The current iteration of the Coins-Worth device version 2 is now about 50% together. Here’s a sneak peak at the end of week iteration 7/3/14.
Taken from above, there are two tall walls that support the smaller functional components. I refer to the left half of this piece as the “stacking station”, since this is where unprocessed coins are stacked. There is a smooth bottom plate, a tube support just above the plate, and sandwiched in between is a coin ejector. A portion of the coin ejector protrudes through the visible slot in the bottom plate, and is actuated (pushed back and forth) by a small servo just underneath the station.
The right half of this piece is the “scanner station”, since this is where the camera takes a picture of the coin. (Technically, not “scanning”, but the name has stuck, so I use it!) Note the coin sits at a 25 degree angle, and the leading edge rests on a small U-shaped piece I call the coin catch. The dark object on the right is a small Logitech USB camera, centered 2 inches above and at right angles to the coin. Not visible is a small hole just under the leading edge of the coin, in which a small servo-actuated rod is used to quickly flip the coin. (Coins don’t always fall heads-up, do they?) This allows the camera to take pictures of the front and back of the coin. After the pictures have been taken, the coin catch is raised slightly, and the 25 degree angle of the scanner station allows the coin to be gravity fed to the sorting station which will start just underneath the scanner station.
The design for the sorting station is nearly done, and I’m just now putting the design into FreeCAD. The sorting station will have a chute which will direct the coins to the appropriate bin, which as a minimum includes:
- A bin for collectible coins
- A bin for nominal-valued coins
- A bin for coins that were not recognized by the image analysis, which can include foreign coins, coins of other denominations (the initial version will analyze pennies, nickels, dimes, and quarters), and coins that the image analysis software just couldn’t recognize. (It only needs to discern denomination, year, and mint; no attempt here at grading the coins.)
I’ll have options for multiple bins, and may try to design coin roll tube holders for the individual denominations.
Folks I’ve talked to have expressed concern for possibly causing marks on the processed coins. Everything that touches the coins will be plastic, and the only real danger is in dropping the coin into the collectible bin. Keep in mind this version of the Coins-Worth device is intended to process pocket change and coins collected in piggy banks, water jugs, etc., which will likely not be uncirculated or mint quality. I do have a design in mind for a second model of Coins-Worth designed to more carefully handle coins individually on fabric surfaces for consistent photography and further analysis. But that’s later!
There are 4 parts not 3D printable: the camera, the acrylic tube, the servo (not visible, but mounted on the wall underneath the stacking station), and of course, the coins!
And just for comparison purposes, here’s a shot of Coins-Worth Version 1, made from plywood, acrylic, and metal. The coin flipper on the scanner station really does work! The sorting station is underneath the two upper boxes, and swings an arm over one of two bins. The Netduino is visible on the right, and ugly servo wiring and other small components are mounted all over.
Yay for 3D printing!
Great stuff Ray! It looks like you have a good handle on making things from your computer. I visited the Lulzbot web site and saw that it can use nylon. Have you tried this material? For my application, this would be a strong solution. I still need to investigate FreeCad. Good job, and you’re having fun too.