EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: khatus on October 17, 2022, 01:49:32 pm
-
Maybe my question will be a little funny. But it will be very helpful if you can give a little information about one thing. In electronics, we calculate power at various points/calculations to predict/understand how much a device will actually heat/ or how hot it is. , My question is. is there any chart or table? So that I can visualize how hot a device is actually getting.
(https://i.ibb.co/QrgLkJM/Screenshot-10.png) (https://ibb.co/4Pqz2Fp)
-
Excel?
-
How hot a device is getting (that is, its temperature) depends not only on its power dissipation but also on its form factor/package and what kind of cooling is being applied to it -- i.e. a heat sink, or moving air or other active cooling mechanisms.
-
Thermal imagers?
-
A lot of datasheets have thermal resistivities/conductivities in them, knowing how much power is being dumped in to a device would let you work with the thermal "resistance" equations to work out how hot the middle will get given a measured temperature on the casing or the pins. You could then code these equations in to something like a python or R or matlab script for your specific setup.
-
Maybe my question will be a little funny. But it will be very helpful if you can give a little information about one thing. In electronics, we calculate power at various points/calculations to predict/understand how much a device will actually heat/ or how hot it is. , My question is. is there any chart or table? So that I can visualize how hot a device is actually getting.
(https://i.ibb.co/QrgLkJM/Screenshot-10.png) (https://ibb.co/4Pqz2Fp)
Hi,
In general this is a kind of complicated thing to calculate. It depends mostly on surface area and power being dissipated, but there is more to it because of the shape of the parts or heat sinks.
As said though, the main points are the surface area vs the power being dissipated. The more surface area you have exposed to the air the more power can be dissipated without raising the temperature of the device too much.
You can also look at data sheets as they often give specs for the different package types. You can then use the power dissipated to figure out the temperature rise.
One rule of thumb is that 1 square inch of surface area dissipates 1 watt with a temperature rise of 60 degrees C. This varies though because as the surface area increases while the area where the power is concentrated stays the same, the conduction of the material become a factor and so the center of the surface area could get much hotter than the area around that.
Testing is almost always used to make sure a device does not get too hot. Sometimes a small hole is drilled into the heat sink with a probe inserted to measure the case temperature. The die temperature can then be calculated by referring to the data sheet where it tells you the case to chip thermal resistance.
The thermal resistance acts like an electrical resistance in that the temperature is analogous to voltage and the power is analogous to current. So if you see a thermal resistance of 10 degrees C per watt junction to case, then if your case temperature is 50 degrees C and your power dissipation is 2 watts, then the junction (die) temperature would be have to be 70 degrees C. That's because 10 degrees C with 2 watts is 20 degrees C, and acts as a temperature drop (similar to how a voltage drop occurs across a resistance), and so to get a 50 degree C case temperature the die must be 20 degrees C hotter than that.
If you have a heatsink that is rated at 5 degrees C per watt then the temperature rise is 5 degrees for every watt being dissipated. With 6 watts that means that the case temperature is 30 degrees above ambient.
If you have a heatsink and the rating for that and you know the rating for the junction to case, you can calculate the expected junction (or die) temperature with a given amount of power being dissipated. The two thermal resistances act in series like two resistors act in series with a current and voltage.