 1.5 ohm one. ;)

Thanks, KEG7GKP & Lefty

My project is going forward, thanks to you guys.

EDIT:

Anyone following this with similar questions: I just discovered this here

http://ledcalculator.net/