Week 8: Attaching the Servomotor Hub Horn to the Feeder Cover

The feeder cover is the only freely detachable part of the feeder once it is in use. As the only part of the feeder that will rotate freely the cover will need to be attached to the hub horn that rotates once it is fitted onto the servomotor. Initially the cover was attached to the hub horn by means of a ⅝" VELCRO Mini Fastener [1]. One half of the fastener was located on the hub horn, the other on the feeder cover. 

Figure 1. ⅝" VELCRO Mini Fasteners

However on testing the feeder it was discovered that each time the cover rotated, due to the unstable nature of the velcro, the cover experienced a "back-spin" effect. As the feeder is designed to be able to rotate to an exact amount of degrees, this was a problem. The solution was to remove the velcro attachment and attach the hub horn directly to the cover itself. Using hot glue, the hub horn was fixed at the centre of the plexiglass cover. This resulted in more accurate rotations and a more stable structure.

Figure 2. Feeder Cover/Servomotor Hub horn attachment

Week 8: Servomotor/Centre-Piece Assembly

After cutting the circular plexiglass servo support and attaching to the servo itself, it was realised that it would not create a stable enough foundation if we were to create shelf-like extrusions on the inner rounded surface of the centre piece. Another factor was that extra support would be needed to keep the servo elevated without the weight of the servo (and stand) dropping at any given moment in time.

As a result, a second structure was cut from plexiglass; a circular ring with an outer diameter that fits exactly into the inner compartment of the central PVC piece, acting as a stable "shelf". The ring's inner diameter is wide enough to fully support the servomotor stand and allow the connected servo wire to lead to the electrical compartment.

This supporting ring was fixed at a distance of __in from the upper edge of the centre piece. This distance was determined by the height of the servo. Knowing that the servo will be fixed at a certain position while the feeder cover is removed on a day to day basis, the shelf was set at a distance where once the feeder cover is replaced, the feeder is airtight and the cover would be able to rotate freely.


[INSERT PIC]

Measurements:
Inner diameter of plexiglass ring - __in
Outer diameter of plexiglass ring - __in
Distance from upper edge of PVC piece - __in

Week 8: Changing intervals

The Arduino is very beneficial for the pet feeder. The pet feeder is going to be presented in front of a panel of people. The pet feeder is meant to rotate 45 degrees every 4 hours, however, the time interval has to be accelerated in order for the judges to observe how the machine works. The fact that an Arduino is being used allows the group to simply change one line of code to change the time interval at which the top of the pet feeder rotates. The rotation will now occur every ten seconds to quickly display the use of the feeder.

However, since the actual interval is for 4 hours, the line that states 10000 milliseconds has to be changed to 14400000 milliseconds.

Week 7: Inner Compartments

This is the general shape of the compartments:

[PIC]

Limitations

There are a few limiting factors to keep in mind when designing the inner compartments:

1. The center needs to be large enough to house the mechanical/electrical workings
2. There needs to be 8 compartments for 4-days worth of food
3. The compartments need to have a minimal number of nooks for easy cleaning
4.  The design must keep in mind the limitations of the printer
• The printer prefers small pieces to larger ones
• The printer prefers a minimal amount of material used
• The printer cannot print dimensions larger than 12 inches
5. The pieces must completely separate the food compartments from the electrical components
6. The compartments housing the food must be adequately separated from its neighboring compartments.

Combating these limitations:

1 The Center

The servos is the only piece of hardware that will be in the space of the of the centerpiece.
The servo's axis of rotation is not exactly at the center:

From servocity.com

C=1.352"
L=0.174"

So the maximum distance from the centerpiece to the edge is 1.526 inches. This means that the centerpiece's interior needs to be larger than 3.052 inches

2 Eight Compartments

The compartments must allow for multiple days of convenience and care. The entirety of the container can be swept radially from the center by 360 degrees, so the mid-line of each eighth needs to be 45 degrees from the previous.

