Automated Garden Owl Scarecrow

by AKeddie in Circuits > Gadgets

1471 Views, 14 Favorites, 0 Comments

Automated Garden Owl Scarecrow

IMG_3014.JPG
IMG_3016.JPG
IMG_3010.JPG
IMG_3012.JPG
IMG_2948.JPG
IMG_2974.JPG
Schematic_OwlBert_2023-07-06.png

**Updated 7-7-2023** Please see last section for updates. I have fixed grammatical and technical errors. Repairs and additions to electronics due to failures has also been updated. I am pleased to present "Owlbert the garden protector". This year I cut an 16' x 8' section off our pool deck which gave us the room to add a nice new garden. We have many squirrels and rabbits that frequent our yard due to the woods in the backyard. Since October of last year, I have been getting more experience in electronics. I have been an automotive engineer for almost 30 years and my electronics skills were basic and limited to simple troubleshooting and wiring. I searched for electronic Owls to setup to watch the garden but wasn't really impressed for the price. I have collected a large cache of electronic components and decided I would like to build something myself. Ultimately it would be cheaper to buy a premade automated Owl unless you have many electronic components on hand. But really where is the fun in that? Owlbert is made from old school discrete components. He is automated with a PIR motion sensor, a speaker with a high-pitched noise output, flashing eyes, and a vibration motor used in cell phones/game controllers. These four circuits are run using one USB rechargeable 9-volt battery. To create more complexity, there are three different voltages being used. After completing Owlbert, an Arduino would probably be an upgrade. Also, I have thought of improvements and will document them at the end. I will explain how I mounted Owlbert, but you may need to modify this setup for your needs. One consideration on packaging Owlbert was how to keep the electronics from getting wet. I created an overhang perch, and the electronics sit below. The project box was a bit big to mount inside and would limit access in case of repairs or adjustments to the PIR. He was created with two perches on either side of the garden so he can be moved so that the animals think he is actually alive. The circuits I used are not uniquely designed by me and I will document my sources for reference and educational purposes. I consider myself an amateur in the world of electronics and would be considered a "Monkey See, Monkey Do" type. This is my first Instructable so please be kind, but also please correct me if I'm wrong. It's the only way I will learn.

Supplies

IMG_3003.JPG
IMG_2887.JPG
IMG_2899 (2).JPG
IMG_2915.JPG

The first section will list the electronics with fasteners and other mounting hardware. The second section will outline will list the items and hardware used to mount Owlbert in the garden area. The third section with outline the tools I used. There are a few items that changed along the way. There are also some fasteners not shown in these pictures, but I will add them to the list here. They will be shown in the accompanying pictures as they were used. Also, some of the items shown in these pictures changed due to improper sizing, accidental damage, and found to be not necessary. I will list them below before starting the supply list.

Item Changes:

  1. The small plastic containers for housing the PIR were not used for my setup. You may consider them for yours.
  2. I did not use the 1 watt black speaker shown. The sound was not loud enough to be useful.
  3. The white Apple earbud speaker was damaged when soldering. I ended up using another Apple earbud speaker from an older pair of earbuds.
  4. I used a larger size plastic plate so that I could seal Owlbert with caulk.

Electonics-Items that I purchased for this project will be bold and the others are parts are already had:

Quantities for each item will be put in parentheses.

Mounting Hardware - This customizable to your needs - Items that I purchased for this project will be bold and the others are parts are already had:

Quantities for each item will be put in parentheses.

Tools:

  • Soldering Iron with Solder
  • Breadboard and connecting wires and 9 volt power source
  • Small Flat Head and Phillips Screwdrivers
  • W-Crimp Tool for JST Connector Pins
  • Razor Blade Knife
  • PCB Holder or Helping Hands
  • Hot Glue Gun with Glue Sticks
  • Super Glue
  • Caulk
  • Step Drills
  • 1/4" Drill Bit or step drill
  • Cordless Drill or Cordless Impact Driver with Phillips screw bit
  • Clamps

Breadboard Prototype Build

IMG_2876.JPG
IMG_2877.JPG
IMG_2878.JPG
IMG_2895.JPG
IMG_2896.JPG
IMG_2894.JPG
IMG_2893.JPG
Automated Owl Breadboard Test
Schematic_OwlBert_2023-07-06.png

