Welded Steel Grill Cover

Our pellet grill kept getting clogged with wet pellets. We had a plastic fabric-type cover to drape over it, but sometimes forgot to put it on. Then the plastic got old and brittle and the chimney made a hole in it. We bought a new one, but it doesn't fit well. Cleaning out the damp pellets was messy and wasteful.

We thought it would be nice to have an umbrella or gazebo to keep the grill dry and make it possible to use the grill when it is raining. But gazebos are expensive and aren't the perfect size. They are often flimsily made with thin short material to be able to be flat packed. Fabric umbrellas are awkward, flammable, and not squirrel proof. I wanted something more "skookum".

After a lot of thinking, shopping, and planning, we decided that a custom welded patio cover with a metal roof would be perfect because I know to weld. 

The patio is 10 feet wide, so we purchased 12 foot pieces of 2 inch square tubing from a metal supply place. The 12 foot pieces barely fit in a rental truck. We bought corrugated metal roof panels from a different metal supplier. We also got plates for the feet, anchor bolts, a couple pieces of angle iron, and screws for the roof.

The square tube components of the roof frame were welded together in the garage, as well as the feet to the legs. Then I welded the legs to to the frame onsite. The exact placement of the legs on the frame changed slightly from the plans based on what position looked the best. After that, I welded extra angle iron to the frame. We used scaffolding and 4x4 blocks to support the structure while I welded it together. The structure was stable, but could be tipped over if we pulled on the back overhang. 

To make it even more stable, the feet were bolted to the concrete. We started by marking the location of the holes, drilling into the concrete, and then hammering the bolts in. The 16 holes weren't lined up perfectly with all 16 bolts, so we had to enlarge some of the holes in the foot plates. In hindsight, it probably would have been much easier to bolt the feet plates first and then weld them to the legs. Once the frame was bolted down, all three of us could hang from the back without it tipping or collapsing.

After the frame was complete, we painted it with several coats of primer and black paint. Then we screwed on the metal roof panels. It was difficult to line up the screws in the middle, especially on the angle iron. The final step was to add covers over the bolts on the feet.

Automatic Plant Watering System

    

plants

    For the Renewable Energy Systems class I took last semester, I created an automatic watering system as part of a group project. The first step in the assignment was to research our motivation for the project: Agriculture uses 70% of the world's fresh water. Watering by schedule can be wasteful and cause overwatering. Advanced smart watering systems use sensors, logic, weather data, and a series of pipes, pumps, and valves to optimize the plant watering. This can make farming more water-efficient and less labor-intensive. Our goal was to create a functional prototype demonstrating automated irrigation technology.

closeup of sensor and moss

    The next step of our project was to select a soil moisture sensor. Conductive sensors measure the electrical resistance of the soil. Over time, conductive sensors corrode because the electrodes are in direct contact with the soil. Capacitive sensors use the change in capacitance of the soil due to its water content, and have coated circuit boards. They avoid direct copper contact with the soil and don't use DC currents. We chose a capacitive sensor because it seems more durable for extended use.

sensor

    Another choice we made was the type of irrigation. Drip irrigation applies water directly to the root zone of the plants instead of spraying into the air. Combined with a pump, drip irrigation is good for the precise water delivery needed for our small-scale experiment. A gravity-fed valve system was also considered but was not practical due to space limitations.
    One team member created a simulation with a Tinkercad starter circuit design for a soil moisture sensor connected to an Arduino Uno R3. Unfortunately, the “water pump” in the simulation doesn’t actually change the soil moisture level. The sensor readings had to be controlled manually with a slider. Tinkercad isn’t the best tool for this because it doesn’t include soil-specific parameters.

Simulation (from group report)

    For our watering system build, I started with a kit that included a pump, relay, sensor, and tubing. The kit can be found at https://amzn.to/4ekJHHS (Amazon Affiliate link). I connected it to an Arduino UNO R3 clone and wrote some initial code to read the sensor value. The higher the sensor output value was, the drier the soil. In the code, I called this value "unmoistness." I had the pump turn on for a specified time when the soil reached a certain dryness. The system would then wait for one minute and measure again, watering if necessary.  The pump was not powered continuously, allowing time for water to soak into the soil.  

pump

    I placed the pump in a plastic container with holes drilled for the tube and power cable, and then sealed the holes with hot glue to utilize the full reservoir capacity. The sensor was placed inside and hose was clipped onto the edge of a large pot with several plants including peppermint, Cuban oregano, chives, bunching onions, chrysanthemums, and local moss. The dryness threshold, pumping time, and wait time were adjusted for optimal plant health.

    I coded a watchdog counter to prevent the pump from running too much in case of a sensor error, an obstructed pump, or an empty water tank. This type of oversight could be helpful in a large-scale system by preventing expensive pumps from being damaged due to running dry or overheating. The watchdog counts the number of consecutive attempts and prevents the pump from running if there have been too many attempts. The watchdog is reset if the moisture reading is within the target range.

