Exactly what is your design power problem?
Bill Schweber - February 14, 2013
Nearly every design engineer says he or she has a "power problem." I get that - but sometimes you have to probe further to find out what the nature of the power problem is.
I divide the power problem into two broad categories: those who don’t have much power available - energy harvesting is the extreme example, of course - and those who can get a little more power from their source if they need to, but whose major problem is thermal and keeping their system from overheating.
Although both groups have power problems, their approaches to dealing with are quite different, of course. The source-constrained group looks at ultra-low-power parts, sleep and wake-up modes, and all sorts of clever power-management techniques. The thermally challenged folks look to lower-power parts and operation, of course, but also dissipation techniques including conduction, convection, and radiation, including heat sinks, cold plates, fans, and sometimes even more-aggressive approaches.
Many times, the thermal group also has the problem of assessing local hot spots versus overall temperature. It's one thing to say that your box is staying cool enough with a temperature below the maximum allowed. However, it's another thing to design and verify that none of the individual components - especially ICs, MOSFETs, resistors, and batteries - aren't exceeding the maximum.
The problem is complicated when your layout uses the PC board's copper layer as a cooling medium. If you read some of the data sheets and app notes, some assume you have perhaps four or six square inches of this copper for heat-sinking that IC. Often, that's not going to be the case with designs where adjacent components are millimeters away, each expecting to have some of that cooling PCB copper to call its own. It's not like the two (or more) components can talk amongst themselves and come to some agreement on how to share that precious resource, is it?
But whatever your power problems, if you think you have extreme power challenges, check out this recent article from IEEE Spectrum, "Finding the Source of the Pioneer Anomaly." It is primarily about the long-term and very, very long-range investigation of the mystery of why the Pioneer spacecraft pair (launched in 1977 and still functioning!) are very, very slightly off from their anticipated course as they travel through our solar system. Much of the article centers on their thermal and power design issues, and how they affected the journey. [BTW, if you want a fascinating look at their 36-year dual missions to the edge of our solar system and beyond (17.9 billion kilometers (119.9 AU) from Earth as of March 2012), and how the vehicles lasted far beyond their intended life, check out "Voyager: seeking newer worlds in the third great age of discovery", by Stephen J. Pyne; I highly recommend it.]
Which class of "power problem" do you have? Which "side" do your designs inhabit? Do you find it difficult to relate to the issues that the other side has?