3 Easy Cleaning

A way to avoid corners made by the meeting of the floor of the container and a vertical ____ is to camber the corners. The assembled piece(s) will slope downward to achieve this effect.

[PIC]

4 Printing limitations

The lab is equipped with a few printers, each with ____ of function. The largest prints a maximum of ______ and it is preferred that the pieces being printed take under __ hours to print and that they don't use under 1 cartridge of material.

5 Separating electrical components

Week 7: Elevating the Servomotor

Now that the central piece and dividers have been made and set up within the Feeder, we need to look at positioning the servomotor so that the aluminium Hub Horn [1] of the servo aligns perfectly at the centre of the Feeder cover.

Measurements (Feeder Cover):
Diameter of plexiglass Feeder cover - 13in
Central point of Feeder cover - 6.5in
Diameter of Aluminium servomotor Hub horn -  __in

Fig 1. Aluminium Servo Hub Horn 

The Servomotor will be located within the closed-off central piece and will need to be elevated in order for the hub horn of the servomotor to be able to attach to the Feeder cover. The basic structure requires the Servo be fastened to a flat surface (circular plexiglass sheet) that will fit exactly into the central piece. This surface will need to have an opening/hole that allows the servo's control wire to lead to a separate electrical compartment (located below Feeder container). Because the hub horn is offset from the central point of the servomotor, the servomotor's position on the plexiglass surface will need to be relative to the alignment of the hub horn and the central point of the Feeder cover.

[INSERT PIC__drawn out image]

Measurements (Circular Plexiglass Sheet):
Diameter of plexiglass servomotor support surface -
Diameter of supporting plexiglass ring -
Height of Servomotor -
Height of elevated surface relative to base o the Central Piece -

Week 7: Inner Compartment Dividers

As mentioned previously, 8 food compartments will divvy up the available space to carry up up to 4 days worth of food. These compartments will be separated into equal volumes by means of 8 compartment dividers.
Based off of internal measurements, the dividers were cut to fit into the remainder of space after factoring in the central piece.

Internal measurements:
Inner diameter of Feeder container (Cake Tin) - 12in
Outer diameter of central piece (PVC pipe) - 4in
Height of Feeder container and central piece - 3in

Divider Measurements:
Length -  4 & 1/16in
Height - 3in
Thickness - 1/16in
Chamfer -

Figure 1. Plexiglass Compartment Dividers

Week 6: Drawbacks & Solutions

This week, the group was met with its first set of drawbacks. We have now reached the stage of project where the Feeder design is concrete and prototyping has officially gone underway. 

Having received the Servo Motor necessary to revolve the plexiglass feeder cover, alongside the Arduino Uno, batteries and the minor parts that complete the electronic section of the feeder, the team has begun working on the programming necessary to run the servo motor.  During the first round of testing; using a Servomotor, the Arduino Uno, and a 9V battery power source to test the rotation of the servomotor, our initial servomotor was burned out. Due to the high voltage of the 9V battery compared to the 4.8 - 6V range of the Servomotor the Servo overheated and was fried. On purchasing a new servomotor, an alternative power source was used. Instead of the 9V battery, a battery pack that consists of four AA batteries was used. This successfully ran the servomotor. In further posts, another reason for the use of the battery pack is highlighted.

The second drawback concerned the circular plexiglass Feeder cover. Due to the brittle nature of the plexiglass; on attempting to sand down the rough edges of the cover, the feeder snapped in half. A new cover was cut, adding one inch to the diameter so as to increase the coverage of the feeder.

Week 6: Soldering

An issue arose recently with the entry points of the servo motor. There was a lead of a wire that broke off inside the entry point of the servo wiring. The lead was incredibly difficult to remove, the wires were separated from the piece that kept them attached.

The lead was stuck in the black piece connecting the wires, which was later removed

The separated wires were later soldered using a soldering iron. The wires were soldered to the wires that would be connected to the Arduino. There were issues with the soldering iron because it was relatively old and could not properly connect the wires together at first. Now, the issue has been adverted and the circuit works properly.

