25 September 2013

Arduino chicken door project taking shape in hardware

You may recall last year we built a chicken coop. Well, since then it's been a daily chore to open and close it dawn and dusk to keep the chickens in and other critters out. I mentioned last October that I was using a microcontroller to automate this task, so now I'm finally soldering it together:


As I mentioned earlier in this blog, I'm also making a smart swamp cooler controller, but as winter approaches getting up to let the chickens out in the snow makes this project a little more urgent though.

Some people like to etch their own printed circuit boards. I've never tried it, but I understand the acid used is nasty stuff. So, I just bought some "perfboard" which is basically a universal circuit board. Connections are made on the bottom side by globs of solder or in the case of my +5VDC rail, a random chunk of wire (a clipped lead from a resistor, I like things that are free). Topside, I made connections with little bits of 24AWG Cat5 wires. (Cat5 is cheap and I had some laying around.)

The project consists of the Atmel Atmega328 AVR microcontroller (the rectangle on the bottom right of the brown board), a ChronoDot real-time clock (blue circle top left), a voltage regulator (green sticky uppy thing top middle), and a twin relay module (off to the right, connected by wires). Not shown are a red and a green LED that I'll aim at the (human) house so we can see if the door is open (red) or closed (green) by looking out the window. Also not shown is the $10 thrift store cordless drill that will actually raise and lower the door by turning a threaded rod like a screw drive garage door opener.

The blue rectangle sitting separately and with the green and white wires coming off it is an Arduino Nano, which is used to download my program from my PC and into the Atmega microcontroller.

The whole thing will be powered by a solar cell (the kind sold to keep your car's battery from going dead) and a little 12V motorcycle battery. To save on battery power, the microcontroller actually spends most of it's time asleep -- only waking once per second to get the date and the time of day from the ChronoDot, calculate the current position of the sun for my latitude and longitude using the TimeLord library, and based on that information activate the relay module to run the motor and raise or lower the door if appropriate. Whew!

The brass brillo pad you see in the pic is something new to me, a waterless soldering iron tip cleaner. It works as well as a wet sponge, and doesn't turn my soldering iron's tip black as much as water does. I also purchased a "sal ammoniac" block to clean built-up black stuff of the iron tip. Makes a lot of smoke and actually works (if the iron is hot enough); these two items should save me some money on buying new iron tips all the time.

10 September 2013

Swamp cooler review

Well, it seems summer has ended and autumn has begun. I've waited this long to mention how our swamp cooler was working, to make sure I could objectively review it's performance.

The verdict?

The swamp cooler has one flaw: it's too cold.

If you think that sounds like a glowing endorsement, that's because it is. I've read many opinions on swamp coolers online, and many of them are quite negative. They make your house soggy, some say. It can't compare to the performance of refrigerated central A/C, say others.

Now, I am admittedly biased since I spent many hours and dollars installing mine, so I'm liable to suffer from "sour grapes" syndrome -- I have a lot invested in my swamp cooler, so I would look foolish if it turned out to suck, and I don't want to look foolish. So, I offer the following objective observations:

The performance of a swamp cooler is directly related to outdoor humidity.

The relative humidity in Idaho is typically too high for good swamp cooler performance at night and in the morning. A swamper might only achieve a 5-10° drop in air temperature. But, outside air temperatures -- even in the peak of summer in of July/August -- often fall to 60-65° at night.

The relative humidity in Idaho is typically low enough for an evaporative 20 to 25° air temperature drop the afternoon and evening, when outside temperatures are around 90°.

When outside temperatures rise above 90°, relative humidity drops proportionally and the swamper achieves a 30 to 32° temperature drop.

What does all this mean, qualitatively? When it's cool outside at night, you only need a 10° temperature drop to make the house downright chilly. When it's 90-95° out, you only need a 20-25° degree drop to make the house a lovely 70°. And when it passes 100°, you still get 70° air coming out of the swamper.

In short, it works, and works well.

