Heat Buildup in Complete Insulation Box

[jbriggs444]

So indoor it could reach 45 Celsius. The operating temperature of the power supply is only -40 C to +40 C.

So even before it is switched on, the power supply is outside its operating specifications. Turning it on will not reduce its temperature.

 

[kiki_danc]

So even before it is switched on, the power supply is outside its operating specifications. Turning it on will not reduce its temperature.

The spec says: "A standard transformer with 220°C insulation and a 150°C temperature rise, will be rated to run full load in an average 30°C ambient environment over 24 hours with a maximum 40°C ambient temperature.".

So if the load is only one half.. and it is not as hot as full load. Let's say the surface temperature of the transformer is 50 Celsius. If the ambient temperature is 42 Celsius. Can the 50 Celsius heak sink into the 42 Celsius ambient temperature?

Or maybe the rule is.. as long as the temperature of the ambient air is less than the transformer.. it will heat sink into the colder environment.. isn't it.

 

[russ_watters]

I was referring to the air in the enclosure. Wont the temperature keep increasing if the heat can't get out (assuming the box is a perfect insulator or thermo)

I'll put a finer point on the situation created by your constraint: Because you have specified a 50C surface temperature, the only easy way to make this scenario work is to use a controller that reduces the power provided to the lamp to maintain 50C as the box warms up. When the box reaches 50C, the controller shuts off the power to the lamp completely.

When I said the lamp was always at 50 C. I was referring to a lamp in open air at constant 50 C. But taken to a sealed enclosure, I realized it was automatically no longer a constant 50 C.. but would increase.

If that's what you wanted, you should have specified it. We're not mind readers here, but we will give you exact answers to specific questions.

 

[russ_watters]

When I didn't make my question clear and Chestermiller replied it would be constant. I actually imagined the enclosure would stop the temperature from increasing. I imagined it was like the enclosure was filled with water (which stands for temperature), and the source of water at center couldn't push it back anymore because the enclosure enclosing it was stronger. So I was thinking maybe the 50 Celsius in the power supply will just stay that way throughout the enclosure as it equilibrate with air inside. When you guys emphasized it would really increase to 1000 Celsius as I imagined. Then I made full stop and would no longer put the power supply inside a sealed enclosure. I'm sharing so you would know how a physics illiterate or newbie would think.

Well going back to the water analogy. Is there a temperature.. maybe one billion degree Celsius.. that it could act like the water analogy where it couldn't increase anymore because the stronger enclosure wall was pushing it? Or would the temperature need to fill up the Planck scale before it could do that? And since the universe can fit inside the Planck scale.. then it's not possible for temperature inside our universe to act like the water analogy?

Newbie or not, you should put more effort into asking coherent questions. This one is pretty much all gibberish.

 

[russ_watters]

The spec says: "A standard transformer with 220°C insulation and a 150°C temperature rise, will be rated to run full load in an average 30°C ambient environment over 24 hours with a maximum 40°C ambient temperature.".

So if the load is only one half.. and it is not as hot as full load. Let's say the surface temperature of the transformer is 50 Celsius. If the ambient temperature is 42 Celsius. Can the 50 Celsius heak sink into the 42 Celsius ambient temperature?

Or maybe the rule is.. as long as the temperature of the ambient air is less than the transformer.. it will heat sink into the colder environment.. isn't it.

An object can only dissipate thermal energy if it is warmer than its environment. The transformer will start at the ambient temperature and then when you energize it it will warm up and dissipate heat.

 

[CWatters]

The spec says: "A standard transformer with 220°C insulation and a 150°C temperature rise, will be rated to run full load in an average 30°C ambient environment over 24 hours with a maximum 40°C ambient temperature.".

So if the load is only one half.. and it is not as hot as full load. Let's say the surface temperature of the transformer is 50 Celsius. If the ambient temperature is 42 Celsius. Can the 50 Celsius heak sink into the 42 Celsius ambient temperature?
Or maybe the rule is.. as long as the temperature of the ambient air is less than the transformer.. it will heat sink into the colder environment.. isn't it.

Yes. Heat always flows from hot to cold never the other way.

Now it's clear you are really talking about a power supplies...

PC Power supplies are designed so that when operated in the rated environment (40C) the temperature of their internal parts does not exceed a higher temperature. Typically the die of a semiconductor must not exceed 125C and to meet that the case of the semiconductor must not exceed 60-70C. That is usually measured where the case of the semiconductor bolts to the heatsink. The size of the heatsink, method of mounting, and airflow through the power supply are designed to keep everything to these sort of temperature limits. They are tested in special environmental chambers where temperature and humidity can be controlled. I've actually done this.

If you want to use a power supply in an environment that is outside it's specification (eg 40-50C) you need to de-rate as you suggest. Eg run it at less than full load to keep the temperature of the internal parts within specification. Unfortunately PC PSU manufacturers rarely provide the information needed to work out how much to de-rate their products to achieve this. I don't recommend putting temperature probes into the power supply yourself because of the dangerous voltages inside.

One thing to know is the effect of heat on electrolytic capacitors. As I recall every 10C increase in temperature reduces their expected life by a factor of 2. So when choosing a power supply look at reviews online to see if it uses high temperature capacitors. Typically these are rated to 105C I think.

Edit: to clarify. It's possible that a power supply operating in an office at 20C will live four times as long as a power supply operating in an office at 40C (all the time). But I can't tell you if that's 20 years instead of 5 years or 4 years instead of 1 year. That depends how good the designer was. Some design engineers have got this horribly wrong in the past and large numbers of their products have failed within the 1 year warranty!