The inspiration for this build came from my imagination along with researching premade decoys. I have used the PIR control circuit for many of my Halloween props so that was a comfortable starting point. The vibration motor was not difficult since it was only installed to the power and ground. I have not previously done the NE555 flasher in this setup by using Pin 3 going high and low. I have also not done a sound generator circuit either so both of these were relatively new to me. The sources I used are listed with each circuit breakdown. There are websites that can provide a better explanation of the circuit than I can at this point. I can get circuits to work and even tinker with them, but I'm not an expert enough to explain them with much confidence. I have always had good luck getting things to work even when I didn't fully understand their operation. I'm good at following direction which benefits me in this type of pursuit. It's not that I don't want to learn how everything works, but time is not on my side with a full-time job to support my wife and five small children. My supervisor helped me through my Halloween crash course on electronics and then challenged me to more complex circuits that we could use to test cables for integrity and under 1.5 ohms resistance. It was several months of research, building, and almost quitting. He would offer advise here and there but forced me to figure it out myself. I attached the wiring to the breadboard as per the schematic and used the rechargeable battery as a power source. Unfortunately, I could not find a proper dual power supply symbol in EasyEDA so I used a picture instead. Not very professional, but gets the job done. I realize this combination of circuits is going to put a strain on the battery so that is why I used a rechargeable one. But I have the owl set to run about 10 seconds after each activation. My hope is the battery will have a long life since it runs short periods of time. I will have to update this Instructable in the next month to provide feedback on battery life.

There are four circuits linked together using three different voltages of 3.3V, 5V, and 9V. To make the circuit less complicated I used one LM7805 voltage regulator and a 3.3V/5V dual power supply. It worked out well for packaging.

The PIR/5 volt Relay circuit was posted on Pinterest by @mohanmahalingamvm. Follow the link for more information (Pin on AVG (pinterest.com). I used this circuit for visual understanding but have modified it to suit my needs. His circuit uses the normal closed side of the relay and I use the normal open. I have made several modifications such as the 9-volt battery is used to create the 5 VDC supply by use of a LM7805 voltage regulator. The PIR can take a 4.5V to 20V input, but only outputs at 3v with not enough mA to run any of the other circuits. The BC547 gives the extra power it needs to open the relay so the bigger 9V can be used. I tried to run the PIR with the 5V output from the LM7805, but it turned on, but would not cycle thus I used the 9V battery input instead.

The easiest circuit was the vibration motor. This one is rated at 3.7 volts which is more like 3.3 volts. These are used in cell phones and video game controllers. All this circuit required was to install the power and ground to the 3.3-volt rail of the dual power supply. Please note that these sensors have tiny wires and will vibrate them off unless secured well. I had two of them shake the wires off. I found that adding more hot glue to the wiring helped keep them together.

The sound generator circuit was meant to make a scream. It really did not work out that way and even with changing capacitor rating, it made no difference. The other issue was the 1 watt 8 ohm speaker was too quiet for my liking, so I tried an Apple earbud speaker which was much louder. Interestingly, the earbud speaker was 32 ohm so I can't explain why it was louder. Unfortunately, the original earbud speaker broke on me when I tried to solder another set of wires on to it. Since I did not have another one in that style, I used an older version of an Apple earbud which was also 32 ohm. The circuit I referenced comes from Circuitstoday.com with the Admin being the author (Sound/Horn Frequency Generator Circuit with Diagram (circuitstoday.com). I only eliminated the push button and used the switched 9 volt power source from the output of the relay. I wish it could have been louder and an upgrade would to investigate an amplifier circuit although there wasn't any room left on the circuit board when I finished.

The last circuit was the NE555 flasher circuit for Owlbert's eyes. I have used blinking circuits before but not this configuration. This one is setup how police car lights work. I found this one on Pinterest as well with a link to YouTube by author Mehdi Jahanian (YouTube | Police lights, Electronics circuit, Circuit (pinterest.com). I decided to run this circuit on 5 volts from the other side of the dual power supply. I was worried that 5 volts would not run the NE555. I had one TLC555 low voltage timer just in case, but NE555 worked just fine.

After installing all the components, I turned it on to check proof of concept. It all worked well, and I was satisfied with the exception of the sound generator.