Control board with watchdog tripped

    An LED indicates the current moisture relative to the target. It blinks n+1 times after each measurement, where n is the number of steps moister than the target value. This blinking system eliminates the need for a complex LCD requiring constant power, unlike similar experiments which have both a 9 Volt battery for the pump and USB power for the Arduino, display, and sensor. Our total system power is approximately a quarter of a watt when the pump is off, and about two watts when the pump is running. 

Flow chart (from group presentation)

    The system was mostly functional, however, I had to do some troubleshooting. The pump occasionally got stuck on because the Arduino got locked up due to the interference from the pump. I added redundant lines of code to turn off the pump, but this revealed that the original code was not the issue. Next, I replaced the relay with a MOSFET, which helped slightly. Then I installed a capacitor across the 5 Volt rail, eliminating the issue of the pump getting stuck on.

Block Diagram (from group presentation)

 

    Unfortunately, after moving the system home, there were many issues caused by loose wires. There was also water damage to the original Arduino clone so it was replaced. The wiring issues were corrected and the system was back up and running with only the need for adjustment of target. In order to allow the system to be used at night on solar power, I added a light dependent resistor to turn the light off at night and adjusted the blink time so it would be visible during the day.

System set up at home with new Arduino

     

    I tested the system for a few weeks before I moved it to make sure it was reliable. The results were better than manually watering a similar set of potted plants I kept in the same location, and there were no percolation losses. I was able to dissolve fertilizer in the water tank to water and fertilize at the same time. The system can also be expanded to have more plants because Arduino Uno can support up to five zones, each with its own sensor, moisture level settings, pump, and light.      

 

References
Our report was:  Sensor-Based Automatic Watering System by G. Colello, J. George, A. Smith

[1] S. Gould. “A Guide to Soil Moisture.” ConnectedCrops. https://connectedcrops.ca/the-ultimate-guide-to-soil-moisture/

[2] H. Ritchie and M. Roser, “Water use and stress,” Our World in Data, 2024. https://ourworldindata.org/water-use-stress

[3] M. Price, “Irrigation Fundamentals – Water Well Journal,” Water Well Journal, Nov. 18, 2019. https://waterwelljournal.com/irrigation-fundamentals-9/

[4] R. Leighton. “The 3 Main Types of Irrigation for Agriculture: the basics.” GrowerExperts. https://www.growerexperts.com/the-3-main-types-of-irrigation-for-agriculture/

[5] “HIRALIY Indoor Drip Irrigation System.” Amazon. https://www.amazon.com/dp/B0D7VYV9JN/

‌[6] V. Sonwane et al., "Automated Plant Watering System with Soil Moisture Tracker," in 2024 5th International Conference on Image Processing and Capsule Networks (ICIPCN), Dhulikhel, Nepal, 2024, pp. 735-739, doi: 10.1109/ICIPCN63822.2024.00127.

Solar Power- Charge Controller Upgrade and Solar Panel Relocation

 The USB port of my second charge controller didn't seem to work right. I hadn't noticed when I first got it, but now that I tested it something didn't seem right. It signaled that it could provide 7 Watts, but it couldn't even provide 1 Watt. 

I bought another new charge controller, and this time, the USB actually works. Both the USB and 12-Volt parts can be turned off automatically or with a button. I haven't measured the exact numbers. This charge controller also displays the load and generation current.

For the last few years, I've had my solar panel in my window or on the porch roof.  When I moved it to the porch roof, I added a thin cable to pass through the edge of the window while still being able to keep the window closed and locked. The porch roof is not the most stable location because I couldn't permanently attach the panel. It also gets some shade from a large tree, and it's hard to get to the panel to clean it. I recently had to redo the pass-through cable and the connection from the wires to the panel after discovering that the battery wasn't charging properly. Because I was already checking the wiring, I decided to move the panel down to ground level.

The location on the ground isn't perfect. The neighbor's house blocks the sun for part of the day. But there's less shade from the large tree. I added a longer cord that goes up to the roof and then connects to the window pass-through cable.  I was able to get a maximum of 36 watts which is decent for a 50-watt panel at a sub-optimal angle with a non-MPPT charge controller.

    Now that I have a working USB port, I hooked up my automatic plant watering system to it. 

Ring Light with 3-Way Switch

 I found a ring-shaped light in the trash, and decided to add it to my solar power system. It was intended go around a camera for photo and video lighting. I decided to hang it from my ceiling because it turned out to be the perfect size to go around the existing overhead light fixture. I chose to hang it from the existing light fixture to avoid making any holes in the ceiling. I wasn't worried about the weight because the electrical box originally held a ceiling fan. I needed something thin enough to slip behind the light fixture without damaging it, so I used some twine. 

After I hung the ring light from the existing mains-powered light fixture, I connected a power cord that crossed the ceiling and then attached to a 3-way switch pair. The ring light is connected to my 12-volt solar power system. With the 3-way switch, I can turn the ring light on from the doorway or from my bed. 

This adds to my collection of devices powered by my 12-volt solar power system. I have 4 LED lights, the ring light, sound system, roller blinds, phone charging station, dinner bell, and a plug for other devices like a pencil sharpener. Unfortunately, the marine starter battery that I've been using to store power is now 6 years old. It is still taking a charge, but might not have as many amps available. I've been doing charge/discharge cycles from mains power to see if that will help.