Current and peak inverse voltage are the things that will get published, and most rectifier diodes will have ratings so high that you needn't worry about either one. The idea is that you take a DC device, and apply an AC signal to it. If you turn on the lights in your room for one minute out of every ten, you have just acheived a 10% duty cycle. If you could do it one millisecond out of every ten, that's PWM. The light will appear to be on at 10% brightness, and it will dissipate 10% of the DC power rating. The modulation part is that you can vary the duty cycle to go from 10% to some other percent. You vary the time the load is on. How it works is that you need to convince some transistor that can handle your load current to switch at an appropriate rate. Doing so involves desingning a circuit that can vary its duty cycle. There are circuits made expressly for this purpose (MIC502 is one). Well, yeah. But it starts to get messy. Especially so since percieved brightness doesn't necessarily map well onto either one. But if you feel a need: http://www.electro-optical.com/whitepapers/candela.htm The answer is yes. As a concrete example, to drive a red Luxeon Star/O (2.95V @ 350 mA) from 5 V, that wants a 5.9 ohm resistor, and the power dissipated will be 720 mW or so. A 5.9 ohm 1/4 W resistor wouldn't cut it, but I used three 22 ohms and one 33 ohm 1/4 W resistors (since that's what I had in the drawer). You can calculate the power dissipated in each one as V2/R.
Linear, damn fine link there! I don't think I have ever learned or gained so much knowledge and unerstanding of a complicated subject from just six pages. Holy hell, that site sent my brain into 110% duty cycle and I think I have it figured out (or atleast a greater understanding of what I was looking at). I am just thankful that I am a quick learner when I am interested. You weren't kidding in the least that it would get a bit messy to do a conversion. Thanks again! About the parallel resistor stuff, could you give me an explanation as to why you have to go from 5.9 ohms to 22 ohms when running them in parallel? My mind would say to use four 5.9 ohm resistors. By your example above, I now know that would be wrong, but I ask so that I can understand and not just follow along. Besides, if I don't ask now, i'd be back asking about something similar appearing to have learned nothing at all. I need to pick up some books on electronics and learn, then I may be able to reverse my role in these forums. Lates, Ryan B. EDIT: Linear, just in case you didn't see it in one of my other posts; kudos to getting Rig of the Month in MaximumPC! I have always admired the attention to detail of your case. From your teaser in the article, I can't wait to see your next case!
