Hi everyone,

I wanted to buy a new fan with blue LEDs for my PC case, but I finally ended up with a fan AND four single LEDs.

I just want to know how to install these LEDs. I was told that they need DC 3.6V - 4.0V. I first thought to plug them on the same circuit of my new fan which runs at 12V, plugging only three of them in series, so the voltage would be divided amongst each, getting four Volts per LED. Am I right ?

But I am now wondering if I could simply plug them along with the POWER LED on the MoBo ? I guess in this case I would plug them in parallel, to get the same voltage, but what happens if they have lower amperage ?

Thank you,

The Smith

If you plug them in parrallel with the power led they will draw more current, rather than sharing the same.

4v is a bit high for led's usually a drop of 2v across each is expected so you might need a resistor to 'soak up' the extra potential (around 200 ohms should do)

Well, this is what the shop vendor said...3.6V - 4.0V.

And if I put them in parallel with the POWER LED, can the MoBo give more amps, as they need ?

And the long pin on the LED is the positive one I think ?

3.6-4.0 is normal for blue LEDs... they'll almost certainly draw 30mA each, so only 120mA total (if done in parallel of four)...

the long pin is positive, yes, and the flat side of the LED base is the negative (handy to know if you trim the lead wires)

you could run:

A) all four, in two banks of two, with a resistor for each bank (so two resistors in total), off a 12V line...
B) all four, in four banks of one, with a resistor for each bank (so four resistors in total), off a 5V line...

A)
resistor = (supply - voltage drop) / current
resistor = (12V - (2 x 3.6V)) / 0.03A
resistor = 160 ohms (use nearest, rounding up to a standard value)

B)
resistor = (supply - voltage drop) / current
resistor = (5V - (1 x 3.6V)) / 0.03A
resistor = 46.6 ohms (use nearest, rounding up to a standard value)

A)
12V====[160_resistor]=======[+LED-]=====[+LED-]===0V
12V====[160_resistor]=======[+LED-]=====[+LED-]===0V

B)
5V====[47_resistor]======[+LED-]===0V
5V====[47_resistor]======[+LED-]===0V
5V====[47_resistor]======[+LED-]===0V
5V====[47_resistor]======[+LED-]===0V


Obviously, you can bridge the lines at each end, and have a single cable going to the supply/ground

You could actually have one resistor for all LEDs... that would actually be a lot neater lol...

Wow, what a detailed answer...I'm sure not to plug them in the wrong way and break them!

I will make the 5 V circuit with one resistor...

Thank you very much and have a good day !

The Smith

Multiple LED's in parallel with a single current limiting resistor isn't ideal, at all, and could actually potentially be quite destructive. The LED with the lowest forward drop will hog current, and the resistance has also been decreased, so more current will also be flowing over the single LED. You can probably see where that's going?

Each LED should definitely have it's own current limiting resistor if you're intending on hooking them up in parallel.

Ok...If it is better to put a resistor to each branch in parallel, I Therefore will make this circuit proposed by Hardnrg:

12V=====[160_resistor]========[+LED-]=====[+LED-]===0V
12V=====[160_resistor]========[+LED-]=====[+LED-]===0V

Meaning two parallel branches of two LEDs in series, with a resistor of 160 ohms to each branch.

Again, thank you very much for your help !

The Smith

Is that because of the variance in manufacturing, like even if you had the "same" LEDs, you'd still have some random difference to the spec?

I'd normally run a resistor to each LED (hence the initial explanation), but thought it might be a bit overkill for just 4 LEDs...

Yep. The draw rate at whatever voltage varies just like any other component. I never design circuits that rely on accurate specs to work; that's just asking for a headache.

One current limiting resistor per LED is the safest and smartest way to go.

Hi guys,

I have installed my blue LEDs and they work perfectly. I now have a new idea...

I thought I could put purple LEDs under my case to make a kind of aura or halo around my computer. I after thought that I could make the aura turn around it, either by putting a small motor which will make the LEDs turn, or with many LEDs which will light one after the other..

