Oh my god. Holy sh*t dewd! Congratulations - That is one hefty piece of work! You seriously got to give us more than an animated gif man, how about a documentary?!?
That is a wounder to behold indeed! You've inspired me, i think i'm going to have to make a 1-d version, mabye using BI colour LEDs
Sure, a report will follow, but first I want to make it fully ok. But I didn't want to keep this pic from you
That one looks very nice. I've been thinking if trying something similar but maybe with FPGA and driving the LEDs with PWM to get different brightness levels to have some additional challenge But there are still so many other things waiting to be done so that it can take a while before I will start with that. Or how about a PWM cube with bi color leds like TheAnimus said. That could be interesting. But building the cube from bi color leds might be quite expensive.
i think an amazing site to behold would be tri color LEDs with varying power to each color, so that you could basically have any color showing from any led.. in my simple mind you would need to triple up on the circuitry, just have 2 more controllers for the two other color legs on the led and then some way to control the power to that LED's color.. man that would be awesome. and... really expensive.
Do you mean tri-colour or RGB? Remember tri-colour LEDs only have two dies with a common cathode. RGB has already been done: http://nw.com/nw/projects/cubatron/
Instead of getting yourself all confused, think of PNPs as negative NPNs. All rules for PNP are the same as NPN if you put a - in front of the NPN numbers.
Yes - I'm well aware of that... Looked more carefully at the schematic and to me it looks like the LEDs should be seeing 4.55 - 2*Vcesat, which I believe typically is about 0.2V, so the LEDs would be seeing about 4.15V each, which methinks is bad... Can somebody please explain why this would work? edit: and I'm thinking that the B transistors would be active high, and the A transistors would be active low, is that right? (much delayed response due to classes taking up all my spare time)
Running a single element through a sim package, it does look like things will blow up, but not because the transistors will saturate. Because of the fixed forward voltage of the diode, and the properties of the transistors, the remaining Vcc that is not used by the LED is split between the two transistors. The circuit may oscillate a few times n getting to that point, but eventually they settle to a balance based on the charicteristics of each transistors. Using virtual transistors with idential charicteristics, the voltage split was evern. Using a pair of 'real' complementary transistors, the split was based on their gains and output charicteristics. The problem is that with the base resistances spcified, the collector currents gets quite big, and the whole thing would fry, transistors, and LED. The base resistors would need to be tweaked to make this work. Personally, I'd get ICs that I know could source and sink the needed currents and drive the LEDs directly. If I needed to fix the above circuit, I'd swap that PNP for a NPN and tie the emitter to ground. Then throw in a current limiting resistor so things can saturate and the only real change would be a swap in the signal you send to the second IC. For what is there, IC1 needs to put out a logic high on the desired signal line, and IC2 a logic low. Both would need to be high for the proposed change.
with 10 maxim controlers, you could make a 10x10x10 grid with 1 Pic and no transitors. thats how i would do it... no lateches, ect... simple = better awsome project though
*thread revival!* i just found this: www.hypnocube.com They make kits for RGB and tri color versions. Pretty neat, byt also very expensive. edit: the kits aren't too bad priced.
Awesome project... I was wondering that is there any schematic for 3x3x3 led cube, 'cause i'd like to build one . Bigger would be cooler but i don't have enough time or money...
Alvin and pwled haven't been online since 2005, so I doubt you'll get a response from them. The schematic won't be the problem anyway, it'll be the program. You could set up the circuit any way you wanted and deal with it in software, so if you can program, you're sorted.
Hi there... Firstly, it's been a long time since this thread was around, back before the faceleft, I believe. Talk about necromancy... Secondly, I have what I believe is a fairly basic question: How do you wire up a matrix? I understand the cathode end; it's an NPN transistor, with the cathode on the collector, the emitter to ground, and the active high input into the base. But I seem to remember seeing somewhere a transistor on the anode end of each line as well. Any idea how it would have been connected?
Somebody revived this thread, and it caught my attention, as I used to be logged in as Alvin, until that account stopped working. Anyway I saw that my clean-up of the windred.dk site had removed the pictures from the "Alvin posts". That should now be fixed. Maybe those illustrations can give "m K o" some ideas. Best of luck.
