So far, this is only an idea, and I need help from the rest of you to make sure that this is a sane idea. This is the schematic I will be referring to. The intended function for this is as follows. When the HD signal is active, triggers the 555 timer, setup in monostable mode, bringing its output high. This output is fed into the Vcc off a LM358 dual op-amp. The op-amps are setup to act as an buffer (IC1A) and inverter (IC1B). The HD signal is also fed into the input of the buffer, while the buffer outputs to a 2n3904 and the input of the inverter op-amp. The inverter op-amp drives a 2n3906. So long as the HD signal doesn't not go into high impedence mode before the 555 timer's output goes low (a time-out of sorts, the HD signal can't be inactive for X number of second before the op-amps, and hence the LEDs, turn off), then the two op-amps follow the HD signal. The transistors feed into two blocks of LEDs (pin-headers). One set is active while the HD signal is active (that'd be the block feeding off the 2n3904) while the block fed by the 2n3906 is on while the HD signal is inactive. Now, I just have a few concerns. Primarily it's with the transistors, since I'll admit it, I know just about jack about them. I think this time I managed to get the right orientation between supply and load, but I'm not too sure about the resistors off of them (they *should* give me 20mA, but the fact that I can't seem to get a concrete hFE figure means I could be way off). Also, I *think* that the 555 can supply the op-amp with enough current, I'm not sure if I read the op-amps power specs right. I think that covers it. Feel free to shoot down my ideas now. Oh yeah, and I'm using op-amps as logic gates because it's a bloody waste to get 2 DIP-14 packags to use only 1 gate off of each.
I thought I said that, but then again, there's a lot of stuff I think I did that didn't do. When the HD signal is oscillating, I want one set of LEDs to light when the signal is active, while another set of LEDs lights when the signal is inactive. When the signal is inactive for a certain period of time, power is cut off to the LEDs (I suppose I could find a way to cut off power to the transistors rather than to the op-amps, probably makes a bit more sense actually).
I think I got the idea.... When the first pulse from the HD indicator is received the LED's power up, and start flashing. They stay powered as long at there are still pulses from the HD. (Still flashing in sync with the HD activity) But if the HD is inactive for a minute or two, the LEDs turn off totaly. (Saving your eyes at night?) First off you don't need an inverter. By using a combo on NPN and PNP you have already inverted the effect. I don't think you need the op-amps at all. I really just hate messing with op-amps. (If you have no problem, use them) I think that you could use an Optocoupler as a switched buffer. You would need only 20mA max for the opto and nearly zero for the 555 timer to activate. The 555 timer then can provide more then enough current to power the Opto's LED directly. (Providing the timer function) The opto then activates the transistors. But only as long as the Opto is powered by the 555. I may be totaly wrong, but I think at least it's less components.
The reason I was using op-amps to act like gates was the very original idea I had for doing something different w/ HD activity involved using flip-flops, and from there logic gates, and so on and so forth until I got that. Your idea is quite attractive, as the part count is lower (9 vs. 13 by my count) and much more compact. Most importantly though, is that I understand how it all works. I'll flesh it out a bit later and update.
Ok, I worked this up, and I'm fairly confident in this one. The resistor values off of the transistors are from an assumed 5v supply (I know it won't be exactly 5v), 4 3.3v @ 20mA LEDs in parallel. Yes, I know I should have one resistor per LED but...do I really need to? The 555 timing is set at almost 1/4s. The reason it's not a really long time is that if it gets too long, then random access by the OS will keep the timer from going low. I'd like to have the LEDs active as long as the HD remains fairly active, it doesn't make much sense (at least in my head) to have the LEDs tell me the HD's not doing anything for a long time. I may stick a pot in series with the timing resistor in case I don't like the time-out period.