Putting It in on the PCB and in the Box

IMG_2908.JPG
IMG_2911.JPG
IMG_2924.JPG
IMG_2925.JPG
IMG_2927.JPG
IMG_2928.JPG
IMG_2929.JPG
IMG_2931.JPG
IMG_2933.JPG
IMG_2922.JPG
IMG_2923.JPG
IMG_2940.JPG
IMG_2944.JPG
IMG_2945.JPG
IMG_2946.JPG
IMG_2948.JPG
IMG_2960.JPG
IMG_2949.JPG
IMG_2950.JPG
IMG_2951.JPG
IMG_2952.JPG
IMG_2953.JPG
IMG_2954.JPG
IMG_2955.JPG
IMG_2956.JPG
IMG_2957.JPG
IMG_2959.JPG
IMG_2963.JPG
IMG_2964.JPG
IMG_2965.JPG
IMG_2968.JPG
IMG_2961.JPG
IMG_2967.JPG
IMG_2970.JPG
IMG_2937.JPG
IMG_2935.JPG
IMG_2972.JPG
IMG_2971.JPG
IMG_2977.JPG
IMG_2979.JPG
IMG_2973.JPG
IMG_2974.JPG
IMG_2975.JPG
IMG_2978.JPG
Automated Owl Wiring Test

Usually after I check my circuits on the breadboard, I source all the same components to add to the PCB. This way if I run into trouble, I can check it against the breadboard. In this case, I thought I would get fancy and not keep my backup on the breadboard. My hubris got me when I installed the sound generator circuit. I thought that since the ElectroCookie PCB is setup exactly like a breadboard, I would not have any issues. After several "lazy" troubleshooting techniques such as changing parts without proof they were defective, I decided to do the right thing and rebuild the circuit on the breadboard. So, I desoldered all the components and setup the breadboard to figure out what I did wrong. Once I got it to work on the breadboard, I used new components on the PCB. In the process, I found my mistake. I put the 56k resistor from Pin 2 of the BC547 to the positive side of the 30 uf capacitor instead of ground. Not only did the circuit not work, the BC547 was getting very hot. It's amazing how much abuse a transistor can take at the hands of an inexperienced amateur.

I first installed the relay circuit and 3.3v and 5v dual power supply. The next step was to add the LED flasher circuit and then the sound generator. I saved the vibration motor for last since it was just a matter of soldering it to the 3.3v rail. I was pretty confident that I couldn't mess that up and luckily I didn't.

I then prepared the battery box and project box for the installation of the completed electronics. Packaging is not something I enjoy because I rarely get it right or make errors that ruin the overall aesthetics. I'm actually pretty satisfied with this packaging considering that I needed it to be waterproof, accessible for adjustments, repairs, and portable without removing components. I was able to accomplish those goals although time will tell if Owlbert's electronics can handle rainy days. But, since he is easily removable from his perch, he can be stored in a dry environment on stormy days.

I started on the box by drilling a 7/8" hole using a step drill. I then had to open it a bit more with a razor knife so that the bulb of the PIR sensor would fit through flush. I also had to grind off tabs on the inside of the box as well. The next step was to drill a 5/16" hole in the lid near the end for the 9 volt battery power supply box. I installed the battery box using 3M double sided tape. This stuff is strong and suited well for this. Then the rubber grommet was pushed in and the wires sent through. You could easily use silicone instead to keep the water out. I moved on to the plastic plate and added a two pin JST connector. The battery box is mounted so that after unplugging the connector you can slide the box open to remove the battery for recharging with a micro usb cable. If you buy the two pack of batteries, you can always have one charged for swapping out. The plate needed to be 3.5" away from the mounting area on the 4 x 4 so it would not interfere with the rear thumbscrews. I used a 5/16" bolt to hold the box and plate together and added four roughly 8mm holes for attachment of the box to the plate. The holes were counter sunk and M8 x 22 pan head bolts were attached with lock nuts. After removing the 5/16" bolt from the center hole, I added grommets. Four 1/4" holes were added to the four corners of the plate about 1" from each end. The PIR was installed with the adjustment screws facing down (future adjustments) with some super glue on the edges of the bulb and them on the tabs to the circuit board after an accidental drop test. A two pin JST connector was installed in case the PIR needed to be replaced in the future. Around the outside of the PIR bulb, I put a bead of hot glue to seal out moisture. I finished soldering on the LEDs, speaker, and vibration motor. Then I added a four pin JST connector to the LEDs. Please do this after you fish the wires through the eye sockets of the Owl. I could have connected the grounds together, but these small JST connectors are hard to get a good crimp. The speaker and vibration motor were each given a two pin JST connector respectively. The last installation step was to put four blobs of hot glue on the circuit board and attached it to the lid. I thought this would make room for extra lengths of wire from the body of the Owl and if I needed to do repairs, it would be easier to access.

