Bizarrely, I got quite a bit of confidence using some of these dc-dc switchers after using one inside a 3W RF out 10GHz microwave transverter.
Odd, I'd be quite surprised if a GHz system noticed a switching converter: there are rarely harmonics beyond 100 or 200MHz (or perhaps more for monolithic converters). If it's a poorly designed system of course (i.e., poor band limiting in any stages), it could be susceptible to lots of sources.
Some exceptions include poorly made chips. There's a National chip somewhere that runs its synchronous buck output stage such that the GND side body diode is driven as a step-recovery diode, resulting in a quite reasonable sub-ns blip.
What is the best way to power an arduino chip (ATtiny, ATmega etc.) from a 12 volt (regulated) source? Is this the Classic LM7805 voltage regulator or is there a more efficent way to do so? (less loss of power thru heat dissipation)
What's the source, how much does power consumption matter? I take it, it's not battery operated? How much do you need at 5V? Just a few mA for the MCU alone, or are there other things attached (say, an LCD with backlight that gobbles a few hundred mA)?
Plug-in hardware, with stuff like the latter, I would be fine with a 7805. An LCD backlight is kind of a pain, but a little heatsink on the 7805 and that's plenty. If this is battery powered, you'll probably want to maximize life, meaning using low-power modes in the MCU whenever possible (cuts current consumption from ~10mA flat out, down to perhaps uA if it doesn't need to do much), and a micro/nanopower SMPS chip to furnish the 5V (roughly doubling battery life versus a low-quiescent LDO).
Tim