But is there a simple electric circuit I can build that is able to light one LED at a time like I want ? ( I studied High-pass and low-pass filters made of a resistor and a transistor at school, so I thought a circuit like this could maybe exist...

The Smith

Hello Mr Smith,

Good to hear that you have got your LEDs working OK.

That mod idea sounds like it could have a nice little effect. A simple circuit can be built which should achieve what you want with two ICs and a few external components. The circuit should blink up to 10 LEDs in sequence repeatedly. It can be expanded for a larger sequence, but it does become a little more complex. If you wanted more complex effects you could learn how to program a microcontroller

Here is a schematic:



I haven't tested it, only simulated. It should work though. Operation of the circuit is quite simple. I will attempt to explain the best I can.




C4 is a power supply filter capacitor. It should be rated at the very least 16v, capacitance isn't all that important. 10uF to 100uF should be fine, you can probably even get away without it if cost is an issue (although it should only be a couple of pennies).

C2 and C3 are decoupling / bypass capacitors. Unlike C4, these should be placed as close as you can to their respective ICs. Their purpose is to filter smaller fluctuations in the power supply. 10nF or 100nF should work I think. You may actually be able to get away without these for such a simple circuit.

The 555 timer, IC2, generates a square wave output ( something like ____|-----|____|-----|____ ) on pin 3. Pulling pin 4 high, to +12v in this case, takes the IC out of reset. C1 starts charging via R2 then through D1. R2 slows down the charging of this capacitor by limiting the current and this creates a 'time constant'.

Pin 2 (trigger) monitors the charging of C1. When the voltage is smaller than 1/3 of the supply it puts the output, pin 3, in to a high state of +12v.
Pin 6 also monitors this voltage, and when it reaches above 2/3rds of the supply it will put the output pin in to a low state of 0v and then connects the discharge pin, pin 7 to 0v. By connecting the discharge pin to 0v, the capacitor C1 starts to discharge through R1. It doesn't skip R2 like when charging, because the diode only conducts in one direction, so the discharging also happens slowly because discharge current is limited.

You may be able to see that this creates an endless loop of C1 charging and discharging, and thus flipping the output pin from 12v to 0v and vice versa at fixed periods. C1, R1 and R2 control timing.

The output pin of IC2 is connected to the clock pin of IC1. IC1 is a simple counter IC and is taken out of reset by pulling pin 15 low, to 0v. The clock input is enabled by pulling pin 13 high, to +12v. The outputs Q0 to Q9 should all start off in the low state. When a pulse is received on the clock pin, on the falling edge it will take the current output low and then toggle the next output high. So, starting off at Q0, it will take Q0 low and then Q1 high. On the next pulse it will take Q1 low, then Q2 high. When it reaches Q9 it simply starts the sequence over again, and so creates a chasing effect.

The small transistor circuit on Q0 should be repeated for each output pin (Q1 - Q9). This is here because the datasheet says the output pins can only handle up to 8mA of current each. A transistor needs only a small amount of current provided to the base to pass a large amount of current through the collector / emitter. The resistor R5 pulls the transistor low 'by default', and R4 limits the current to the base when the output pin goes high. The transistor acts like a electronic switch. When it has no base current, it is off and no current can pass. When it has some base current the transistor is switched on and current can flow from the LED to ground, thus lighting it up. R3 is the LED current limiting resistor.

Pins of ICs are easy to determine. Pin 1 will be indicated by a small circle / dimple on the casing. Opposite this will usually be the highest pin (eg for an 8 pin 555 opposite pin 1 will be pin 8). Then just count down / up for pin 2 - 7. For example:

1 8
2 7
3 6
4 5

As for the transistor pin out, if you hold it so that the flat side is facing toward you and pins going down toward the floor, then the pin on the left will be the emitter, the middle pin the base, and the pin on the right the collector. The emitter in the schematic symbol is represented by the small arrow pointing towards ground, and the collector is the pin with the LED connected to it. You should be able to work out which is the base

Pin outs can sometimes vary between different manufacturers, so it may be wise to search for data sheets just to make sure!





