LED Flashing Array! Please Help!

[mindctrl]

Hello everybody,

This is my first post and hopefully will be the start of a good relationship with the board. I'm hoping that somebody here can help me, seeing as this looks like a relatively simple project.

I would like to build a flashing LED array. Basically when a switch is thrown, a series of, say, 20 LEDs will begin to flash on and off, all at once. The flash rate would be a constant, require no modulation, and be fairly fast(11-14hz). I am going to be dealing with high intensity infrared LEDs, so I figured I'll probably need some capacitors for a quick charge/discharge time. The whole system will run on a 12VDC, but it will be automobile so I'll need to have some kind of a resistor to keep the current flow smooth. I'm not a complete beginner, I do know my way aorund a soldering iron and can figure out a circuit diagram. If anybody can give me some help, I would really appreciate it.

Now here's the kicker to the whole shebang: I need to somehow be able to adjust the speed of the flashing LEDs fairly precisely(11-14hz) I'm talking maybe a 1hz -/+ room for error, and even that is undesirable. I need it to be exact as possible.

Here's what I thought I might need: LEDs, a timer chip(555?), capacitors, and resistors. I know that sounds pretty basic but I really have no idea how to go about this. Project summary: Array of 20 infrared LEDs that flash at a certain hertz when turned on, remain flashing constantly at that speed until turned off. Thats it.

Any help is greatly appreciated and I look forward to hearing from you all. Thanks again and have a good weekend.

--mindctrl

 

[TheOne]

It sounds like something (maybe uP) with a crystal based clock will be needed if you want that sort of accuracy.

What will be your step size? 1Hz, 0.1Hz?

 

[mindctrl]

step size would be 1hz. woudl a crystal clock have the accuracy I'm looking for? Never really heard much about them

Thanks,
mindctrl

 

[chemelec]

I'm quite sure a 555 would be stable enough at that low of a frequency.
You just need to use a Good Quality Capacitor on pin 6/7. Perferably Not an electroyltic type.

 

[TheOne]

If it is a stable crystal clock reference with temp compensation it should be OK. The crystal reference would be much higher and then divided down to get the 11 - 14Hz

 

[mindctrl]

Sounds great to me. Do you think I should go with a 555? I've never had any experience with assembling a circuit with an IC. I've got the basics down in theory(solder in a socket, etc) but if you guys could point me to maybe a diagram? That would be awesome. Thanks everybody.


Also, one last thing. Would it throw too much of a complication in the works if I wanted to add some way of determining how fast the LEDs are blinking? Like maybe an LCD display? I would probably have to have a counter circuit with a controller for the LCD display, so the answer is probably yes, it's too complicated.... but I thought I might ask. Thanks.

 

[gerty]

If you go with the 555 you would get a finer resolution with a multi(10-15) turn pot...

 

[TheOne]

If you are very lucky you will get by with the 555 if you willing to tweak every now and then due to temperature changes, component tolerances, aging etc. I personally won't use RC time constants for accurate timing applications, crystals is much better for that. Would you want a wristwatch with a adjustment that you need to tune on a daily basis depending on the temperature? :wink:

 

[mindctrl]

OK let's say I want to go with crystals. What crystal should I use, and more importantly, how does the whole circuit look? Is there any diagrams around I could take a look at?

 

[TheOne]

You will need a type of programmable divider fed from a crystal oscillator or a microprocessor to be able to give the required frequency from the selected division ratio.

If you are not comfortable with micro's you could use a CMOS programmable divide by N counter like the 4059

https://www.electro-tech-online.com/custompdfs/2004/12/cd4059.pdf

which is fairly easy to use.

Crystal oscillator could be made by using a 4060

https://www.electro-tech-online.com/custompdfs/2004/12/cd4060.pdf

which also have some dividers in it, that you could use to do the prescaling, and then feed it to the 4059 for further division. Fig 13 shows how to connect a crystal to it.

You will have to look at vendors to see what standard crystal frequencies are available to work with your division ratio's. Some vendors like Digikey.com sells programmable crystals if you want a non-standard frequency for your application. You can browse their online catalog at www.digikey.com

 

[mindctrl]

Thanks!