Week 6: Portability

The selling point of this pet feeder is that it is portable. Because it has to be portable, the power source of the Arduino cannot be 5 volts of power the computer is delivering to the Arduino. Therefore, the best method to make the pet feeder portable is by attaching a five volt battery pack to Arduino in its external power source pin.


The initial issue that the motor could not handle a nine volt battery because it was a 4.8 volt to 6 volt motor. Therefore, an ideal voltage for the motor is 5 volts. However, a battery pack wired in series would be perfect as a external power source. Most likely, there will be four 1.5 volt AA batteries supplying a total of 6 volts of power. In addition, the battery pack is fairly small so it will properly fit in the section underneath the tin pan.

Week 5: Pet Feeder Cover (Plexiglass Sheet)

Plexiglass is a very versatile polymer that is formed when giant giant carbon molecules combine chemically. Plexiglass, also known as Acrylic, is a synthetic plastic material which can be used for a number of applications because it is so easy to work with; it can be sawed, drilled, routed, glued etc.

Fig 1. Plexiglass Sheets

It has many beneficial properties that include:

• Glass-like qualities -  clarity, transparency, translucence - but at HALF the weight
• Stronger impact resistance (10 times greater than glass) and more durable
• Ability to be coated, tinted, coloured, mirrored or made opaque to enhance features such as: scratch-resistance, anti-fogging, solar-reductive
• Smooth surface that would provides little to no resistance when in motion

Due to the easy accessibility and low cost Plexiglass is the most cost efficient material whose physical properties meet the needs of the feeder cover. We were able to mould the sheet to the exact shape necessary to cover the Feeder, as seen below:

Fig 2. Pet Feeder - Plexiglass Acrylic Sheet Cover

Basic Dimensions of Feeder Cover:

Outer Radius (In contact with cake tin) - 6 inches

Inner Radius (Length of compartment opening) - 4.5 inches

Thickness (Of Plexiglass sheet) - 3/32 inches 

Angle (Of Compartment opening) - 45˚

Week 5: Back to the Arduino (Official Switch to Digital Approach)

After considering realizing the difficulty and the danger that would come with working with high the high voltage that would be sent to the timer switch from the outlet, we decided against using the timer switch. The main problem was as I said, the issue with the danger associated with the high voltage and high amperes, therefore, we went back to using the Arduino. The Arduino will be connected to a motor, which will rotate according to the certain degree angles that we input into the Arduino program.

Format of a sample Arduino code

The initial issue with using the Arduino was that we were unsure on how to place it in the system, but now that we have created a hole in the bottom of the cake pan, we will simply lift the pan and place the Arduino underneath the system. We now realize that the Arduino language is a very learn-able language with a section where the user has to input the variables and another section where the user has to insert a prompt that will continue to loop. The Arduino is perfect for our pet feeder system. However, we will need a servo motor for this situation because with that, the Arduino will send pulses to the motor, which will rotate at preset angles. Making a simple DC motor rotate at certain angles and time intervals, would be quite difficult. Therefore, we will use a servo, which we will speak about later.

Week 4: Timer Switch Use (Pros versus Cons)

One issue our group is facing is how the wiring is going to be setup. We need to figure out how to run the timer using a breadboard. Some problems we may face are actually getting the timer to run and programming it run at certain periods of times. The timer runs on 120v AC and 5 Amps, we need to figure out how this is going to be put into a circuit because  it uses a lot of power which is very dangerous.

Fig 1. Timer Switch (Closed)

Fig 2. Timer Switch (Open)

Week 4: The Frame (Pet Feeder Container)

Parts are finally starting to arrive in the mail! We have ordered several parts (some which have since decided against using) and expect to start assembly once they are all here.

The frame --which all pieces, parts, and covers will be assembled around arrived --earlier this week. It is essentially a cake tin that is 3 inches deep and 12 inches diameter. It is a lot larger than I imagined it would be. This is okay because I (speaking for myself only) was afraid that it would be too shallow.

[PIC]
But all is well :)

The tin encompasses 332.29 square inches of space. This is approximately tantamount to 23 cups.

Coming soon, there will be more information on the inner compartments which will divvy up these 23 cups. Ideas of how this intersections will be placed can be made when the rest of the mechanical pieces arrive in the mail.

The image in the post below (Fig.1) is an idea of what the compartment might look like, but it will change. For one, it will have to have a hole in the middle. Second, the slope closest to the center axis will likely change to accommodate a specific configuration of mechanical pieces, which have yet to be designed.

More on this next week.

Week 4: Segmented Food Tray

Now that the cake pan (feeder container) has arrived, the next step is to design the inner compartments that will be used to separate set food amounts. The most ideal method would include the use of the Creo Parametric 2.0 CAD program to design one tray that has 8 distinct compartments for pet food (See Fig 1.).

Each compartment would be of equal size and volume and walls between compartments would be sufficiently thick enough to prevent pet from accessing the next compartment's food supply.

Use of a CAD program and the 3-D printer ensures that the tray (that has specifically been designed for this model) will fit into the cake pan perfectly, leaving no room for food to escape or move around once pet feeder has been activated.

Fig 1. CAD Tray Design in 3-D

Specific details that relate to measurements needed for the CAD design process can be found on the Tutorial page

Week 3: Timer Switch (Timing Intervals of Cover Rotation)

We are going to use a timer switch instead of an Arduino. We found this to be more effective in the way our cat feeder is going to work. A timer switch is a timer that operates an electric switch controlled by the timing mechanism. The switch is connected to a circuit operating from a circuit connected to a breadboard. We will wire the switch so that the motor that will be connected to it will be given power for only certain amounts of time and throughout the day for as many days as the user would like.

Week 3: Arduino Use (Possible Switch to Digital)

An Arduino is a microcontroller that can run on a DC battery or a USB port in a computer. This system allows the user to input certain commands using the Arduino programming language. The commands allow several real life applications. The Arduino can do things like create a light show that controls several LEDs, or gauge water flow. However, initially, this group intended to use the Arduino to control the movement of a motor that would rotate the top of the pet feeder.

This is an Arduino Uno, which is an example of an Arduino the group was going to use

In the past lab session, the group was trying to figure out a ways to incorporate the Arudino in the project properly. The initial issue with the Arduino was trying to learn the language for the Arduino, which was very different from any of the programming that the group has done in the past. However, the group still tried to learn the coding by looking at certain samples. After trying to insert code that the group created and not having it work properly, some of the members found examples that would run a simple DC motor, but they did not work properly either. Even though the group did not receive any results, it still continued to research more on how to make the Arduino work. At the next meeting the group decided that using an Arduino in this situation would not be practical because if a person were to purchase this pet feeder, he or she would need to have a computer, and knowledge about programming. In addition, the group determined that there may be some unpredictability with the Arduino is something went wrong with the code. Finally, the group decided that a timer switch would be best for this project, which the group will write more about in the future.

Week 3: Gear Research & Calculations (Mechanical Approach)

Since we are now sure that gears will be an integral part of the project, it is time to decide how exactly.

The RPM of the motor which connects to the power system via potentiometer will have to be reduced using gear ratios. With 8 compartments and the maximum number of feedings per day being 4 (see first post) the pet feeder needs to rotate 45 degrees (360degrees/8compartments) every 6 hours (24hrs/4feedings).

So a hypothetical motor that rotates with a speed of 400RPM would need a lot of reducing. While it seems lofty, this is possible with 4 sprockets.

[PIC]

The first unit is the motor itself rotating at 400RPM
The second unit (1) rotates at 1RPM*
The third (2) rotates at 1 rotation per hour
The fourth (3) rotates at one rotation for every 6 hours in units of ___ degrees
The fifth unit (4) rotates at one rotation every 48 hours in units of 45 degrees

*The connection between unit one and unit 2 may be expanded.

Research on the Geneva gear:
The final unit can operate in a number of ways. It needs to rotate with 1/8th of it's teeth in short bursts every 6 hours. The Geneva gear is a mechanism that can help our system execute this kind of motion.
Here is an example of this dynamic connection type:

<---Unit 4 rotating 45 degrees every 6 hours




<---Unit 3 rotating 360 degrees every 6 hours



In simulation of this connection type, a dilemma was presented. For unit 3 to rotate the (approximately) 120 degrees needed to move unit 4 for one 45 degree step, 2 hours would need to pass. In other words, it would take 2 hours of motion for one compartment to open.

This is a problem. Other avenues will have to be explored.

Week 2: User Notification (Use of an LED)

One issue that the group was thinking about was how to notify the owner once the pet feeder has rotated a certain amount. The group reached the consensus that the best method of making the owner aware that the container is empty is by placing an LED light at the side of the pet feeder.



Initially, the group was considering to implement an alarm that would send out a low noise when the feeder would have rotated a full rotation. However, the group figured that an alarm would scare a cat off so the group decided to use an LED. The group believes that this would work when the top of the feeder fully rotates, it will set off a switch,which will in turn turn on the LED notifying the owner to replace the food. Then the owner will rotate the top of the feeder back around and turn the feeder back on. This seems to be the most effective, inexpensive and pet-friendly method of making the owner aware of the empty pet feeder.

Week 2: Circuitry (Digital Approach)

There are many different ways that the circuitry could be designed in order for the pet feeder to work. Based off of the knowledge that a DC motor would control the revolutions made by the pet feeder cover, the group debated the potential use of resistors within the circuit.

The group discussed two methods in particular. The first being a circuit that would make use of a 12V DC motor connected to different resistors. Each circuit path would have its own set resistance; which would correspond to a specific (user-input) feeding setting that allows for a set degree of rotation of the pet feeder cover. The difference in resistance would be due to an increased number of resistors (in series) for each circuit setting (see Fig 1.)

Fig 1. Pathways of Different Resistance

The second option would make use of a POTENTIOMETER; which is a small sized electronic component whose resistance can be adjusted manually (a manual variable resistor - see Fig 2.). It is used in various electronics e.g. volume knobs in music systems, fan regulators. Based on this concept the group was able to visualise how the user-dependent dial would help control the number of revolutions made by gears and ultimately the degree of rotation made by the feeder cover.

Fig 2. Standard Potentiometer

Week 2: Use of Gears (Mechanical Approach)

An essential component of out projects is creating gears that would revolve the top of the pet feed at certain times and snap the feeder into place at certain points.

The component we desire to create are similar to these. There will be several notches that will allow the top of the pet feeder to snap into place after certain times. In addition, we believe that the most effective way to create these gears are with 3-D printing. With a 3-D printer, the group could tailor the gears to the feeder and have them properly fit into the feeder. The program that the group intends to use is Creo Parametric to design the gears and this is also one of the programs that are compatible with a 3-D printer. 

There are going to have to be various gears, which the group still has yet to determine. In addition, creating these gears and find how exactly they will work will be a challenge, but the group hopes to have that issue sorted out but week 3 after extensive research has been done. 

Week 1: Preventing Overfeeding (Reduced Food Accessibility)

As mentioned, overfeeding needs to be avoided. The machine needs to make sure that the pet does not have access to more than one serving at a time. This can be done be convering the uneaten portion or by disposing of the uneaten portion --All before feeding again. This also helps to make sure that the pet has access only to fresh food.

• Covering:
  The ability to cover old potions before making a new portion available would logically suggest compartments. Compartments could be rotated between, clicked between, or opened and closed.
  Clicking between compartments would be the most ideal.
• Disposal:
  There are a few ways to take the uneaten potions of food out of reach
• Trap door to a waste bin below (would make using a scale nearly impossible)
  PIC
• Sweeping to a waste receptacle to the side
  PIC
• In order to dispose of the food in this way, a logical statement would be needed before the next feeding, outlined here:
• If not empty
     Dispose then refill
  Elseif empty
     Refill
  End
• The terms empty, not empty, full and refill would be of replaced by terms needed to use whichever actions (outlined above) chosen.

Week 1: Feeding Intervals

The only factor that determines the constraints of the proposed apparatus is the consumer's needs. It needed to be decided for whom exactly this apparatus would be designed for.
Would it be for the busy pet-owner that needed help with daily feedings? Or the frequent traveler who would need a week's worth of feedings and daily appropriation?
Furthermore, would this consumer be a cat-owner or a dog-owner? Maybe even a rodent or bird-owner?

So, how much to pets eat anyway?
Pet	Age/Size	Amount/meal	Meals/day
Kittens	2 months	3 Tbsp.	four
	3 months	3-5 Tbsp.	four
	4-5 months	0.25 Cups	four
	6-12 months	0.25 Cups	three
Cats	5-9 lbs.	.25-.5 Cups	two
	10-14 lbs.	.5-1 Cups	two
Puppies	Varies with breed. Can be read on back of food container
Dogs	3-20 lbs	.75-1.5 Cups	two
	21-50 lbs	2.3-3.25 Cups
	51-100 lbs	3.5-5 Cups
	100+ lbs	5-7.5 Cups
Rodents/Birds	All day, every day

Dials could potentially be used to customize feeding intervals. One dial could determine the amount of food per feeding (in ounces or cups) while another determines the time interval of feeding (in hours).

Programming to intake button responses is also an options.

Now, the machine would need to be able to tell how much food is in the bowl --in other words, when to stop feeding. This can be completed in a number of ways:

• Scale: Using weight measurements as input, the machine would know how much kibble would for example equal one ounce and would multiple that amount by the preference of the owner.  Heavy programming would be required.
• The owner would need to be able to rare the scale as necessary
• The owner may also be able to use a set and rare button to input desired weights, in lieu of a dial





Possible design of a infrared gates.
Cross-sectional view.
• Infrared: Using pairs of gates on either side of the bowl at certain marks (say 1/4cup, 1/2cup, 1cup & 2 cups), the interruption of a gate would mean that the mark has been met. This option would require light programming and could be completed in two ways:
• Programming based on consumer input via the dial would determine which gate was used.
• Each pair of gates could have its own circuit completed by the dial (optimal).
• Measured output: This option is purely mechanical and is ideal for this group's skill sets.
• This option has a multitude of execution possibilities. For example, a small scooper or fan (for a lack of a better word) could revolve allowing only a certain amount of kibble to pass per revolution. The kibble output (in ounces) per revolution could be measured and used as the factor which determined the number or revolutions needed to reach the consumer input (via dial).



Depiction of a a trap door

• This would be completed in a similar manner as option 2 of infrared option: mechanically, with different sized gears.
• Regardless of the complexity of this machine's user input, the output (the kibble) should trickle out rather than be dumped or fall.
• This is to keep the feeding area clear of stray kibble and to ensure that the amount put out is as close to the user input as possible

Another important feature would be a bell that would indicate to the owner that the feeder needed to be refilled.

Week 1: Brainstorming

It was decided that an automatic pet feeder would be made. Many idea were explored. It was important that the group improved upon products that were already on the market.

Some pet feeders drop the food from above, others poured it down a slot, and some just let gravity take over.

Crestuff Automatic Portion Conrol
$40 on Amazon

Replenish Pet Feeder
$16 on Pet Solutions

Looking at the currently available automatic pet-feeders, it was noticed that none of the products available on the market made provisions for waste. What would happen if the pet were to not finish his food in time for the next portion of kibbles? The pet would have access to way more than he needs daily.

That scenario should be avoided...

for obvious reasons.

The idea is the owner will not be there to supervise.

So, this feeder needs to 

1. Dole out precise amounts of food on a tailored interval
2. Dispose of uneaten food after a certain amount of time, before the next feeding

More on how that might be made possible, later.