The time the 555 pulses last can be calculated by the following

Thi (s) = ( 0.7 x (R2 x C1) ) / 1,000,000
Tlo (s) = ( 0.7 x (R1 x C1) ) / 1,000,000

So for the values in the schematic (10k / 10k)

T = 0.7 x (10,000 x 47) / 1,000,000 = ~0.32 seconds.



You may want to include variable resistors to tweak the time. I don't recommend you simply replace R1 and R2 with variable resistors. You should leave R1 and R2 but reduce their value (~680R to 1k ish), then add a higher value (~100k) variable resistor in series with each resistor. So one VR between 12v and R2, and the other between pin 7 of the 555 and R1.



Hope that helps and wasn't too confusing!

Ouch, this is more complex than I thought !

I will study this and I will learn how to solder, and I will make this circuit. I made my circuit of blue LEDs with a small board which has small holes to plug the wires in them without soldering, but this time I don't have the choice...

And if I want the LEDs to fade slowly, should I put a transistor somewhere ?

Thank you very much for your complete answer...I should have taken you a lot of time to write this. And where did you get the circuit picture ? Did you created it with a program if you say that you simulated it ?

Ok...I don't understand very well what are the IC1 and IC2...Does the "555 timer" is the name of this piece ? Because I will need to go buy all these pieces at the electronics shop so I must know exactly what are all the components.

Thank you again,

The Smith

I think I know the type of board you speaketh of. You should still be able to construct this circuit on one of those, soldering would be better though really. It's fairly easy to do!

For fading... I think you might be able to put a capacitor from the transistors base, after the resistor, to ground. Not completely sure on that though. It might not be a particularly accurate fade either.

The schematic I posted was drawn with CadSoft Eagle. It's a schematic / PCB design software. It's got a bit of a learning curve to it and there are other things for drawing PCB layouts and schematics you may find easier to use.

The simulator I used is called LTSpice or SWCADIII. It's quite complicated. I don't even know how to use many features I do remember back in high school we had something called Crocodile Clips, which was awesome and would probably be more helpful to you! I'm not sure if that's freeware though.




Sorry, IC2 is an "NE555N" or "LM555N" and IC1 is "CD4017".

NE555N - http://search.digikey.com/scripts/DkSearch...amp;x=0&y=0
CD4017 - http://search.digikey.com/scripts/DkSearch...ds=cd4017%20dip

If I can't make them fade and light slowly, I don't think it would be like if the aura was turning around the case...

Maybe it would be better with a small motor. Or else I just don't make the aura turn. But I will probably make your circuit though...

I'll think about this and when I will have finished this mod, I will post pictures so you will see what's the final result.

Edit: I thought that if I make them light fast, it will resemble as a turning aural. It's just because I had in my mind the image of a very slow aura, like 5 sec/rotation, for 6 leds.

Can you explain me how to plug these integrated circuits ? On your schematic, does IC1 has 11 pins on one side and 3 on the other ? On the site you gave me, it has 8 pins on each side.

I think it will probably look OK without a fade if you do it quick enough, but then it probably does have something to do with the area you want to light. If it's fairly small and the LEDs are close and evenly spaced, then each LEDs lighting area will overlap. This might look OK. I'm fairly sure replacing the resistor from each transistors base to ground with a capacitor, about 100uF 16v will make a crude turn off fade that may work well enough. Otherwise, personally, I would do it with a microcontroller. The circuit would be simple, but you would need hardware to program the chip and you would need to write code.... so you may find that a bit overwhelming really.


The 555 has 8 pins, in reality the IC will have 4 pins on each side. The 4017 has 16 pins, and again this will be half on one side, half on the other. The symbols in the schematics are just to make the schematic easier to draw / read. Each pin on the schematic is labeled with the pin number of the IC, so all you have to do is translate each pin labeled on the schematic to the pin on the physical package. Pin 1 on ICs is usually identified by a small dot or dimple next to it.

