Building a miniature 'Warm Box'....need advice

No those battery sources will die out way too quickly to be of use...

That would be one of my first choices...

Likely not, even though some of the high wattage ones do produce heat, a bulb is better...

Sure...

There is another thread her about doing this exact thing to grow mushrooms... Lots of ways to go about doing it, it's more a matter of how involved you want to get, how experienced you are with electronics, your budget, among other things...

 

Coincidentally I came across these yesterday:
http://www.coolcomponents.co.uk/catalog/heating-5x10cm-p-1024.html

Theyre made by sparkfun if I recall (that site resells alot of their stuff).
You could use something like that. As for batteries, the minimum feasable would probably be D cells, maybe even a couple of banks in parallel or a lantern battery (which actually have about half the amp / hour rating of D cells)

I was thinking of getting some for heated animal boxes / blankets for Hessilhead (wildlife rescue centre where I volunteer)

 

Resistors make very efficient heaters.
As the others have said, of course unless your application really demands battery power, you'll be getting through an excessive number of batteries.

There are some nice example circuits dotted around for proportional temperature controllers, typical sort of op-amp feedback circuits. This type of arrangement will avoid the 'bang-bang' control characteristics of using a thermostat (more like a thermal limit switch) you linked in.

 

Thats a very good point, I'll be entirely honest and state I didn't do any calc's to base my response on.

You could do a good first order approximation from thermal resistance of a material similar to what you've made your box from.

A colleague of mine is currently using a similar sized box made from three layers of 15mm thick expanded polystyrene (measuring losses from a power converter). The internal temperature of the box raises about 5 degrees C per watt of power dissipated in the box. If that puts you in the ball-park of what to expect

 

In a closed box, a resistor, a light bulb, nichrome wire, or a heating element will all produce the same amount of heat per Watt i.e. 1 Watt of heat for each 1 Watt of electrical power. All of the power is going into heat.

The amount of power needed to maintain a temperature will depend on the amount of thermal insulation, the surface area of the box, and the difference between the inside and outside temperatures.

Bob

 

Look at how heaters work, if the device is only pumping out 110 degrees it will take a LONG time to raise the ambient temp of of a larger area, also the volume of heat exchange is a huge factor... For example your furnace in your house produces about 1000-3000 degrees F when the flame is burning over say a 16x20 inch heat exchanger with forced air flowing over it, all to maintain your house at a happy 68-74 degrees... When you strike a match it's about give or take 1500 degrees F, how far will that match get at heating something?

Now look back at your 110 degree heater and it;s short comings as a 'heater' should be more clear...

There is also insulation and lose factors to consider, that is why you basically 'brute' force heating and cooling in most cases vs taking the long slow road... A highly isolated never opened 8x8x8 cube will not take much at all to maintain once it reaches a temp, but if there is a lose like opening it or lack of insulation that bleeds the heat, maintaining it can be troublesome... If you want the temp to recover quickly or at all you need to pump the heat in or wait a long time for that to happen, if it happens at all...

 

What is the purpose of this 'warm box'. If it is something that is just going to be run for a long period of time with quite slowly varying external temperature then the answers should be slightly different to if it needed fast recovery or had very rapidly varying temperature extremes.

If you don't require very fast temperature recovery times, then lower temperature heating elements such as resistors would be more beneficial, whereas if you need a very rapid transfer of energy from the heater to the air then a higher temperature (lower thermal mass) element would be required.

Then in terms of selecting the controller, your requirements of accuracy would be useful. If you don't mind a bit of temperature overshoot and undershoot (see bang-bang control) then a very basic bimetallic thermostat would be right up your street, whereas tighter regulation of temperature would require a much more delicate solution.

 

I like the look of the controller you've linked in, it's pretty much what I would suggest. The key to keeping the overshoot minimum with a controller like that is finding a way to keep the thermistor as close the the air temperature as possible (the thermistor temperature will lag that of the air because it takes time for the air to transfer energy to the sensor), I would advise using a small aluminium heatsink, they're as good at taking energy from the ambient as they are at giving it.

The other key to a project like this is to experiment a bit, thermal stuff is very difficult to predict without painfully accurate maths. Try putting a few resistors, perhaps 10 or so 1/4 watt ones, pick a value any value and drive them with enough volts to make the dissipate near their 1/4 watt rating and see what the temperature does. Good practice in ohms law as well!

If things are a bit slow, you need more power.