Mini-fridge for I.T. in loft ?

[Ronald Raygun]

Yes. Against.

If you're using the fridge only for the thermal insulation it provides, instead of activating its cooling capability, then its internal temperature will swing less wildly about the mean than that of the roof space generally.

Or that would be the case if you didn't put heaters into the fridge, which in a manner of speaking is what the equipment you're hoping to put there are. They may not be consuming a lot of power, but remember that the fridge is well-insulated, so the heat can't get out quickly, and so the internal temperature will rise. Perhaps even to more than outside.

And if you do power up the fridge, you may end up with condensation inside.

It's a daft idea putting IT equipment in the loft. If the reason people have their nerve centres up there is that it's easier to run comms cables through the ceilings of all rooms and then just randomly above the rafters, then why not put the nerve centre on a wall shelf in the hallway, mounted high up so that it's just below the ceiling, next to the loft access hatch? All the wires can then come down from the loft into it, it won't be in the way, and it will be in a reasonably temperate climate.

[www.GymRatZ.co.uk]

I have a G-bit switch and loft box in the roof along with a Network storage device as this is the heart of all home network terminations and have been pondering about creating a more controlled enviroment against the summer ravages, then stumbled upon the idea of a mini-fridge. The type with a glass door. I'm thinking this would provide a much more friendly and dust-reduced enviroment for these low-powered items. Not powering the fridge up obviously just allowing it to do it's thing. Any thoughts for or against?

[Dave Liquorice]

Why would you want to put heat producing objects inside a sealed, insulated, box, then place said insulated box in an enviroment that can get stupidly hot and quite cold?

[Huge]

Put a small patch panel in the loft and the "nerve centre" in the cupboard under the stairs.

[Ian Jackson]

If you put a heat-producing object in a perfectly insulated box, the object will simply get hotter and hotter and hotter until such time that it fails (burns out, catches fire, explodes etc), and afterwards, is no longer capable of producing heat. It's a variation of the "What happens to a sausage in a fridge in a perfectly-insulated room?" question.

If the box is not perfectly insulated (such things don't exist), the object will still get hotter and hotter until such time as a state of equilibrium is reached, and the heat lost from the box equals the heat being produced by the object. Obviously, the condition of the object will depend on how good the insulation is.

I've seen an example of the 'well-insulated box' effect when I ran a temperature test on four Cable TV set-top boxes, left running in a thermal chamber over the weekend at minus 20C. Unfortunately, the power for the cooling failed and, on Monday morning, I found that they had been 'well cooked'.

[Skipweasel]

IT's (that started as a typo, but made sort of sense) is going to get rather hot in a nicely sealed and insulated box.

[Huge]

The "Bang goes the theory"(*) programme once cooked a chicken in a well insulated box with a 100W light bulb ...

(* A perfect example of how deeply shit popular science programming is. Ghod knows why I was watching it. Oh, 'cos I wanted to shag Liz Bonnin. I've gone off her a bit now.)

[Alan Deane]

I've been running a gigabit switch, NAS and loftbox in a crude loft-room (not insulated much, wide temperature variation) for years without any issues. I keep an eye on the NAS disk and CPU temperature on hot summer days (and it's configured to email warnings of over-temp anyway) but have not seen anything to be concerned about.

[Andrew Gabriel]

Reminds me of one of my stock physics questions...

You leave a freezer operating in a perfectly thermally insulated room. Does the room a) get colder, or b) stay the same temperature, or c) get warmer?

I like it because the 3 possible answers tell you how much the person has understood about the thermodynamics of a freezer (or any heat pump system), as in nothing, or a little, or a lot.

[Andrew Gabriel]

Like others said, not a good idea.

You could hack the thermostat so it was set to 21C, and power it, which would be slightly better. However it gets up to 50C in my loft, and I wouldn't be surprised if that would kill a fridge. If you had to open it on a hot day, the kit inside would get condensation on it (although with the door shut, a fridge is a dehumidifier).

If you had an old one knocking around to try it with, might be worth it for experimental value. I wouldn't buy one for the purpose though.

Some of the small glass door ones I've seen use Peltier effect heat pumps, which are way less efficient than a compressor.

[Skipweasel]

Talking of which - I took apart a cheap slow-cooker the other day. The damned thing has no insulation at all - the heating element is on the inner wall of the shell and there's a bit of an air-gap, but that's it.

