Well..expand the string to 100 LED's and the voltage requirement becomes an= issue with the client. They will deal with potential issues of non-consis= tant brightness across the design, rather have 12V than 200V sitting there. David VanHorn wrote: On 9/12/06, alan smith wrote: > > > Dimming LED's *can* be done by just PWMing the device. How about if I > have 3 or 4 LED's in a string.....ie...constant voltage setup vs constant > current....seems reasonable that I can PWM the base of a transistor > (assuming it can sink 20mA) and it should do the same thing....dim the LE= D's > that are in series. And of course, since the PWM dims them by rapidly > turning them on and off...this should also conserve power as well.... You want constant current, rather than constant voltage. current =3D brightness. So, if that is true....I'd need some sort of SEPIC convertor since its > running on batteries, to keep the voltage constant, and would be even bet= ter > if the SEPIC could be PWMed rather than the string, but how fast can a > typical SEPIC actually turn on and off? Better to just stick with the > transistor approach maybe...? Well, either way, you're going to throw the converter out of regulation, unless you use the transistor to short out the LEDs, and have the CCS still dumping X mA through the system. -- = http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist = --------------------------------- Yahoo! Messenger with Voice. Make PC-to-Phone Calls to the US (and 30+ coun= tries) for 2=A2/min or less. -- = http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist