Ok, here is something that I am pondering, and looking for guidance on... I'm wanting to create a flickering effect in my led's. I'll have 6 led's on this, and each of them I want to always maintain about 70% of their brightness, but I want them to randomly "flicker". Not strobe, not off, I want them to just dim a bit (say 60% - 70%), then back to full intensity. Think of the light a fireplace puts off when you are looking at the ambient light on the walls. Also, the ability to turn off the flickering and have all LED's running full bright, or chasing. They'll be 5mm UV LED's, with a max forward voltage of 4.5V, Reverse voltage of 5V (I dont understand this concept yet...), forward current of 30mA. If I am understanding this stuff correctly, I should be able to pull this off with a 5V line from the power supply, and run these in parallel with a 3.3 Ohm/.15W resistors should "theorheticaly" give me good performance @ peak without allowing too much through and damage the LED's. What I am mystified by, is running the 6 leads back to some sort of controller that would "route" the electricity to the proper LED making it flicker. Heck, if there is a controller that would have several "switchable" effects that I could toggle through, that would be cool too! If you guys know of an existing LED controller that is premade - I'm not too proud to buy one! If there is a diagram of one already, I'd be willing to try my hand at putting one together too! I've just got NO idea on electronics/PCB design/circuit design. I'm learning, but no where NEAR what this group has coarsing through it's veins.... I appreciate any and all assistance on this.
For the dimming of the LED, you have two options (or more) 1. Have a varying resistence or voltage. 2. Pulse-Width-Modulate the current being fed into the LED so that the duty cycle determines how bright the LED "appears" to be in human eyes. This site contains a plethora of LED circuits; some of which might apply to your needed application.
Ok this post is not going to be a whole lot of use (sorry) but the problem with flickering is that your brain is quite good at picking up patterns and repetition. To acheive a flicker you need something to be random, randomness is really quite dificult to generate. With a computer program the best that you can hope to acheive is a pseudo random chain of numbers generated from a complex formula. Into this formula you must put a seed, it is where this seed comes from that is difficult. In the UK we have Premium bonds, they use a PC to select random numbers, this takes its seed from a number of very sensitive temperature probes. I think that the new version measures the changes in temperature of its own CPU or so but I am not sure. The Americans listen to space noise to generate their random numbers. So how you generate something truely random is a problem, one way would be to take a very sensitive ADC and hook it up to your 5v line in your computer. If you ignore the first say 5 digits ie 4.9849382734 is the actual value and you just look at 382734 then any small fluctuation and this will change completely. Well this has just been a bit of a ramble but you never know it may make you want to attach a satalite dish to your PC Rod
any half decent pysudo random algo will sufice. enless your a living maths legend you won't be able to notice the pattern
One of the other things I was thinking of (after reading some stuff on metku mods) is coming up with a circuit design that accepts the activity pulses from the LAN adaptor and another one that takes pulses from the HD activity.... I could then have, for example 4 LED's tied to the HD activity, and then throw in 2-3 LED's of another complementary color, and have that tied to the LAN activity! That would make the flames sick, would give them flicker, AND would make the display something functional! What do you think of that? edit: Found this on MetkuMods (LINK). I believe with a little twiddling, 2 of these will work!!!! One for the HD and the other for the LAN!!!!
EXCELLENT LINK!!!! Thank you, zz300! If it does not make it's way into this mod, I have another in the wings that this will be perfect for! Sweet!!!! I got lots of blinky stuff now!!!!!
Needing a bit o' advice here... Ok.... linky to the schemeee (Impressive wooo doggie circuit) Ok, souple of questions have popped up in my witto tiney bwain, 1) How to attach up to 10 LED's to this (not 1 or a paralleled 2). 2) How do I translate this into a circuit board? What I want to have this circuit perform is the safe but simultaneous flickering of approx 10 LED's (maybe less, but I have the proper colors on order, and 10 will be an easy shot if I want that many). As well, since I will have various LED's out there, is it possible to have the base voltage to hit @ 5v, so that I can keep the reisitors to 1/4 watt. The schematic shows resistors on the PCB to regulate the LED's, but there is nothing keeping me from putting these in line with the LED wiring... correct? Using various rated ohms on the different colors, I can dial down the voltage to what each LED needs. Everyone gets nice and bright. I am very new to electronic component level diagrams and such, so please... be gentle! Is this possible to perform with this linked circuit board? or am I over doing it a bit for this unit? TIA!!!!
According to a book that was imply titled "Chaos" nothing random exists. The things that we see as random can be described using complex formulas. But as noted by TheAnimus it doesn't have to be truely random (which is good since nothing ever is truely random).
Just had another "branch" inspiration on this circuit design. How would I modify this so that when one set of LED's are lit, another would be dark? Example: HD Activity light is on-off, therefor, 2 channels of LED's (1 for on-off, 2 for off-on). When there is no activity, I'd like the 2nd of the LED's on. Then when activity hits, they each flash opposite of each other..... Thoughts?
Erm, I'm not sure if this would entirely work, but maybe you could feed the output through a NOT gate, and have a feedback path going through the LED circuit, so when the input to the NOT is high (your HDD circuit, for example), the output is low, so the 'alternate' LED circuit is OFF. EXTRA THOUGHT: Due to the nature of these logic IC's, you will get a propogation delay, it'll be very small and I doubt you'd even notice it, but you would need to be aware of this fact if you were to control the circuit with a microcontroller for example.
Use a PNP switch for one set, NPN for the other; a second NPN chained on to give an inverted signal; or use macroman's Bit-Tech circuit.
cpemma.... you are a GOD among men! Thank you, thank you, thank you... My boss and I have been playing with ideas on this all day long! Now we can get back to work and be productive!!!!
This is one way I was playing with some months back, minimum parts (AFAICS) and OK for light loads only. Both transistors are working as emitter-followers, but the current through R1 is so low when the opto is off (led current/transistor gain) there's basically still (12-0.7) volts for each string of leds. And when the opto is on, R1 is high enough not to overload it. The opto-diode cathode doesn't go to ground, but to the other mobo pin - shown 'as is' just for simulation.
