Last month, we ginned up a little temperature sensor that will give us a digital output for a preset temperature from near room temperature (68° F or 20° C) to below freezing (32° F or 0° C). Now what we have to do is find out how to keep those low temperatures from affecting the materials and/or aircraft that we fly. In short, we need to heat up the paint box or preheat the engine to keep the effects of cold from making mush of the paint or tar of the engine oil.
The three light bulbs used for testing. Left to right, the 100-watt floodlight, the 60-watt clear incandescent and a 250-watt true heat lamp.
A very short test in some moderately cold temperatures (32° F or 0° C) showed that for any reasonable amount of protection from freezing weather, a heater for a “garden-tool”-sized outdoor shed that can handle 5 gallon cans of paint or other liquids plus small tools and accessories would have to heat 72 cubic feet (about 2 cubic meters) and should be at least 250 watts. Translating that into current, a 12-volt aircraft battery would have to supply nearly 21 amps of current. For the typical 35-amp-hour battery, that translates into a run time of a little less than 2 hours. That isn’t going to work.
With the 12-volt supply now off the design table, the next most available source would be standard 115-volt AC “house power.” Given that supply, our minimum 250-watt heater would work itself out to a little over 2 amps. That starts to be reasonable.
The small relay on the left will work up to about 700 watts of heater power. The larger relay in the middle works all the way up to the 1500-watt heater. The mini-box on the right will house the small relay, but the next size larger will be necessary for the big relay. Part values associated with these parts are Magnecraft W88KDX-2 for the small relay, W199SDX-2 for the larger relay and Hammond 1411BU for the mini-box. None of these parts is critical. The usual suspects (Mouser and Digi-Key) carry them all and Jameco has cross-references to them.
The next week of fiddling around (excuse me, heuristic engineering) led me to some inescapable truths. One is that there are good ways of design-ing semiconductor replacements for the venerable mechanical relay. The next truth that I discovered is that one little wiring error and you rapidly find yourself in a room full of smoke and sparks. Not a good thing, especially if you haven’t been doing this sort of thing all your life. That and the local fire investigation board may take a dim view of putting an unproven home-engineered device into a box of flammable materials.
Better that we go with the tried and true: relays. Now the problem resolves itself into finding a suitable relay and a heat source.
Field Testing
I did a whole bunch of testing in what has been the coldest California winter in my memory. I started off proving my initial theory that it would take 250 watts of heat to provide even marginal protection from really cold weather. Nature cooperated for a few days with temperatures into the low 20s to high teens (0° C to -5° C). But what to use for a heater? Then the light bulb (literally) went off, and I remembered that a light bulb gives off about 0.5% of its energy as light and 99.5% as heat. We’ll use light bulbs as the heat source.
On to the house light-bulb junk box. In there I found a 60-watt, 100-watt and 250-watt bulb. I tested each one of them for 2 hours to see exactly what sort of temperature rise I could get with the cabinet doors shut as tight as I could get them. Here are the results in Fahrenheit (Celsius). Δ Temp is simply the heat rise over ambient (OAT) over time:
Skyfall
Then disaster nearly struck! Quite by accident, as I was changing from the 100-watt bulb to the 250-watt bulb, a chunk of snow fell from the roof, missing the hot bulb by less than an inch. Had that snow hit the bulb, we would have had shattered glass all over the hangar from the popped glass bulb. A bit of thought said that if there was condensation of any sort in the cabinet during unattended operation, then the odds of that bulb surviving a direct hit on the glass by dripping cold water were nearly zero. The result then would be a shattered bulb, but even more important, the filament wires inside the bulb would still be connected to the AC line.
While pure water is a nonconductor, little rainwater is pure, so now we have a wet bulb, a wet cabinet floor and the odds for an electric shock have risen dramatically. This is not an acceptable engineering outcome.
What to do? A little rummaging around the local hardware store turned up a small electric heater with a forced-air fan that cost little more than the big light bulb and was completely housed in a plastic case. Not a bad solution, especially as it had a heat control on the top that would shut the heater off if the little temperature sensor we designed last month failed in the “On” mode. Now that is a value engineering compromise that I can tolerate. A few bucks for double redundancy and a lot more heat if we need it.
See the table above. Note that now our margin for low temperatures is raised from 25° F (14° C) to nearly 50° F (27° C). That is, the temperature can plummet all the way down to -17° F (-27° C) and we will still be protected. More? You want more? OK, that same heater has a “HIGH” 1500-watt mode, and we can now roughly guess that those margins will just about double.
The inset shows the circuitry for the temperature sensor, while the rest of the schematic shows how to wire up a relay and a heat source.
So what shall we do for circuitry? Pretty simple. From the output of our sensor (last month) we simply provide a relay, a transistor to turn the relay on, a diode to protect the transistor and “electricals” that you feel comfortable with. I’d recommend a small metal housing for the relay with a 3-wire grounded cord for the input and a true weatherproof box for a duplex outlet for the heater. Simple, cheap and safe. That’s the way we like it.
Inexpensive forced-air heater. Note the power level (right) and temperature-limit controls on the top.
So, what’s up next month? Well, I can’t promise, but I’m working on a scheme to get Internet service to the hangar from a remote location. I may or may not have that experiment done in time, so rather than promise that which I’m not sure of being able to do, let’s just leave it that I’ll research something interesting. And remember, it was Neil Armstrong who said, “Research means to investigate something that you do not know or understand.” Vaya con Dios, Neil.