The pin layout for a 555 is simply



The 4017 is much the same

Oh boy, I hope they sell these pieces at the electronics store because the shipping is quite high for 2 pieces of 0.40 $ each...

Ok, I'll be able to make this now, I understand how to plug the pieces. If I have another question, I will let you know.

Thank you again for your explanations !

Edit: As for my LEDs positions, I Thought to install one on each corner of my case and the other two in the middle of the sides.

The 555 is a very, very common chip. They'll almost certainly have those at a local electronics store. Not sure about the 4017, but you could search ebay "CD4017" to see if there's anything on there with a cheaper postage.


edit - no problem



Oh, and if this is your first time working with ICs and electrolytic capacitors.... both of these are likely to explode if you get the power supply connections reversed. From experience, both make quite a loud BANG when they do, and in the case of ICs the plastic package may do quite a spectacular little flying stunt. I would advise not having your head hanging over the circuit when you first apply power, just in case something does decide to explode in your face

I already looked on Ebay for these pieces. The shipping is approximately 5 $ ( I live in Canada )

Can you explain a bit more how to connect this circuit on my PC power supply ?

And is R3 here to limit the voltage through my LED ? So it's resistance will depend on my LED's voltage needs ?

You can use one of the big 4 pin molex plugs. Probably the easiest way would be to buy one of those 4 pin molex splitter / extensions. It will have one side which plugs in to a power supply lead (male connection) and the other end (female plug) will plug in to a hard disk etc. You can cut off the female plug. The Yellow wire is +12v, the Red wire is +5v and the two black wires are ground.

R3 is indeed there to limit the current through your LED. You should calculate the resistance as explained by hardnrg before. If you will only be using a single LED per transistor rather than two or three in series, then you can use a 5v power supply to reduce the wasted heat dissipated in R3. If you use a 5v supply you should also change R4 to ~2.2k and R5 to ~22k.

Other questions... But this time I think they are the last ones !

What is IC1P ?

What is X_Power-1 and 2 ?

For plugging it in my power supply, can I put both +12V circles in series ?

IC1P is just indicating the power pins for IC1. In case it's difficult to read, pin 8 is ground and pin 16 is +12v.

X_POWER is a screw terminal connector. You can just think of it as power supply input. It doesn't necessarily have to be a connector of any kind.

All GND points should be connected, as should both of the +12v points. These are just symbols which save routing supply wires all over the schematic. It can become rather messy if you do it that way!

OK thank you very much !

I think I'll be ok now...

LM555N is EVERYWHERE, and should be very easy to find. Not sure about the 4017 timer chip.

Have your tried MultiSIM 8 or 9? It is both a circuit design suite and simulator in one. It has its quirks(I use ver.8 since I got a copy through one of my circuit design classes and have never tried 9), but is very easy to use and easier to get into its advanced features. The library it comes with comes with just about every component from common basic IC's to complex pre made circuits to import into your own work, and each component can be customized for the specific values needed(temperature tolerance, failure rate(components can randomly fail just like in real life), voltage tolerance, etc).

The simulation is also straight forward, and the timing and such can be adjusted from very slow to watch which parts break down first under extreme conditions(like improperly connected LEDs ) to realtime simulating, just like with some of the bigger programs. At first I hated it for its small quirks(like having to stop/start the simulation after changing parts or the results sometimes dont change), but have come to find it pretty handy overall.

Maybe I will try to get one of those programs someday...

Hi,

I bought all the pieces I need to make this circuit, except IC1, I ordered it. I have two questions about plugging.

1st: The screw terminal connector. When you say that it is a power supply input, how do I plug it in ?

2nd:Which pin is the base current on the Q1 transistors ? And which is the positive ?

Thank you,

The Smith

It doesn't appear to be free




I'm not sure of the question. 12V goes to 12V, GND goes to ground. You will probably have to make up a 4 pin molex plug with two wires, one 12v (the yellow wire) and one ground (the black wire next to the yellow wire). Put 12v wire to 12v point on your board, put ground wire to GND point on your board.








