How to measure heat from electronics 1

Search for "thermal imaging" . .
Here are a few:


Expensive. .


[link ] [/url]

Get an IR meter for individual measurements

You need to be looking for a Calorimeter.

Thermal imaging cameras or IR thermometers measure temperature, not heat.

Yes, MintJulep.

I've been looking for a calorimeter, but haven't been successful finding one that would work. Basically I'm looking for a controlled environment box, maybe 5 ft by 5 ft, that I can set to various temperatures. Then I'd place the components inside and it would be able to measure how much heat (in watts) the device was putting out.

I would even be willing to have one custom made if anyone knows of a company that would do something like that.


I appreciate the replies thus far.

The amount of thermal watts it puts out will be equal to its electrical watts consumed.

heat generated =(power in) less ( power output }

or neglecting signal input power

for efficiency is defined as

power out = eff x (power in), then

heat generated = (power in) x ( 1-eff)

so the thermal wattage is equal to power in only for 0% efficiency.


Most heat test are done in a flow tunnel with controlled air flow unless nat. convection is sought and monitored temperatures.

I'm with Compositepro -- if you can measure the input watts, you have your answer for most electronic devices.
An exception would be if it's a fairly efficient radiator of power in another form (radar maybe? or light?) -- in which case you might need to subtract the amount of power that is known to leave in the form of electromagnetics instead of heat.


Good on y'all,

Goober Dave

While the suggestions offered in this thread are all good ways of measuring temperature, they are not means of measuring heat, if it is in fact heat that you are looking for. While doing a research project for NIST when I was in school we rigged a device to measure the real area of surface contact between two objects, and as a consequence devised a contraption that can measure the heat load from a semiconductor with high accuracy. Conductive heat transfer is q=kA/L*delta T, you can mount your device on a substrate of known conductivity, k, and of known area, A, and known thickness, L, you can then measure the temperature drop from one side of the substrate to the other. (I use thermocouples). Now that you know k, A, L and DT you can back out heat, in watts if you stick to SI unites for all of your parameters. If you want the best results perform the test in a vacuum to eliminate convection, and unless you are operating above the magic smoke range for most electronic devices, radiation shouldn't be an issue.

One other point: yes, while the device cannot dissipate more power then it put into it a la the Second Law of Thermodynamics, neither can any device 100% efficient, a la Law #3. This means there are losses, and much of that is in the form of heat. Another consideration is that if the device is being used to drive other devices then most of the power into it is diverted elsewhere to other components. This requires a circuit wide thermal energy balance.

For example: I am driving four CREE MC-E multichip LEDS at 12W each for a total of 48W. All four of them are driven via a single device, but it only dissipates 150 mW of power. The remainder goes to the LEDS and line losses in the traces on the board. This illustrates that power in is only equal power out when the entire control volume is considered. Each component has it's own contribution to power dissipation and efficiency.

will a temperature chamber tell you have much heat it takes to maintain the set point?

If not you could just turn it off when it is at the temp you want.

A thermal test rig will not tell you directly how much heat it takes to maintain a set point, but you could back a good swag out empirically given the data you got from the tank. What you could do however is heat things up to steady state then tweak your input power until you reach the point you desire. This is without convection however, so if you are using a fan or other active cooling measure you should establish the same operating condition with your active cooling without operating in a vacuum and tweak away. In my line of work I am forbidden to use active cooling and must therefore use only passive conduction and convection. Were I to run the same test I would do it in an enclosure similar to the one I am trying to design into. Most of the data on the component sheets are fairly conservative (so far), but nothing beats a day in the lab to find out what's really going on. In any system there are variables that were not present during the manufacturer's lab testing, so take what they give you for what it is worth and invest in testing facilities.

The reason that you can't just turn it off when you get to where you want to be is because unless you have reached steady state then you would be shutting off in the middle of your transient. Unless that's how you intend your devices to perform, it's not going to give you the real set point but only a data set of one, and you can't get any meaningful data from a single point.

You can build your own calorimeter.

> Insualted vat of Fluorinert, or similar
> Measure starting temp
> Dunk board in and power
> Measure ending temp
> Multiply specific heat by deltaT and divide by test duration

TTFN

FAQ731-376