Finally, I did a test before closing up the box. In the process, I set the time on screw for about 30 seconds. After final installation outside, I opened the box and set if down to around 10 seconds. 30 seconds seemed too long and could drain the battery too fast. Also, I believe that multiple short on times might create more attention.

Wiring and Mounting Owlbert to the Plastic Plate

IMG_2984.JPG
IMG_2985.JPG
IMG_2986.JPG
IMG_2987.JPG
IMG_2988.JPG
IMG_2989.JPG
IMG_2990.JPG
IMG_2991.JPG
IMG_2992.JPG
IMG_2993.JPG
IMG_2994.JPG
IMG_2995.JPG
IMG_2996.JPG
Owl Scarecrow Test

I started by cutting groves in the eye sockets and installing the LEDs as flush as possible after fishing the wires through the body. I did not install the LEDs into the eyes themselves because they are solid molded. My though was the drill would scuff the plastic and make the LEDs dull. I used hot glue and super glue to hold the eyes to the head. I cut grooves in the white plastic washers that go behind the eyes that increase brightness. The eyes were reinstalled with a dab of hot glue and the grooves open over the LEDs.

I placed a piece of double-sided tape on the back of the vibration motor and placed inside the base. In hindsight, I should have only placed the tape on the wires so that the motor bounced around inside a bit. My only concern was breakage of the wires and sensor itself.

I intended to mount the speaker in the circuit box under Owlbert due to decreased volume. A coworker convinced me it would fail due to the elements. I put some small holes at the base of the Owl and using double-sided tape, I installed the speaker on the base facing up at the holes. In my opinion this was a mistake as it is difficult to hear the high-pitched sound versus the vibration motor while running. It's possible to hear it, but you have to listen for it. It does change pitch with the flashing of the eyes which must be some electrical noise in the circuit.

I used two machine screws about an 1/8" in diameter to mount Owlbert to the plastic plate. After which, I laid a bead of caulk around the base for a moisture barrier.

I did another test to ensure the circuit was still operational and it all worked.

Preparation/Mounting the Perches and Owlbert

IMG_2915.JPG
IMG_2981.JPG
IMG_2980.JPG
IMG_2982.JPG
IMG_3003.JPG
IMG_3004.JPG
IMG_3006.JPG
IMG_3007.JPG
IMG_3008.JPG
IMG_3010.JPG
IMG_3011.JPG
IMG_3016.JPG
Completed Automated Owl Scarecrow
Owlbert with shorter on time
Owlbert at night

I gathered up the wood 4 x 4s and the plywood sections which I stained. I placed the plastic piece on the two pieces of wood which were clamped together. This way the holes would be the same on both perches. I outlined the holes and used a 1/4" drill bit open them up. I held one end of the plastic plate at the end of the perches that were clamped to together. I outlined the circuit box onto the wood perches and cut it out using a jig saw. The holes were cut slightly bigger than the circuit box to accommodate the PIR, make it easier to remove, and give some slack in case I incorrectly measured. It all worked well luckily for me. The plate mounts to each perch using four 1/4"-20 thumbscrews, washers on each side, and 1/4"-20 wingnuts. This hardware allows for easy removal and installation.

I screwed the 4 x 4s one on each side to the deck with three countersunk 4" wood screws. The perches were mounted with three 3" wood screws. Since one of the 4 x 4s was warped, I added two 1/4" washers between each 3" screw under the plate.

I mounted Owlbert using the thumbscrews and wingnuts. Two more tests were conducted. The PIR was set for about 30 seconds of on time and maximum range setting of 20 some feet. This test was done during the day. It went well and the PIR was quite stable with no false reactions. The automation was not as dramatic as I hoped but worked as planned. I then reset the PIR to about 10 seconds and did another test in the late evening. The outcome was more dramatic as the eyes where more noticeable.

