Heat producing device with Arduino

I'm trying to build a arduino controlled heating box with temperature control. Roughly I want to have a temperature sensor (using LM35) to verify current temperature, a fan for cooling, and now I'm trying to find how to produce heat.

The box would be something like 30x30x30 cm, and I would need to have somewhat uniform temperature inside it, ranging from min 10ºC to 40ºC max. I've researched heating pads and resistance wires, but am not sure if and how any of them would be suitable for this, since I'm trying to heat an air mass (so to speak) instead of a surface.

Any ideas/help? :cold_sweat:

Hi, light bulb, no brainer. If you can still buy them, incandescent type. Probably only 25 or 40Watt for the size box you have.

Tom... :)

The most simple way would be an incandescent lamp. Keep the fan going all the time to homogenise the air in the container and adjust or turn on/off the lamp with a PID to maintain required temperature.

I did think of an incandescent bulb, but in terms of energy consumption and dissipation I guess it isn't the best solution, or am I totally wrong? (also they're getting increasingly hard to find :P)

Didn't think about keeping the fan going to homogenise temperature, thanks! :)

Also: if I do use an incandescent bulb, how would I go about using a single power source? Since most bulbs are 220v, should I use a 220v source to get the bulb running and then convert it to the 5v for the arduino, sensors and whatnot?

Maybe I'm saying something absurdly wrong, I'm a newbie in what concerns power sources and all that, so I'm kinda thinking out loud :drooling_face:

ladansedesdamnes: I did think of an incandescent bulb, but in terms of energy consumption and dissipation I guess it isn't the best solution, or am I totally wrong? (also they're getting increasingly hard to find :P) You do loose a little efficiency as your also producing light but your only talking about a couple of percent. ;)

If you’re only looking for a few tens of watts, you could use an automotive 12V bulb, driven via a suitable circuit. For example various dual h-bridge motor driver circuits can handle a few Amps at 12V.

That keeps all the power electrics nice safe easy low voltage DC which makes it easier to control and avoids all the safety issues related to mains voltage electrics.

Was checking on 12v bulbs from your advice, and it seems to be a pretty good choice. Pretty easy to get 12 down to 5 with a regulator, and 12v wall warts are common enough to find a generic one which will work.

So my question would be if 12v bulbs will generate enough heat for this purpose?

Decide what wattage you want. Pick a bulb or selection of bulbs that use this much power. For example, 5W and 11W are common automotive lamp ratings so a small number of these would give you what you need. If you use an h-bridge driver to power them you'll probably have several outputs (four is common) so you have a lot of flexibility to turn them on and off or PWM them individually, or just turn everything on if you want full power. If you're planning any sort of closed loop feedback such as PID then having the ability to control the amount of heat output would be a significant benefit.

Thanks a lot for the advise Peter :D

I know I'm a complete newbie :disappointed_relieved:, but is there a way to estimate wattage from desired heat generation? Is there anything concerning heat dissipation on the lamps' datasheets or something of the kind?

I was thinking about using a simple SSR to turn heating element ON/OFF, but having individual control would indeed be very useful. I've used H-Bridges to run motors before, but in all honesty never fully understood how they really work (shame on me), just followed instructions, so not really sure how I would go about using one for this application. Am googling around a bit, but if you happen to a have a nice link or something that might enlighten me, I'd be very thankful :blush:

Hi, an incandescent lamp is a very inefficient light source, however it is a very good heat source, almost 99% heat efficient.

Tom...... :)

Well, come to think of it that does make a lot of sense... They're inefficient light sources because of massive energy dissipation through heat :) So that actually makes it the best choice for this haha

To estimate the power in watts required to heat the inside of your box, you need to know the R-value or thermal conductivity k of the insulation (i.e. the wall material), the inside and outside temperature and the total area of the outer walls, ceiling and floor of the box.

The outward heat flow in watts is then very simply calculated from the thermal conductivity equation. See http://hyperphysics.phy-astr.gsu.edu/hbase/thermo/heatcond.html

Great link! Thank you very much! XD

So, according to this, for a box of dimensions 40x30x30cm (A=0.66m2), considering cork walls (k=0.04), 2cm thick (disregarding surface over cork - plywood or acrylic, for example), to achieve an inner temperature of 30ºC with outside temperature of 5ºC, the conduction heat loss rate would be of 33W.

Does this mean I can work it out with roughly 30W worth of lightbulbs, or should I have MORE than that to compensate for heat loss through wall dissipation?

ladansedesdamnes: is there a way to estimate wattage from desired heat generation?

If you're asking how to choose the bulb to generate a given amount of heat, the bulb Wattage is how much heat it puts out so estimate how much heat you need in Watts and choose the bulb(s) accordingly. If you want to control the temperature it would be sensible to have some excess power available. The nice thing about this approach is that if your initial estimate is too low you can simply swap bulbs to increase the heat available, and if it's too high you can always reduce the heat output by turning bulbs off or PWMing them. Just make sure that the power supply and driver circuit have a suitable margin of excess capacity available, because you don't want to have to replace those if you can help it.

According to that calculator jremington was kind enough to provide a link to, considering worst case scenario of having an outside temperature of 5ºC and needing to maintain an even 30ºC, there would be 33W dissipation through the walls.

I found this calculation somewhere, checked the math and it does look ok.

Say 70% of 50 watt of Halogen Bulbs converts into heat (hypothetically)
i.e. = 0.750 = 35 watt of heat
Bulb is on for 5 minutes = 60
5 = 300 sec
1 watt = 1 joule/ sec, thus 35*300 = 10500 joules = 10.5 KJ
Specific Heat of Dry Air (Cv) = 0.716 KJ/kg.K
Thus, 10.5KJ/0.716KJ/kg.K = 14.6 Kg.K
Density of Air = 1.3 Kg/m3

Thus 14.6 (Kg.K)/1.3 (Kg/m3) = 11.23 K-m3
This means temperature rise will be 11.23 K per m3 of dry air.
Converting Kelvin to Celsius
11.23 C temperature rise.

So, to keep that calculated ~11ºC rise in 5 minutes, considering the 33W loss from wall dissipation, I would have to have 2x50W lightbulb to get roughly 37W of heat.

Am I correct or completely wrong?

EDIT: No, wait, 50W to get 11.23º/5 minutes is for 1 m3 air mass. Considering my box is considerably smaller than that, I’d need less wattage to achieve the same.
Damn, can’t think, too sleepy… Will get back to this tomorrow morning.

The heat flow estimate says nothing about the time it will take to heat the box up -- for that you need an estimate of the heat capacities of the wall material, the air and whatever else is in the box, plus an estimate of the maximum heating power available. You've started that process with the most recent post, but you have to take the walls into account too.

The heat flow equation allows you to estimate the minimum power required to maintain the temperature difference of 25 C in your example. It may take quite a long time to get up to that, if you supply only the minimum heating power (or maybe never, if your estimate of the heat flow is too low).

So let me see if I'm getting this right: to be able to maintain that 25º difference, I must have AT LEAST the calculated 33W. If for any reason I need to make the difference broader (rise the internal temperature, considering no outside temperature change) I should provide more than that power to be able to do it.

The time it takes to rise is actually not that important for this project, I'm just trying to understand as best I can (which might be hard haha) the power needs and consumption for it to work, so I can start thinking about power source and how to drive the bulbs. If I end up only using 1 single bulb, maybe a relay is better than using an h-bridge; on the other hand, if I do need multiple bulbs, an h-bridge probably makes it easier to manage.

Just found this while lurking around ebay: http://www.ebay.com/itm/AC-DC-12V-PTC-heating-element-max-50W-heater-warmer-melter-unfreezer-defreezer-/291076787974?pt=LH_DefaultDomain_15&hash=item43c5874f06

Would this be a good alternative to using a lightbulb? If so, what would be a good heat dissipation surface to maximize the heat radiation? A common ceramic plate?

So, summing up everything until now: - use either lightbulb or that ceramic heater as heating element; - around 40-50W needed for described purpose; - 12v voltage would do, while keeping it safer;

New considerations: - considering 50W@12v, the heating element would use around 4.17A; - Arduino draws under 0.5A, with sensor, lcd screen, LEDs, H-Bridge and pots; - pwm fan would draw somewhere between 0.12 and 0.4A (from what I managed to research).

12v for powering the heating element and PWM fan, and a 7905 to get it down to 5v for the rest, and that would be it for power.

So in total I would need a 12v@6A power source, correct?

Would this be a good alternative to using an H-Bridge, considering only one heating element? https://www.sparkfun.com/products/10636 Just turning it ON/OFF as needed?