This is some great stuff to look through. I hate to ask this, but I feel a little lost. Does anybody have perhaps a step by step tutorial or can give me a straight circuit diagram for this? I got a little lost when you were talking about feeding through the 4059 and dividing and such... I am also not familiar with prescaling or dividers. Sorry if I sound a little out of place, but I'm just starting out. Thanks again.

 

[Nigel Goodwin]

It's by far simpler to use a micro-controller, such as a PIC, this will give a one chip solution, with crystal controlled accuracy, and accurate adjustments for the flash rate. You could also easily add an LCD text module to display the flash rate.

What's it for anyway?.

 

[mindctrl]

Proof of theory concept to make a cheap, durable and rugged duress beacon for outdoors use, such as backpacking, cycling or winter sports. This is just the beginning, and I want to learn how to use the infrared LEDs because they are oh so neat

I also want to test out an idea of mine that the videoconferencing unit at work can see infrared, and translates it into pure white light.

But right now I just want to build the thing

 

[Nigel Goodwin]

Proof of theory concept to make a cheap, durable and rugged duress beacon for outdoors use, such as backpacking, cycling or winter sports. This is just the beginning, and I want to learn how to use the infrared LEDs because they are oh so neat

So why IR?, why not visible light?.

I also want to test out an idea of mine that the videoconferencing unit at work can see infrared, and translates it into pure white light.

Video cameras are sensitive to IR, simply point a remote control at it!.

But right now I just want to build the thing

Check my IR remote tutorial for suitabe hardware, basically one transistor, a couple of resistors, the LED's, and a PIC.

I would suspect you're better off modulating the IR LED's, as they are in remotes - and pulsing them with high current. It lowers current consumption, while maintaining a high output.

 

[mindctrl]

would modulating the LEDs give me a constant output? Or would it group them and make them all funky... for better lack of a word. I guess I'm not understanding what you mean by modulating? Also, where can I find your tutorial? Thanks Nigel, you're the best.

 

[Nigel Goodwin]

would modulating the LEDs give me a constant output? Or would it group them and make them all funky... for better lack of a word. I guess I'm not understanding what you mean by modulating? Also, where can I find your tutorial? Thanks Nigel, you're the best.

You pulse them far too fast to see them flashing, and my tutorials are listed in my signature!.

 

[mindctrl]

Ahhhh... gotcha. Looking at the tutorials right now and this is exactly what I need. Thanks, Nigel. If I want to go more then 8 LEDs, how could one do that? Would I just need to daisy chain the chips? If so, what Pins to what Pins should I chain?

Thanks again,

--mindctrl

 

[Nigel Goodwin]

Ahhhh... gotcha. Looking at the tutorials right now and this is exactly what I need. Thanks, Nigel. If I want to go more then 8 LEDs, how could one do that? Would I just need to daisy chain the chips? If so, what Pins to what Pins should I chain?

If you're only wanting the LED's to flash all together you can use a transistor fed from one PIC pin to feed all the LED's. The existing IR tutorial hardware uses two IR LED's in series, with a current limiting resistor - you could put more of these in parallel wioth the existing ones, each pair of LED's with their own resistor, just like the existing one.

You would have to uprate the transistor though, to provide the higher current.

 

[mindctrl]

I think I'll just use the LED flashing tutorial and find out the voltage & current demand of all the LEDs wired in parallel, then get a transistor that can handle it, put a resistor at the beginning of the LED string, and it should work... right? Or do I need to wire each LED with a resistor of their own... For example, T = Transistor, L = LED, R = Resistor:


T---R---L---L---L---L

or...

T---R---L---R---L---R---L

 

[Nigel Goodwin]

I think I'll just use the LED flashing tutorial and find out the voltage & current demand of all the LEDs wired in parallel, then get a transistor that can handle it, put a resistor at the beginning of the LED string, and it should work... right? Or do I need to wire each LED with a resistor of their own... For example, T = Transistor, L = LED, R = Resistor:


T---R---L---L---L---L

or...

T---R---L---R---L---R---L

You only need one resistor per chain of series LED's - how many LED's you can put in series depends on your supply voltage.