It's the model that appears to be badge engineered with every name conceivable on it, and sold from a tenner to forty quid.

one.

[www.GymRatZ.co.uk]

Mine has also been running fine for the last 3 years or more too. ;¬)

I held back at putting the router up there though so that and the modem are sat on/under the office desk where they're happy.

Did think of passive cooling of a box by a chimney vent and bottom inlet from the room below. Just one of those random ideas that popped into my head. Too many other projects to sort out first. :¬)

Thanks for putting my mind at rest though. My Linkstation can mail through alarms for fan failure etc etc but doesn't have any temp. monitoring so as long as it's all backed up the other bit's aren't a problem or big expense to replace.

Cheers All for the input though. Pete

[Ronald Raygun]

I reckon the answer must be C.

If the experiment begins with the freezer not having been operating, so that its inside is at room temperature, then consider first what would happen if the freezer door were left open. Then cold air would pour out of the front, while the black coils on the back would get hot. Thus all three answers are possible, as part of the room will get colder, and part of it warmer, but on average it would stay the same, so by this reasoning the answer should be "all of the above, but effectively B".

If you close the door, it's not much different except that now the cold "leaks out" more slowly. However, with the inside of the freezer sealed from the outside, it is no longer meaningful for the "room average" to include the freezer's interior. The freezer's inside gets colder, the outside gets warmer, so the answer is C, at least initially.

If the freezer enclosure itself is as perfectly thermally insulated as the room, then the room will get warmer until all the heat has been pumped out of the inside, which cannot get colder than zero kelvins. From then on, there is no more heat inside capable of being pumped out, so the outside can't get any warmer. So the answer is first C then B, i.e. the room will get warmer than it was, but will then settle at that temperature.

Now the heat pump is in practice going to have inefficiency losses, and they will contribute to warming the room. So the ultimate answer is C, i.e. it will get warmer and keep getting warmer forever.

Also, in practice, the freezer won't be perfectly insulated and it will be impossible for the interior to reach absolute zero. At best a certain temperature differential can be sustained. Now, with the outside getting warmer, the inside must also be getting warmer (while staying perhaps 30 degrees cooler than the outside). With the pump consuming more and more energy, the room will get hotter and hotter until even the inside of the freezer is like an oven.

[Chris J Dixon]

What strange language that review is written in. I wonder if it was Google translated?

Chris

[www.GymRatZ.co.uk]

Thanks Andrew and everyone else. I guess blowing off the dust when the NAS needs re-starting following the occasional power cut will suffice for the time being. Perhaps I will re-visit this idea when (if) something actually dies. Got a working one for our dogs insulin in the project cottage but if I stumble across one at the tip I may play around with a top/bottom vent, perhaps an old PC fan to draw house air through when required.

[Skipweasel]

Dunno - didn't read the review.

[BartC]

I thought the answer is usually (c) simply because the room has a net input of N Watts of electricity, likely to end up as extra heat energy. That the appliance is a freezer is irrelevant.

(It does seem odd though, that with much of the UK at the same temperature as the inside of a fridge for several months of the year (as I tried to explain the concept of Cold to someone in the Caribbean once), we waste huge amounts of money recreating the same conditions inside a box in our heated homes!)

[www.GymRatZ.co.uk]

LOL. Over Christmas I foolisly suggested to the bringer of food that because the fridge was full and no-one could get to us to help consume the abundance we could simply put the extra stuff in a box in the garage as it was around 2 or 3 degrees C. Needless to say only the wine and some veggies were allowed out there as they were.

8¬O

[Andrew Gabriel]

It is C. Unfortunately, I missed out an important part of the question, that the freezer door is open. It doesn't make any difference to the answer, but it provides a possible reason to incorrectly select a).

This gives the 3 options for level of knowledge:

a) The room gets cold because the freezer cools it due to door being open.

b) The room stays the same temperature, because the heat extracted from the freezer is emitted again from the rear.

c) It requires energy to pump heat energy from a cooler to a hotter location (Carnot's Theorum), and this additional energy is emitted from the rear too, or to note as you do that the freezer consumes power, which will end up in the room.

[Owain]

They're all strange.

I think they're written by people in China for 5p a review and intended to attract search engines.

Owain