Conclusions and Possible Upgrades

Overall, I'm pleased with Owlbert Garden Protector circuit and enjoyed this project. It will be interesting to see if he keeps pests out of the garden and how long the battery will last between charges. I will update as the summer goes on.

Possible upgrades.

  1. Use Arduino which would make wiring easier with better control.
  2. Louder sound circuit with a waterproof speaker mounted on the outside of the Owl.
  3. Mount the vibration motor so that it bounces around more inside the Owl.
  4. Better power source.
  5. Installation of a solar panel/charger.

Humorous Update After 24hrs of Deployment

Owlbert Vibration Motor Bouncing

I went out this afternoon to check Owlbert's operation. He was still working well except he must have taken one of my updates seriously. I noticed that he was much louder than yesterday. Apparently, the vibration motor came loose from the double-sided tape and is now bouncing inside the body of the Owl. It makes the operation more dramatic which improves the ability to scare off pests. The speaker output is still low and can be heard when you are very close to the base. Time will tell if the wires or motor will fail due to bouncing around. I will update of Owlbert's deployment as events occur. I have posted a video of the new operation. Thank you reading my whole long Instructable.

Update, Repairs, and Improvements

IMG_3115.JPG
IMG_3116.JPG
IMG_3119.JPG
IMG_3122.JPG
IMG_3124.JPG
IMG_3126.JPG
IMG_3127.JPG
IMG_3128.JPG
IMG_3129.JPG
IMG_3130.JPG
IMG_3131.JPG
Schematic_OwlBert_2023-07-06.png
Improved Vibration Motor Design
Amplifier Test
Test of Repairs/Additions
Improved OwlBert

After a couple of days of the vibration motor bouncing around inside Owlbert, the ground wire broke. About another day after that, the battery needed charging. I can only guess, but I believe he was activated about 50 times. It could have been less or more. Since the battery needed to be charged, I thought it would be a good time to repair and upgrade the electronics. The new and improved OwlBert was deployed on the evening of July 6th. I will update again when the battery needs charging. The best way to provide feedback on the battery will be in days and not activations since I cannot track them consistently.

I removed the caulk around the base of the Owl and separated it from the plastic base plate. I removed the vibration motor and speaker.

I have a large number of LM358 op-amps in my parts bins, so I researched an audio amplifier circuit for use. The LM358 is really not considered a wise choice for audio amplification, but for my purposes it would be fine. I still had my sound generator circuit on the breadboard, so I added the amplifier circuit I found. The design came from LM358 Audio Amplifier in Non-Inverting with Split Resistor Biasing Configuration experiment 1 | ee-diary with the author being ee-diary. In the end I did not use the circuit as shown completely since I found that is worked better with a peizo buzzer than a speaker. I have posted the latest schematic of what I did. There is also a video of the buzzer versus the speaker. Please note that when using a piezo buzzer there is no need for the sound generator or amplifier circuit. It could have been directly wired to power and ground from the relay output. Since I already had the circuit on the board, I just added to it instead of taking it all apart. I believe the sound from the buzzer has a different pitch due to how I wired it, but again it is not necessary. It also was a learning experience for me to work with another circuit.

The vibration motor hardware needed to be upgraded to avoid breakage. I came up with a solution for that as well. After repairing the wiring, I modified a plastic pill holder I bought some time back from Amazon. I added a small hole in the side for the wires. I placed the vibration motor inside the holder and hot glued it to cover all the wiring especially the solder joints. I took care not to glue the motor housing down completely so it would have room to vibrate. I also hot glued around the hole in the side of plastic holder where the wires came through. After testing, I liked the setup.

The new amplifier circuit soldered together, but not using a pcb. It was mounted to the bottom of the electronics box with hot glue and the wires attached accordingly. I decided to mount the piezo buzzer outside of the Owl. But I made sure to setup the jst connector is such a way that I could mount the buzzer in the Owl at a later date if necessary. My only mistake is that I mounted the buzzer on the bottom of the base plate. That required me to open up the electronics through hole on both perches.

To put it all back together, I put the new vibration motor back inside Owlbert was reattached the plastic base plate. After an operational test, I re-sealed the base of OwlBert to the plate. I re-deployed him in the backyard and did another test. The combination of the piezo buzzer output and vibration motor output is balanced with neither one drowning out the other. I am much more satisfied with this setup and time will tell if the vibration motor and speaker setups are robust.