# Electronics box in the heat situation

i'm designing a little project but i'm running into some environmental heat issues that i'm trying to work out. i just need a little clarification and insight from anyone willing.

I have this box made of sheet steel that is going to be mounted to the side of a building. the box will theoretically be air tight so that none of the environmental elements will harm the components inside. the components inside include electronics that will do up to 70W of power and are rated to work in an environmental condition (inside the box) up to 60 C. The thing i'm trying to figure out is what temperature it would have to be outside of the box (i'm first trying to do this without considering direct sunlight because that's really difficult to do) where it would put the box temperature over 60 C and destroy the components.

this is what i'm thinking. after some reasonable amount of time of letting the electronics run, some sort of equilibrium will occur between the temperatures inside and outside the box. at that point the rate of heat transfer due to conduction through the sheet metal will be equal to the rate of heat production by the electronics (70W).

so using fourier's law (the rate of heat conduction):

dQ/dt = k*A*(T_in - T_out) / x

where:
dQ/dt = rate of conduction
k = steel's conductivity constant
A = surface area exposed to environment
T_in = temp inside box
T_out = temp outside box
x = thickness of sheet metal

we can conclude that the temperature inside the box will only be slightly hotter than the temp outside.

So if what i did was true then the temperature would have to reach a really high 58 C or so to compromise the electronics functioning capacity?

berkeman
Mentor
You are going to need electronics rated higher than 60C to sit in a sealed metal box outside. Standard Industrial temperature range is -40C to +85C, and even +85C rated components may be problematic in a metal box outside in the sun. Why are you trying to use Commercial-grade components in an application that is at least Industrial-grade, and maybe even Automotive-grade?

the components are always a matter of price and availability to what is required for the project.

but can you comment on my analysis taking into account that the box will always be in the shade?

berkeman
Mentor
If it will always be in the shade, that will help some. Thermal conduction and heat rise for electronics is generally characterized in terms of a thermal conductivity theta, in units of degrees C per watt. A good heat sink will have a fairly low Theta (a few degrees C per watt or less), and straight convection cooling with stagnant air (like in your box) will have a much higher theta (many degrees C per watt).

Are you directly heat sinking this 70W to the metal box, or is only through the air inside? It's probably easiest to measure the theta numbers, as opposed to calculating them. They depend on how components are mounted to metal structures and such. BTW, your power components should have theta(J,A) (junction to ambient) numbers in their datasheets, for their various package and mounting options. All of the thetas add up to give you the overall temperature rise. You can look at a heat sink manufacturer's website for more info on the theta numbers and calculations.

i understand what your saying but i'm not sure if we're on the right page or if i'm just completely wrong. the components that i'm using are rated at 60C ambient temperature, meaning the temp inside the box. if they give a rating like that i'm assuming that they take into account the worst scenario of cooling (stagnant like you said) so that the components will function in an environment of 60C, not an internal operating temperature of 60C.

am i making any sense?

berkeman
Mentor
i understand what your saying but i'm not sure if we're on the right page or if i'm just completely wrong. the components that i'm using are rated at 60C ambient temperature, meaning the temp inside the box. if they give a rating like that i'm assuming that they take into account the worst scenario of cooling (stagnant like you said) so that the components will function in an environment of 60C, not an internal operating temperature of 60C.

am i making any sense?

I'm speaking from a designer's standpoint -- I design circuits and modules with components (transistors, resistors, ICs, etc.), and those components typically have either Industrial temperature range parts (-40C to +85C) or Commercial temperature range parts (0C to +70C). That temperature rating is for the body temperature of most parts, for example. There will generally be a different max junction temperature for semiconductor parts, higher than that max body temperature, and it's my job as the circuit/product designer to be sure that I'm using adequate cooling to ensure that the max junction temperature is not exceeded when the ambient environment is at max (70C or 85C).

It sounds like maybe you are dealing with an already-designed product that you want to put into an additional metal box for operation outside? 60C is a strange temperature rating as I mentioned, since it does not match standard component temperature operating ranges. If it is a product already, it assumes some ambient conditions associated with that 60C rating. Does it assume no heat sinking and only stagnant air up to 60C? If so, then you can just put 70W of light bulbs in the box and test the temperature of the air in the box after it stabilizes to see if it gets over 60C.

So when I design a circuit module or product, it is my job to use the theta(J,A) numbers for the semiconductor parts, using whatever cooling method I have (dead air, forced air, heat sinking, or whatever), and calculate

yes i am using prefabricated parts that i'll assemble to make my system. i'm working off the spec sheets of each part which says max ambient of 60C.

unfortunately i can't test the temps in the box because, 1) i don't have the box yet, and 2) i'm not in the environment that i'm going to be assembling it. that's why i'm going through all this trouble working out the math.

berkeman
Mentor
I think I'll move this thread to the ME/Aero forum for a bit to get some different views. There should be some folks there who do thermal conductivity work for metal enclosures in outside environments. If that doesn't work, I may move it again to EE.

I design equipment for the military that is fairly similar to what you're describing (electronic components in metal enclosures).

My immediate suggestions would be to A) vent the bottom of the box to allow cooler air in while still disallowing moisture (rain or snow) an entrance way and B) possibly provide a fan and/or heat sink over the known warmest components to help alleviate internal temperature rise. Also, if possible, consider making your enclosure out of aluminum, not steel. The coefficient of conduction is much higher (~180 W/m-K, depending on your source) than steel (12-27 W/m-K, depending on your source and material).

Your analysis in the first post was simply relying on the ambient air and the thermal conduction of the material to produce an inside temperature for the box. You are not taking into account heat produced from the components that becomes 'trapped' in the box.