The emitter goes to ground. The collector to the LED.

Oops...I didn't re-read your post before asking the question on the transistor...Sorry for that.

Edit:This is what I have done so far for the power, as you said, but I don't know what to do with the X_Power:

You just need one more ground connection to the GND point near X_POWER and it looks like you're all set!

Euh...yes, I just forgot to draw it, I was in a hurry to go eat.

But what do I do with the screw terminal connector ? I bought a small piece to be soldered on my board which leads to two screws. I thought I would need to put this as the X_Power 1 and 2, but are you saying that X_Power is not a physical piece ? Nothing is to be placed there ?

It can be a physical piece. I didn't realise you had bought the connector.

Molex wires goes to screw terminal, join all power connections on the board.

edit: sorry, that image was slightly wrong!

edit: this one should be right-er

Ok, now I understand !

I directly plug all the +12 V and Gnd to the molex connector OR I plug the X_Power, joining the +12V together and the same with Gnd.

I will be refunded for the 1 $ piece I bought.

Thank you very much for your patience, Markiemrboo.

The Smith

You can of course solder the 12v and GND wires from the molex to some point on the board, then join all of the boards 12v and GND to those points... just like the screw terminal way, but without the screw terminal. It would be neater and easier than having lots of separate wires coming out of the molex. Or perhaps that is what you meant.

No problem anyway. I hope we will get to see pictures and / or a small video (webcam?) of it working!

Hi again,

I received all the pieces I needed for the circuit. Just want to be sure of the pins of the ICs. This week, we got a demo at school of a electrolytic capacitor exploding, and I don't want my ICs to do the same !

So, I am a bit confused about the small dot to identify the 1st pin. Which of them is it ? I think it is the round one, not on the side, but I wanted to be sure.

And for the variable resistor, too. It has 3 pins, one in the middle which is not aligned with the other two. Which pin is what ?

Thanks in advance!

Oh, in this case it'll be the U looking one.





edit - The middle pin tends to be the wiper. This might help you understand a bit better. http://www.kpsec.freeuk.com/components/vres.htm

Oh...A luck I asked...

And on the circuit, black dots between lines represent intersections, and when there is no dot where lines cross, I guess they should not touch ?

That's right

And I will put only 6 LEDs for the moment. Can I plug them on any Q0-Q9, or there is specific pins I should let unused ? And will they flash at regular intervals or or there will be "holes" in the flashing ?

It should count from Q0 to Q9 sequentially. There will be a gaps if you don't fill it with 10 LEDs. For testing I would just put them in starting from Q0.

Hey,

The circuit construction is going very well. I decided to plug the LEDs on 5V, as you explained. But can they use the same Gnd as the 12V part with the ICs ?

Edit:And why are the pin not in order with the QXs ?

Good to hear. It should be fine to use the same ground.

edit - The pins are up to the IC designer. I guess it was a design constraint / choice.

Hi,

bad news this time... I finished to build the circuit, but it doesn't work...

I tried to figure out why.

1st: There's only one pink LED that is always on, and it is not flashing. When playing with VRs, I got 3. Can it be that this VR idea doesn't work and I should use normal resistors instead ?

2nd: Are you sure both 5V and 12V parts can use the same ground ? Because if the ground is the negative(I think), the electrons in the wire are flowing from negative to positive, that is to say the inverse way of the current convention, which is the movement of the positive charges. So I'm not sure it's working, but if you say yes, you are probably right, I trust you...

Here's some pictures of my circuit: (Yes I know that I used 1/4 of the whole board and I could have used all the room, but this way I have room left for future circuits.)

Hi,

It sounds like the 555 isn't oscillating as it should, or perhaps at a much slower frequency. It is, unfortunately, hard to tell what's wrong from the pictures of your circuit. A potential source of error would be the VRs. Have you simply replaced the charge / discharge resistors, R1 and R2, with VRs, and connected the wiper (the middle pin) to either end (either of the two remaining outer pins)? What value did you use? 10k? If so, set them both to 10k and make sure they are wired properly.

