Electronics Cooling Article Library

Between December 2000 and August 2010 these articles, packed with advice on cooling electronics, appeared in the monthly on-line magazine of CoolingZone.com.  In response to popular demand, they have been re-formatted and collected into this convenient one-stop library.  It begins with the opening article from December 2000, on why heat sinks never seem to work as well as you think they should, and will eventually contain every wise and wise-cracking article in the series.  Additional articles will be added approximately every week, so if you haven’t found your favorite Cooling Q&A yet, keep coming back.

December 2000 — Heat sinks are the most popular method of reducing component temperature.  Here are four reasons why they don’t work.

January 2001 — Does it take any special training to enter the field of cooling electronics?  Can the circuit board which is already there act as a heat sink?

February 2001 — Does heat have “inertia”?  Why does component temperature seem to overshoot when the electrical power and fan are turned off at the same time?

March 2001 — Converting watts to BTUs, and is it possible (and ethical) for an IR camera to see through clothes?

April 2001 — Vias in printed circuit boards conduct heat, but how much?  What books besides his own does Tony recommend?

May 2001 — As chips shrink and logic voltage goes down, isn’t cooling digital electronics actually getting easier with time?  And did Electronics Cooling save Star Trek?

June 2001 — Fan noise is getting pretty annoying.  Aren’t there any laws that can help reduce the noise?  Of course.  They’re called the Fan Laws.

July 2001 — Sometimes Something I Know Is Wrong.  The June 2001 article gave an incorrect Fan Law.  Here is another stab at it.  Plus, Do-It-Yourself Thermocouples.

August 2001 — An apology to PEs.  And a tribute to the garage-and-basement pioneers of electronics cooling — The Overclockers!

September 2001 — How accurately can a CFD tool like ANSYS or FloTherm predict component temperatures?  To get a handle on that you have to realize that CFD tools can’t predict temperature at all.

October 2001 — How much do thermal engineers add to the company’s bottom line?  Nobody buys a laptop because it has a great new heat pipe, right?

November 2001 — It’s not the heat, it’s the humidity (at least at the family reunion picnic in July in Michigan.)  Do changes in humidity have any effect on electronic component temperatures?

December 2001 — There’s a thin piece of tape between your temperature probe and a hot component.  Could that tape make the temperature reading wrong by as much as 100°C?

January 2002 — Finally!  The absolute ultimate answers to the three most nagging questions in cooling electronics:

  • Fan on the inlet or exhaust side of the box?
  • What is the best color for a heat sink?
  • Should you use Typical or Maximum Power when modeling a circuit board?

February 2002 — How come when you attach a heat sink incorrectly, it makes the component hotter, but when you attach a thermocouple incorrectly, it makes the component cooler?

March 2002 — Which CFD software is the best?  And why Type J thermocouples are always the most appropriate (for me, at least.)

April 2002 — How come you can’t just look up the formula for ? J-A in a heat transfer textbook?  And finally the answer to the long-standing dispute:  which temperature scale is better — Celsius or Fahrenheit?

May 2002 — If you know power dissipation, you can calculate temperature.  If you know a component’s temperature, does that mean you can find its power?  You’ve heard of “thermal resistance”, and maybe “thermal impedance.”  What about “thermal intransigence?”

June 2002 — Legend has it that Ludwig Prandtl was fond of saying, “The longest thermal analysis begins with a single hand calculation.”  How will you begin that next electronics cooling assignment if you can’t start with hand calculations?  Here are some tricks to get you started.

July 2002 — Fans are coolers, but they can also be heaters.  Heat from the electric motor raises the temperature of the air stream.  How much of the fan’s electric power turns into heat?

August 2002 — How to become a thermal guru in six short weeks (maybe less), the Wizard of Oz way.

September 2002 — Are suppliers of Thermal Interface Materials exaggerating their thermal performance?

October 2002 — You can use a chimney to cool your electronics, even if you are no longer using coal to power your boiler.

November 2002 — Theta j-c — the thermal resistance between junction and case.  You don’t need to know exactly what its value is to know it is not a safe way of calculating junction temperature.

December 2002 — Theta j-c is no good, so how do we figure out junction temperature?

January 2003 — Everybody has heard of the speed of light.  But what about the speed of heat?  You can actually figure it out for yourself without having to pretend to understand the Theory of Relativity.

February 2003 — Thermally-conductive epoxy:  better than non-thermal epoxy, but still a pretty good insulator.

