28BYJ-48 driven Fish Feeder

thingiverse

**Update: 6/16/18** I just got a fresh set of 28BYJ-48 steppers, suspecting my old ones were weaker than normal. The new ones seem to have more torque and are resulting in fewer jams. Running them at 9v instead of 5v also helped eliminate jamming even more. But what made the biggest difference was adding a mount for our tank. Our cheap 30gallon tank has a really flimsy plastic LED hood with a 62mm x 23mm opening for feeding, so I made a little mount that replaces the end cap on the feeder and fits in there holding the feeder at a slight angle. That angle seems to have solved most of the jamming issues. In fact, I switched back to 5v and while it did jam once going through an entire container of food, it cleared itself the next cycle. I've run a full container of food through it 2-3 times now, and it only jammed up 2 or 3 times and cleared itself the next cycle on all of those. I'm still running my test script that runs a dispensing cycle every 2 seconds - and the cycle first starts by running the auger in reverse a few hundred steps to help clear jams before running it forwards. I've added the tank_mount.stl file, and since this is most likely to need to be modified, I also exported a .STEP file for it. (Though again, I've included the link to the onshape project which is public so anyone can access the full project directly.) **Original Summary Follows** This "Works" but is still a work in progress as it jams up too frequently for me to trust it right now. A few years ago my daughter won a few fish at the county fair, unlike most fair fish we actually put them in a decent tank and cared well for them so they've thrived and are now large and healthy. But I'm tired of having to find someone to feed them every time we want to go out of town for a few days, and I'm too cheap to spend $15 on a feeder at the store when I should have enough junk laying around to just build one. I Initially tried this simple feeder: https://www.thingiverse.com/thing:2539750 but found it tended to jam up on the pellets our fish prefer and just wasn't reliable enough. So, I decided to redesign it with a few key changes. Since I have a bunch of 608 bearings on hand, I decided to redesign it so the auger would be supported at both ends. At this point, I basically started over on the design and only took inspiration from the designs I listed this as a derivative of. In addition to being supported on both ends with bearings, the auger now has a thicker center for extra strength, and a fillet where the screw meets the shaft to help prevent food pellets from jamming up. Since the bearing blocks the end, the tube now has an opening for food to fall out of on the side. The hopper opening is shorter and filleted so it can print without supports. The screw holes for mounting the motor are sized so you can either self-tap M3 screws or use an M4 tap. The hopper is a tight press fit - you may have to clean up the opening for it in the tube just slightly - printing the tube with no supports I had a tiny bit of dropped extrusion that I had to cut off at the top of the opening. The lid for the hopper is also a tight press fit - but if you don't push it on all the way, it's still fairly easy to remove. The end cap is a little looser but is really just decorative. Neither cap is necessary, just kind of nice to have. It now works much more reliably - but I'm still fine tuning the hopper inlet because the auger will sometimes still get jammed when pellets feed in. I've tried tweaking my code to run the auger in reverse for half a rotation before dispensing which helped...but still hasn't completely prevented jams with the size of pellets we use. Flakes seem to dispense fine - but our fish don't like them and flakes jam up in the hopper. You could add a stirring system to the hopper like some other designs have to use this with flakes - but since we don't use flakes, I haven't pursued that.