24 volt power supply question:

[MichaelGP3]

Electrical contractor installs a 24 volt, 3 or 5 amp (not sure) DC power supply used for HVAC controls (fed by 120 volt AC) in a 15" x 15" x 4" NEMA 1 enclosure in an electrical room, no battery charger present. AC in on one conduit, DC out on another. Enclosure is not sourced from the power supply manufacturer; power supply occupies @ 20-25% of the enclosure's volume. Does this enclosure need to be vented? I asked the EC and he doesn't know. I don't either.

 

[Cow]

I'll ask the obvious question, why don't you read the instructions? If you don't have instructions, usually you can google them.

I decided to do the legwork for you, here's instructions for an AB PS:

http://literature.rockwellautomation.com/idc/groups/literature/documents/in/1606-in005_-mu-e.pdf

 

[MichaelGP3]

It seems that so long as clearance for air space is provided within the enclosure a vented enclosure is not necessary. Thanks.

 

[gar]

121201-0915 EST

The answer would be very dependent upon the type of power supply.

If the supply has a maximum output power of 5*24 = 120 W and has an efficiency of 95%, then the power dissipated within the enclosure is less than 7 W. In that size box 7 W will not produce an excessive temperature rise. It will be modest.

Now suppose the power supply was an old style linear regulated supply, then you might have 50% efficiency. Then power dissipation in the enclosure would be 120 W, and that would produce an excessive temperature rise. Probably well over a 70 deg F rise. With an ambient of 70 plus a rise of 70 the interior would be 140, or 60 deg C. This would probably be above the maximum working temperature of the power supply.

.

 

[G._S._Ohm]

And the block diagram implies it is a switch-mode supply, as does the spec'ed efficiency.

Since smaller supplies in larger boxes are surrounded by thicker layers of air, I'd think they'd run hotter than supplies that almost fill the box.

 

[gar]

121201-1405 EST

G._S._Ohm:

Where in the original post is there a block diagram, or even much of a specification on the power supply? The point of my post was to try to get MichaelGP3 to think about the fundamental factors in answering his question.

On heat transfer. To a large extent if you put X watts of thermal energy inside of a box with no forced air inside the temperature rise in the box will be a function of the outside surface area of the box. In other words the inside to outside thermal resistance of the box. There are gradients inside and outside and therefore temperature variations inside the box.

Cut the outside surface area of the box in half and for the same interior power dissipation the interior temperature will about double.

You could power a 100 W incandescent bulb in the center of the cabinet and get an interior temperature rise measurement that would be reasonably comparable to a larger electronic instrument that dissipated 100 W in that cabinet.

.

 

[G._S._Ohm]

121201-1405 EST

G._S._Ohm:

Where in the original post is there a block diagram, or even much of a specification on the power supply? The point of my post was to try to get MichaelGP3 to think about the fundamental factors in answering his question.

On heat transfer. To a large extent if you put X watts of thermal energy inside of a box with no forced air inside the temperature rise in the box will be a function of the outside surface area of the box. In other words the inside to outside thermal resistance of the box. There are gradients inside and outside and therefore temperature variations inside the box.

Cut the outside surface area of the box in half and for the same interior power dissipation the interior temperature will about double.

You could power a 100 W incandescent bulb in the center of the cabinet and get an interior temperature rise measurement that would be reasonably comparable to a larger electronic instrument that dissipated 100 W in that cabinet.

.

The link in #2.

The rest I'll think about and I'll be assuming no convection currents.

 

[gar]

121201-2226 EST

The author of post #2 was not the author of post #1. Thus, we really don't know if the power supply is a switching type regulator. However, that it is is a good assumption.

Put a fan inside the box and get better heat transfer from the source in the box to the inside surface of the box, more uniform heat transfer, and a reduction of the internal thermal resistance, and there will somewhat better transfer to the outside. But still to a large extent it is the box size that has a major impact on temperature rise.

Put fins on both the inside and outside of the box and the thermal resistance from inside to outside ambient will be lowered.

At one time we tried using cooling fans inside. These are a mechanical problem. We changed to reducing the internal losses, and increasing the box size. Much better reliability. In a factory environment you do not want holes in the enclosure. Even with no holes oil or grease (gunky stuff) still gets inside. By using no fans and other design improvements reliability was increased by years.

.

 

[Besoeker]

It seems that so long as clearance for air space is provided within the enclosure a vented enclosure is not necessary. Thanks.

Assuming the enclosure is wall mounted with space round the sides, the bottom, and the top and is made of steel, then it can dissipate about 22W for a 10C internal rise.
Given that your PSU is rated at 120W if 5A, the box should be OK.
If it's plastic or any other not very good thermally conducting material it could be problematic...

 

[G._S._Ohm]

it can dissipate about 22W for a 10C internal rise.
Given that your PSU is rated at 120W if 5A, the box should be OK.
If it's plastic or any other not very good thermally conducting material it could be problematic...

A thermal resistance, Theta (case to ambient) of ~0.5 C/W is not too bad.