And yes, the 5v and 12v can use the same ground. If you open your power supply, both the 5v and 12v rail will share the same ground anyway.

I replaced charge discharge capacitors as you said on the first page of the thread:



I used 1K resistors and 100K VRs...Uh-Oh... Is this the error ? Was it supposed to be 10K instead of 100K ?

Besides that, I re-check my connections for the Xth time and we'll see...

Edit: I knew the pictures wouldn't tell you nothing, I just send them to show you my work after learnig how to solder...And I used this freshly-learned skill to make my vdroop mod lol

Ok cool. The value of the VR shouldn't matter too much, so don't worry about that. Could you confirm that they are connected like the following:



If you have a multimeter then it could be useful to check a few things. Set to voltage measurement, DC, and put the black probe to ground.

With the circuit powered.

Red probe to pin 4 of the 555, this should read 0v.
Red probe to pin 8 of the 555, this should read 12v.
Red probe to pin 1 of the 555, this should read 0v.
Red probe to pin 15 of the 4017, this should read 0v.
Red probe to pin 13 of the 4017, this should read 12v.
Red probe to pin 16 of the 4017, this should read 12v.
Red probe to pin 8 of the 4017, this should read 0v.
Red probe at either pin 3 of the 555 or pin 14 of the 4017, set the VRs both to 100k. You should see the voltage reading spike up to 12v and then back down again over and over.
Red probe at pin 7 of the 555, VRs set both to 100k. You should see the voltage slowly climb and then suddenly fall to 0v over and over.
Red probe at pin 6 or 2 of the 555, VRs set both to 100k. The voltage should slowly climb and then slowly fall, but not necessarily to 0v. This should repeat over and over.


With the circuit off, set the multimeter to measure continuity, or if it doesn't have this feature, resistance.

Measure between pin 3 of the 555 and pin 14 of the 4017. It should beep or read close to 0, or both.
Measure between pin 2 of the 555 and pin 6 of the 555. It should beep or read close to 0, or both.


You could also try a 100nF cap (if you have one spare) between pin 5 of the 555 and ground.

Ok...I found the problem. Thanks for the schematic.

I used this website to know how to plug the VRs http://www.kpsec.freeuk.com/components/vres.htm , and I used them as a rheostat, connecting only the wiper and one pin.

I now see that you mentioned it at the end of page 3 but I was so sure about my VRs connections that I simply didn't think that it was what you meant, and understood that the wiper and another pin should both be conencted to the circuit.

Now that I saw the small picture of it in the schematic, I know that I need to use it as a preset, as it is showed on the website. So I'll fix this problem and if I still have some, I'll check what you said with my multimeter(Yes, it has the continuity function).

I'll tell you if everything is ok, and I'll post pictures of my case very soon !

So long as the wiper is connected to something it should be OK. If you connect the wiper to the "free" pin, if the wiper momentarily comes off of the track when adjusting it, it will still act as a fixed resistor (in this case of 100k), rather than momentarily going open circuit. It probably wouldn't matter too much in this circuit if it was just the wiper and either side connected.

Doing those quick checks might show something hopefully. Make sure the circuit doesn't have power when checking continuity.

Looking forward to some pictures when we get it working!

Oh my! Sorry about this. I think I may have just figured out what is wrong.

Pin 4 of the 555 (reset) should be pulled high. Connect pin 4 to 12v rather than ground.
Pin 13 of the 4017 (enable) should be pulled low. Connect pin 13 to ground rather than 12v.

Very sorry! I think it should start working after that.

Hey, good news !!

It's working !!! There's only one LED not flashing but this time it is probably a wrong connection of my part; I'll check this.

I'm so happy !!!

But to make them turn the fastest they can, I need to decrease the VRs ? They are not turning so fast. I'll see when they will be installed in the case, but if it's not fast enough, how could I change the resistors in order to get a faster speed ?