March 2003 — Heat sinks always seem to be anodized.  Doesn’t that add an insulating coating that defeats the purpose of the heat sink?.

April 2003 — How much can thermal grease reduce the temperature of a microprocessor (without the heat sink?)  Can a laptop computer produce burns on the user’s “lap”, or is this another urban legend?.

May 2003 — My components are cooled only by conduction into the board.  I need Theta J-B to do thermal analysis, but the component datasheets never have it.  Can I safely substitute Theta J-C for Theta J-B?

June 2003 — My product is exposed to the sun.  The sun makes it run hotter.  Can I test it by just running the thermal chamber a little hotter to simulate the sun?  How many degrees hotter?  Also, what is the spreading angle for heat conducting from a small component into a big heat sink?  My text book says 45 degrees.  Doesn’t it depend on the material?

July 2003 — How to explain the difference between heat and temperature to your mother-in-law.  And, are the laws of physics stacked against the electronics cooling engineer?

August 2003 — Is there a Thermal Interface Material (TMI) that changes thermal conductivity with temperature?  With luck, the conductivity would increase with temperature.)  And, is it possible to define the “still air environment” so that a method of testing electronics in it could be invented?

September 2003 — Is there a textbook equation to predict the average convection coefficient for a pin fin heat sink?  Does a push-pull fan alternate between pushing and pulling?

October 2003 — It was a joke, but now it’s real:  Purdue University really did develop an air mover based on the traditional Japanese hand fan.  Plus, did updates in the JEDEC standards make Theta J-C any safer to use?

November 2003 — Which cell phone is better at giving off heat to the surrounding air:  the one with a uniform skin temperature or one with a local hot spot? (Assuming both have the same average skin temperature.)

December 2003 — Air in motion is a coolant. A trapped layer of air an insulator. Can we predict whether a particular pocket of air will be one or the other?

January 2004 — What is the temperature of the sky? You might want to know, if your electronics are used outside, and can radiate heat to the sky. And can thermal problems be solved remotely? (Such as over the phone?) The answer might surprise you, considering this column is giving advice over the Internet.

February 2004 — A more complicated method of determining component power by measuring its temperature, except it is not quite complicated enough. And, do fans make air cooler? It sure feels cooler when I stick my hand in front of the fan!

March 2004 — It’s not the heat, it’s the Relative Humidity. Most of the time cooling electronics is not affected by humidity. But if you are cooling below ambient, watch out for the Dew Point Wall.

April 2004 — Burn-In Testing: does it weed out infant mortality in electronics, or does it just burn up time in the manufacturing cycle?

May 2004 — Now that gigabytes and disk space are practically free, why do we need to be careful about how we mesh our CFD models? Shouldn’t we just mesh the heck out of them to get the most accurate results every time? Why take a chance of losing accuracy just to save a few grid cells?

June 2004 — A disk drive has a maximum operating temperature of 35°C in “still air”? If you blow air across it with a fan, can it operate at 40°C? There’s a difference between being the feasibility and advisability of such an action.

July 2004 — If you slow down a fan because it makes too much noise, the circuits get hotter. Is there a formula to convert RPM to degrees C?

August 2004 — The wiring inside an electronics box often reminds me of a bowl of spaghetti. Wiring and pasta have more than just a passing resemblance, though, when it comes to heating.

One single comment

  1. Keith Dixler says:

    Hi Tony,
    I have both of your Electronic cooling books and have read most of your column’s in cooling design. I’ve been an Electronics Cooling Engineer my whole career as well (going on 25 years). I’m an an Engineer (just a BS or full of BS). I’ve often thought about your thermal “overshoot” article and wanted to relay something interesting. I agree with your explanation of the overshoot in terms of the die never increasing in temp after power is removed (but the case temp can increase). However, I’m also a cook and many chefs also claimed that the internal temperature of a roast or turkey will increase after it is removed from the oven. I proved to myself that this truly does occur by watching the thermometer (which measured the internal temp of the meat) rise after a turkey was removed from the oven. My only explanation is that the turkey external surface is where the heat source hits first and the internal temp is cooler then the external. So the heat flow is in the opposite direction with a hunk of meat in an oven (compared to a die on a heatsink. However the ambient immediately drops when the turkey is pulled from the oven so you would think the heat transfer direction would change (but maybe not right away). I though maybe there was some bio / chemical explanation involved in why the internal temp continues to rise for a while before falling. Maybe the meat juice movement has something to do with it? It good to see you kept all your columns. Thanks for the entertainment and education all these years! Best regards, Keith

Leave a Reply