sat, 14-dec-2013, 15:37

I spent most of October and November building a dog barn for the dogs. Our two newest dogs (Lennier and Monte) don’t have sufficient winter coats to be outside when it’s colder than ‒15°F. A dog barn is a heated space with large, comfortable, locking dog boxes inside. The dogs sleep inside at night and are pretty much in the house with us when we’re home, but when we’re at work or out in town, the dogs can go into the barn to stay warm on cold days.

You can view the photos of the construction on my photolog

Along with the dog boxes we’ve got a monitoring and control system in the barn:

  • An Arduino board that monitors the temperature (DS18B20 sensor) and humidity (SHT15) in the barn and controls an electric heater through a Power Tail II.
  • A BeagleBone Black board running Linux which reads the data from the Arduino board and inserts it into a database, and can change the set temperature that the Arduino uses to turn the heater on and off (typically we leave this set at 30°F, which means the heater comes on at 28 and goes off at 32°F).
  • An old Linksys WRT-54G router (running DD-WRT) which connect to the wireless network in the house and connects to BeagleBone setup via Ethernet.

The system allows us to monitor the conditions inside the barn in real-time, and to change the temperature. It is a little less robust than the bi-metallic thermostat we were using initially, but as long as the Arduino has power, it is able to control the heat even if the BeagleBone or wireless router were to fail, and is far more accurate. It’s also a lot easier to keep track of how long the heater is on if we’re turning it on and off with our monitoring system.

Thursday we got an opportunity to see what happens when all the dogs are in there at ‒15°F. They were put into their boxes around 10 AM, and went outside at 3:30 PM. The windows were closed.

Here’s a series of plots showing what happened (PDF version)

The top plot shows the temperature in the barn. As expected, the temperature varies from 28°F, when the heater comes on, to a bit above 32°F when the heater goes off. There are obvious spikes in the plot when the heater comes on and rapidly warms the building. Interestingly, once the dogs were settled into the barn, the heater didn’t come on because the dogs were keeping the barn warm themselves. The temperature gradually rose while they were in there.

The next plot is the relative humidity. In addition to heating the barn, the dogs were filling it with moisture. It’s clear that we will need to deal with all that moisture in the future. We plan on experimenting with a home-built heat recovery ventilator (HRV) that is based on alternating sheets of Coroplast plastic. The idea is that warm air from inside travels through one set of layers to the outside, cold air from outside passes through the other set of layers and is heated on it’s way in by the exiting warm air. Until that’s done, our options are to leave the two windows cracked to allow the moisture to escape (with some of the warm air, of course) or to use a dehumidifier.

The bar chart shows the number of minutes the power was on for the interval shown. Before the dogs went into the barn the heater was coming on for about 15 minutes, then was off for 60 minutes before coming back on again. As the temperature cools outside, the interval when the heater is off decreases. Again, this plot shows the heater stopped coming on once the dogs were in the barn.

The bottom plot is the outside temperature.

So far the barn is a great addition to the property, and the dogs really seem to like it, charging into the barn and into their boxes when it’s cold outside. I’m looking forward to experimenting with the HRV and seeing what happens under similar conditions but with the windows slighly open, or when the outside temperatures are much colder.

Meta Photolog Archives