In fact, on many occasions, it works too well, outputting 65° air and anyone in the family room has to put on a blanket. We could turn the cooler off when this happens, but it's a hassle to go turn the cooler on and off all the time. A thermostat would solve this, but more on that later.

How about humidity?

Well, as I mentioned, Idaho is very dry ("semi-arid", whatever. Locals call it a desert). Many of our family members suffer from dry skin (elbows, knees) so the added humidity is very welcome. Our house has never gotten soggy because we open a few windows halfway, on the far side of the house. Exhaust air is essential for swamp cooler effectiveness! If you keep the windows shut, all you get is wet with no cooling.

Table salt does not clump. Doors do not stick. Furniture does not warp.

Ah, but what about the smell? My answer: what smell?

When our pads were new, they smelled wonderful (I love the smell of aspen). Sadly, this only lasted a month or so. After that, nothing. I have installed a "bleed kit" that constantly removes a trickle of water from the cooler whenever it is running, which brings in fresh water. This means the water is not in the cooler long enough to spawn life forms. Also, each night I run the cooler on "vent only" (the water pump is shut off) which allows the pads to dry. Bacteria and mold have a real hard time growing on dry pads.

But you say, the window thing is an issue: can't leave it running when you're not home (open windows being an invitation to burglars) and it's a pain to keep opening and closing them all the time. Well, this is true.

The solution is up-ducts in the ceiling of each room. These automatically open when you switch the swamp cooler on and exhaust air into the attic (thus cooling the hottest part of the house, killing two cooling birds with one stone). I have some up-ducts, but haven't installed them yet. Maybe next year.

The great thing about up-ducts is you can put a thermostat on your cooler and at that point, it's just like central refrigerated A/C -- set it and forget it, it takes care of itself. I should have that ready to go by next summer too, soon as the up-ducts go in.

At this point, I'm laughing all the way to the bank. At the peak of summer this year, our electric bill was less than $70 (and that includes a lot of irrigation water pumping). Electricity is cheap in Idaho, but from what I understand, most Idahoans pay $100-$150 a month to cool a 1,700sqft house. So the savings are definitely significant.

And while the cooler cost quite a bit more than I expected -- closer to $1,500 -- that's still a fraction of the price of central refrigerated A/C, and I can fix it myself with a screwdriver if it ever breaks. Try asking an A/C tech how much it costs to replace a coil or compressor, and get ready for sticker shock. Even the diagnosis for A/C is typically $100, whereas I can fix pretty much any part on my swamper for less than $100. And I can do it the same day and not wait weeks for the tech to fit me into his schedule.

Swamp coolers are useless in humid environments -- so they are, admittedly, pointless in much of the US. But, in the arid parts of California, east Oregon, east Washington, Utah, Colorado, Nevada, and New Mexico... they rock.

02 September 2013

Lumberjack, part II

What do you do when 40 years ago someone planted a row of pine trees spaced about 5 feet apart and now they're all scraggly and shading each other? And what if one of those trees is leaning towards the house? Well, a smart person would call a tree removal service.

I'm not smart.

I cut off the lower limbs and tied a couple of ropes to the trunk up as high as I could.


Then I tied the other ends of those ropes to the bumper of my little car. Yay for 1980's bumpers.

Then I cut most of the way, but not quite all of the way, through the trunk just below my ropes. That's probably a good 15 feet off the ground. You don't want to actually cut all the way through, because that trunk could go anywhere and if you're on a ladder, your only escape route is straight down. Your choices then become limited and all of them painful.


With the tree structurally weakened, I climbed in the car, revved the little 52HP diesel engine and popped the clutch! My car only weighs 2,200lb but it won the tug-of-war:

Once the top is removed, the lower section can be cut conventionally because it's too short to fall on the house.

I definitely do NOT recommend this. However, nobody died today (except the tree) so I guess it worked out ok.

We actually cut down three more trees today. The biggest was right in front of our living room window. I didn't have to do anything special there, because it was naturally leaning away from the house. It was a textbook cut, fell right where I aimed it. Here's the new view: