Card Feeder for a Trading Card Machine




Introduction: Card Feeder for a Trading Card Machine

Card Feeder for a Trading Card Machine

The Background

When I was young, I collected tons of trading cards, but for some years, the passion for collecting has been decreasing. In the meantime I have children and slowly but surely they also start to get interested in trading cards. So I'm back in the game. :)
Meanwhile, we have more than 10,000 cards. Most of them are from World of Warcraft, but we also have many others:

Some of our cards are sorted and protected in UltraPro 9-pocket pages and UltraPro Albums, but most of them "stored chaotically" in boxes. We tried to sort them by hand, but after a few days we gave up frustrated. Especially with these quantities, handling and administration is very laborious and time-consuming.

The Solution

A few months ago, I stumbled across an interesting article published in the MagPi Magazine Issue 71, Jul 2018. Michael Portera explained there how he built a Card Counter.

It was exactly the motivation and inspiration that I need. Which that said, I want to create a machine that can handle our trading card collection.

Some goals; Trading cards should automatically be

  • managed (Which cards do I have?, Which ones are missing?)
  • sorted (Block, Language, Set, Series, etc)
  • rated (How valuable are my cards?, How much money do I have to spend for a full set?)
  • traded (Buy and Sell)

Because of this ambitious goals, I decided to split the huge machine into 3 parts.

  1. Card Feeder - a machine which grab and transport a single card out of a card stack
  2. Card Scanner - a part where the cards will be analyzed
  3. Card Sorter - a machine which will sort and store the identified cards.

Let's do it!

In this instructable I will show you Part 1 - How to create the Card Feeder.

Step 1: Tools and Materials

Tools and Materials

I decided to create the Card Feeder out of cardboard. It's cheap and easy to handle.

Here's what I used:


    • Cutter
    • Cutting mat
    • Ruler
    • Rubber bands
    • Scraper
    • Wire cutter and stripper
    • Screwdrivers
    • Pencil, markers
    • Sanding paper
    • Files
    • Acrylic paint and brushes
    • Hot glue gun
    • water-containing adhesives or solvent-containing adhesives.
    • Soldering iron and tin solder
    • Table saw


    Step 2: Prototyping


    I would like to share with you my ideas, thought processes and decisions in the Prototyping and Design phase. In my opinion, this stage is the most fun part during a project. From nothing to a first idea.

    The Mechanism

    To understand the whole mechanism, I analyzed one of the best feeders that exists; a printer. It does exactly that what I need. I bought a cheap 8€ printer on eBay and disassembled it until I could take a closer look at the needed mechanism.
    I tested a lot:

    It worked perfect. To be honest I didn't expect that behavior, but even with the trading cards the mechanism was able to pull out only one card at a time.

    From my point of view the secret of the mechanism is:

    • an area which press the paper to the major roll
    • a second smaller roll which has direct contact to the major roll
    • the second and major roll are covered with rubber


    It was time to copy the printer functionality. I made the 1st tests with a piece of IKEA cardboard to get a feeling for the sizes. Very simple and reduced to the feeding mechanism.

    After that I created the 2nd version with a container for the cards out of MDF. Very ugly, but it's working. While testing, I realized that:

    • I need some ramp for the cards which came out of the Card Feeder.
    • It's not necessary to have a separate container for the cards.

    In the 3rd version I switched back to cardboard, which I felt in love during the creation of this prototype. With the right tools and the right handling the material is sooooo great.

    Step 3: Design


    With the experiences from the 2nd and 3rd version and some thoughts about symmetry and equality, I created the final and 4th prototype on the computer. I've designed everything in Fusion360/Inventor and used all the built in functions like:

    • Rendering
    • Animation
    • Drawing

    I took the important measurements from the prototype with a caliper gauge and transferred them into the 2D sketch. All other measurements I made with the help of dependencies in Inventor and calculations. That's the reason for the strange dimensions. All angles and many distances are the same. The size of a Trading Card 63 x 88 mm was also a big factor in the design process.

    The basis of all further planning was the outside [2]. All other parts are oriented to it. The inside [1] is mainly for stabilization. Front [4] and back [5] prevent the cards from falling out. The ramp [3] serves as a slide for ejected cards. The cards are inserted into the machine via the top/back and placed on the card tray [10].

    [x] the numbers are references to the drawing


    • Outside 210,79 x 185,32mm
    • Inside 188,74 x 185,32mm
    • Front 73 x 84mm
    • Back 79 x 185,32mm
    • Card tray 70 x 67mm
    • Ramp 67 x 73mm

    Step 4: Basic Shape

    Basic Shape

    (You can find all pdf-files in the last step!)

    The individual shapes I had created during the design phase were printed on A3 paper. I glued them to a piece of 3mm corrugated cardboard. It is very important to apply glue and paper (scraps) on both sides. If not, the cardboard will bend. To be safe I put some weight on it during the dry time.

    I also would say (not sure at the moment) that water-containing adhesives will cause more bending that solvent-containing adhesives. Do you have here some experience?

    For a nice and clean look it is also important to distribute the adhesive evenly. Otherwise, bubbles will appear later. I used a spatula/brush to spread the glue.

    I doubled the backside for stabilization, which means I cut out two times the 3mm shape and glued them together. I also took care of the different orientations of the cardboard. A horizontal oriented outside will later be glue to a vertical oriented inside.

    After the drying time I cut out the shapes with a cutter and a ruler on my cutting mat. I worked my way through the process in several steps. Not the whole material at once!

    Step 5: Stabilization


    After everything was cut out, I used my small table saw and cut some small strips out of a big piece of wood to stabilize the cardboard. These strips have to fit inside the cardboard, so they have to be very thin (around 2mm). It is also a good way to protect the corners and edges. As you can see in the pictures, I only stabilized the critical areas.

    I put them, with some glue, into the cardboard. After the glue was dry, I cut off the excessive wood with a wire cutter, so they are flush with the shape. I mounted everything together and hold it in place with a rubber band.

    For a better optical effect and additional stabilization, I filled the cardboard "holes" with some filler and sanded them. I also applied some black color to create a bit of contrast on the inside. (It is visible on a later picture)

    Step 6: Mechanical Parts

    Mechanical Parts

    Since I didn't want to disassemble the printer, I looked around for other components. I found what I was looking for in the Lego collection of my children.

    With Lego Technic components I copied the mechansim from inside the printer.

    I used a bunch of Lego Bushes and Lego Bushes 1/2 to hold each wheel or gear in place.

    The distance between the sides is determined by the width of the card tray. The 4 washers are spacers to achieve exactly this width with the axes.

    For the right pressure, between the wheels, I added two springs on the outside of the axis. Direct contact between the wheels is necessary.

    One secret and a very important part are the rubber bands on the main wheels. Without them it will not work very well. Each wheel has three rubber bands in an 120 degree angle.

    I made the card area flexible. It can be adjusted in height in relation to the major wheel. I did this with two M8 threaded rods and 4 M8 nuts.

    After that, I glued all parts together with some hot glue.

    Step 7: Electrical Parts

    Electrical Parts

    The whole Trading Card Machine will be managed by a Raspberry PI, even the Card Feeder.

    If you need some Raspberry PI basics => Raspberry PI Class

    A continuous rotation servo, which is connected to the Adafruit Servo Bonnet via a standard 3-pin female connector, will move the 40mm major wheels forward and so eject a card for the card stack. The Adafruit Servo Bonnet is mounted directly on the Raspberry PI.

    I glued a Lego 24 Tooth Gear to a servo arm and screwed everything onto the motor. The whole unit fits perfectly below the card tray. I mounted it with some hot glue. The Lego 24 Tooth Gear has contact with the Lego 40 Tooth Gear on the major axle.

    I soldered some wires to the Adafruit Micro Switch and "connected" it temporaly to the Raspberry PI. With the switch i can turn the servo On an Off.

    I connected an external 5V/10A power supply to the Adafruit Servo Bonnet terminal block to power the servo. At the moment the power supply is oversized, but it should be enough for the whole machine.

    Step 8: Coding


    If you need some Raspberry PI basics => Raspberry PI Class

    I used the Raspbian OS and the Adafruit Python PCA9685 lib to interact and control the Card Feeder.

    The Adafruit Servo Bonnet will "talk" via I2C to the Raspberry PI. Therefor I had to activate I2C in the Raspberry PI Kernel and install 2 packages.

    sudo apt-get install -y python-smbus
    sudo apt-get install -y i2c-tools

    My code is based on the Adafruit example. I used it out of the box, just implemented the Adafruit Micro Switch to turn the Card Feeder On and Off.


    I'm a coding noob, so there are better ways to write the code. It was good enough for my test, but I still have a lot to learn.

    # Simple demo of of the PCA9685 PWM servo/LED controller library.<br># This will move channel 0 from min to max position repeatedly.
    # Author: Tony DiCola
    # License: Public Domain
    from __future__ import division
    import time
    import RPi.GPIO as GPIO
    #import the PCA9685 module.
    import Adafruit_PCA9685
    GPIO.setup(18, GPIO.IN, pull_up_down = GPIO.PUD_DOWN)
    # Uncomment to enable debug output.
    #import logging
    # Initialise the PCA9685 using the default address (0x40).
    pwm = Adafruit_PCA9685.PCA9685()
    # Alternatively specify a different address and/or bus:
    #pwm = Adafruit_PCA9685.PCA9685(address=0x41, busnum=2)
    # Configure min and max servo pulse lengths
    servo_min = 385  # Min pulse length out of 4096
    servo_max = 407  # Max pulse length out of 4096
    # Helper function to make setting a servo pulse width simpler.
    def set_servo_pulse(channel, pulse):
        pulse_length = 1000000    # 1,000,000 us per second
        pulse_length //= 60       # 60 Hz
        print('{0}us per period'.format(pulse_length))
        pulse_length //= 4096     # 12 bits of resolution
        print('{0}us per bit'.format(pulse_length))
        pulse *= 1000
        pulse //= pulse_length
        pwm.set_pwm(channel, 0, pulse)
    # Set frequency to 60hz, good for servos.
    print('Moving servo on channel 0, press Ctrl-C to quit...')
    while True:
        if GPIO.input(18) == GPIO.HIGH:
            pwm.set_pwm(0, 0, servo_max)
            pwm.set_pwm(0, 0, servo_min)

    Step 9: The End

    The End

    That's it!

    I've created a video. It's all about the Card Feeder building process. I hope you like it.

    To be honest, I would have expected a good to mediocre result at the beginning. But it's perfect. It never happened to me that either none card or more cards were fed at once. I am very surprised and pleased how well the whole thing works.

    I will use it as a Card Feeder for my Trading Card Machine. However, the principle can also be applied to other areas. Card Dispenser, Card Dealer, etc. Please let me know if you build a dealer for card or party games, blackjack machines or something else. I would like to see your ideas.

    I would appreciate any criticism, comments or improvements. Whether in relation to the Card Feeder, Photos, Skills or Writing/Language.

    I will move on to the next part of the Trading Card Machine; The Card Sorter. In my next update I will show you how I (will) built it.

    If you do not want to wait until the next update, you can see some news on Instagram.

    Thank you for taking time to read about my project!

    Have a great time.

    Servus and cu next time!

    Step 10: Attachment

    Here you can find all shapes as pdf-files. They have to be printed in an DIN A3 format.
    If you need anything else, feel free to ask!

    First Time Author

    Participated in the
    First Time Author

    Be the First to Share


      • Cold Challenge

        Cold Challenge
      • Baking Contest

        Baking Contest
      • Make it Glow Contest

        Make it Glow Contest



      1 year ago

      I just want to say thank you. I’ve been looking for so long for a affordable sorter but I guess it’s just not a good business because all I found was one for 8000 and then 50 a month after that. So glad I found your article. I’m pretty new to raspberry pi and I was trying to get an idea of how much this will cost. I came up with ~213 pretty good price all things considered. The one question I have is a few parts accriss the 3 articles have the same name. (The power supply and and raspberry pi 3) am I supposed to order multiples of these? I don’t think so and will maybe have to take a closer look at the article to find out but I just wanted to say thanks again for your hard work it is very much appreciated


      Reply 1 year ago

      Hi, my prototype is made out of cardboard. So it's very very cheap. For a business it would not work. You need a more stable material. 3D printed or aluminim, maybe wood.


      1 year ago

      do you have the 3d model of your prototypes?


      Reply 1 year ago

      Hi, no sorry. I lost all data and files for this project.


      Question 1 year ago

      Do you happen to have documents with the actual dimensions of the the parts?

      I'd love to recreate CAD files for these so they can be 3d printed.


      Answer 1 year ago

      Hi, no sorry. I lost all my data and project files for this project. If I want to move on, I have to start at the beginning.


      2 years ago

      Hello love the project. by chance is there a way i could get the raw dwg files i am having trouble getting the pdfs and converted dxf to work with my laser?


      Reply 2 years ago

      Hi, thanks.
      Sadly I created this project on another PC which I don't have access anymore. So most of the original files are lost.

      But I found some files from a previous stage which I can send you, so please send me a message with your e-mail.


      Reply 2 years ago

      I have loved your project. I am trying to make a model but with 3d printing. Would you have the document to take the measures? I can't find several and it would be of great help.


      Reply 2 years ago

      Hi, no sorry. I lost a lot of data.
      You can use the attached pdf's to measure the size.
      Sorry about that.

      First off, thank you for this *instructable*, I have been, for the last year, proto-typing my own storage, bin and OCR/RML/LTSTM machine, and suffice it to say, you're brilliant. You took everything in my mind that has been twisted in to a knot, and made me think "KEEP IT SIMPLE!"

      I have built step 1, and there are some tricks I had to deploy in order to get it to be 1, tensioned correctly 2, angled correctly.

      Starting with the lego parts:
      1. "Axle 16" (16cm). This specific part is no longer in any new lego sets. So for those of you whom cannot wait for parts to be shipped from eBay or, don't worry! Although, Axle "15, 14, 13" do not exist, "Axle 12" does. It leaves you with 4cm less of an axle, but it works just fine, you have ~ 1/2cm overlap on each side to attach a spring and a 0.5 bushing.

      2. 40mm tire is actually referred to as "Solid Rubber - Offset Tread - 40mm Technix Tyre: Large" This is also an impossible wheel to find in any lego set on the shelves. I was lucky that I had one from a lego set from my childhood, but I only had 1 (one). This is the IDEAL tire. It has everything you need for the "drive" or "pull" tire. The offset tread allows for a rubber band (yes the rubber band is imperative) to groove nicely and not slip. These tires can be bought online, but, like I said, I built this over the course of a few days, and I had no patience to order anything online. Make sure you have at least one of these tires. This build can be accomplished with only two tires. (The idler tire, or small counter-clockwise turning tire, and the clockwise turning drive tire, or the 40mm (really 43mm) technix tire. (Last seen in a lego set in 2004).

      ***Also, the "Small" wheel is not 10mm in diameter. It's 20mm (I used 17mm, as the "Large Tyre" is 43mm). Or that's the wheel(s) I used. Maybe 10mm in radius is what was intended to be put down here?***

      AXLE 12 (12cm) can be purchased in the "Lego Movie Two: On the Scene" lego set. It has exactly 2 (two) and that's exactly as many as you need.

      3. 1-3mm square piece of rubber. This is what I used to act as a second tire. It encompasses the whole front wall and overhangs so to keep all other cards, but the one-at-a-time you desire, back.

      4. Cardboard vs. "Bass" Wood.
      Cardboard is 3mm thick, as is craft "Bass" wood. Not to be confused with BALSA wood. Bass wood is extremely sturdy, and allowed me to skip the steps of strengthening cardboard. Bass wood can be bought at any craft store, and is also 3mm thick. With a dremel, jigsaw, or even a hand-saw you can make the same shapes very easily, and you're left with a sturdy shape. BEWARE OF BUYING THE "Bass Wood KRAFT Glue." This is a very weak glue, and will not help you achieve the strength in the joints that you will want. Gorilla Glue is what I used to seal every joint. Before applying Gorilla Glue, I made sure to take a damp rag and soak the joints in water, then applying the Gorilla Glue. Because this glue is a semi-epoxy, it uses the damp wood as a transport device to seep into the pores of the bass wood. Based on some strength tests, the joints or "connections" are actually stronger than the Bass wood itself. I don't suggest using wood glue, as it's generally messy, hard to sand off, and does not soak in the pores as well, also it has a much longer drying time.

      If you do not have a micro-switch, i'll include a pin-out of how to turn a regular button in to a 5v gate that when pressed passes a pulse of 3.3v so to act in the same fashion.


      Thank you so much, this thing has now ran through 10,000 cards, without ever, once, pulling two cards. Also, I have used the same 400 cards, and there is no visible damage to any one of the cards, as the rubber band/ rubber patch/ technix tire are all very gentle with the card.

      THANK YOU!

      I'll post step 2/3 when I receive the parts.

      Thanks again.


      Reply 2 years ago

      Hey, thanks for the comment! You really helped me find some of these harder-to-find parts. I'm still having difficulties with the small wheel/tire, however. Did you use a 17.5mm diameter one? I've seen those around. Any chance you could point me in the right direction?


      Reply 3 years ago

      Also I have my servo on the outside. Know that the distance between the 24tooth and the 40tooth are exactly 4 lego holes. (4cm)


      Reply 3 years ago

      sorry for the late reply, but I'm very busy at the moment.
      Thanks for this awesome comment full of improvements, other ways of doing things and ideas.

      I'm really impressed to see and read about your Card Feeder version.
      Thank you very much. This makes me so happy.

      Regarding the Lego parts, as you already mentioned, some of them are from my childhood (now they belong to my kids). So it will be hard/impossible to find them in newer Lego set.
      Sorry if my wheel specification is not correct.
      This is the big wheel, this the small one.

      What will be your next steps?


      Reply 3 years ago

      Well... I was kind of hoping you were looking for a job! Or a collaboration?

      If no,

      Continue working with LTSTM to completely automate and objectify “Card Grading.”

      I love this machine, it’s simple it works, it’s gentle and can be made to be even more gentle, but what I need is something maybe only you could design. Your creativity and mechanical understanding and conceptualizing goes far beyond any colleague or mentor I have ever had You think in such a clever and concise way., very much unlike those. Including myself, I’ve interacted with.

      Out of respect for you, I am building the entire project. I have had steel 1/16cm disks machined, and am waiting on a few parts.

      Sincerely, thank you for this. It really felt like it cleared my clogged thought process out and made me start to delineate between what is needed and what is wanted.


      Reply 3 years ago

      thank you very much for your praise of my work.

      Collaborating sounds great, but there is a problem.
      I'm working in the background on another project. Complex and with a lot of new topics. Therefor it needs my full attention.
      It will last quite a while. (Maybe mid of 2020)

      But please, tell me what you are looking for? What you are planning?


      4 years ago

      I have been thinking of this project for two years now. I have about 1/2 million baseball cards, some worth in the thousands, but I could not get past my fear that running the cards through using any type of "rolling" mechanism would decrease the value of the card because of rubber trails appearing after the scan is done. My thoughts now is to use a rolling ramp like you have here but once the card reaches the rubber rollers it would use a vacuum mechanism to move the card into the photo arena. Once the card is safely in place there would be a top and bottom cam run with a RaspberryPi to get both sides at the same time and forwarded to my computer. Nevertheless you actually did what I have been thinking about. I just don't want the mental and emotional pain of destroying the value of a card. I buy and sell quite a bit of baseball cards, hence the 1/2 million cards I have. OK fine call me a hoarder. As of right now I only provide stock images that someone else took but offer the potential buyer an actual scan from my 3 in 1 printer upon request as baseball card collectors want to look closly at the corners, edges and stains on the front and/or back.

      Any thoughts on this comment? You did a nice job by the way for cards other than cards that are worth 1000 times more than Yoogio (sp?) cards. Did you know that porn stars also have trading cards? I didn't until a few weeks ago by accident. Some of those cards go for $300+ each. I found that fascinating. And NO I have zero interest in trying to make a buck or two on anything related to porn or anything I wouldn't want my kids to find as kids know where everything hidden in the house CAN be found. :-)


      Reply 4 years ago

      First, thanks for your comment.
      I really like to discuss different options and ideas around a project. Great!

      In the intro I wrote about my “huge” collection, but that word fit better to yours.
      0.5 million WOW.

      Yes, I understand your fears. It is a torture for the cards and it is hard to watch, because the bending is quite a lot. But… I can not see any effects afterwards on the cards. On the other hand I used only scrap cards at the moment. If I would have very valuable cards I can not say if I would torture them with my machine.
      It’s made for the bulk process.

      Back in the days, when the idea grew in my mind I thought about possibilities to transport a card. This vacuum option was also on my list. But after the first test this idea failed. You can try this as well.
      I used a suction cup gun from my kids. The dart has a suction cup. I pressed it onto a card a lift it up. It works perfect and without any mechanical issues for the card, but (and this was the reason why this idea failed) it never happened the only one single card was lifted up. There are some physics behind, maybe static charge or static friction. I don’t know. But on every test two or even more cards were lifted up. Sometimes it lasts only a few seconds and the other cards fall down, but quite often even after minutes they are still stucked together.
      Maybe your baseball cards don’t have these problems.
      Maybe this problem can be solved with some air flow that will blow off all the additional cards.
      But I don't focused on this idea anymore .

      Pornstars trading cards? Interesting!
      No, I never heard of that.

      Yes you're right. Kids will find everything. :)


      Reply 3 years ago

      What is happening is called the Bernoulli effect. Even though air is not actively being blown between the two cards, they have not discharged their imbalances (+\-) charge. The reason it’s inconsistent in timing of when the card below the one being pulled up is because the amount “static electricity” or in this case known as static cling. If attached to the inside of the suction cup was a small copper wire that was grounded, or had a minimal current, the discharge or charge, whichever would work, would be instant. Basically, some sort of conductive material needs to make contact with the card being pulled to free, or balance it’s electrical build up.

      Seems really strange that this would have such an affect on a piece of cardboard, but there is a very good reason felt is used in casino’s. Ground your lifting mechanism and watch magic happen!

      I’d love to see what you come up with using fluid dynamics!


      Reply 3 years ago

      Matter of fact, just run a current through whatever the cards are sitting on, if there is a current on the tray, it will align the entire stack, hence no negative pressure or increased velocity of the air between them.