Still frustrated with your Hextraction trap tiles not performing as expected? These trap tiles have been optimized for absolute maximum consistency. I have tested these against trap tiles from other packs and they perform more consistently at slow and fast speeds than any other I've tested. Probably more work could be done to improve these designs but these are the best you'll find right now.

These tiles were optimized for 10mm balls. They still work with 3/8" balls though.

This pack includes three tiles. Engineering notes on all of them are included here in case you're curious why these tiles are an improvement over other packs. I am not a mechanical engineer; I just did what made intuitive sense to me and iterated on the designs until they stopped failing. Each tile took about four iterations to get right. I printed these in PETG with 15% infill, 0.2mm layer height, and 3 wall loops.

Double Trap Pro: The tricky thing about this tile design is making it work in both orientations. (From here on out I will call the orientation that comes in the side and out the bottom the 'horizontal' orientation and the other the 'vertical' orientation.) The base tiles by Zack Freedman have severe issues in the vertical orientation with a second ball getting hung up on the first. This happens because trapped ball is at too shallow of an angle against the passthrough ball to give it much of a horizontal push and the geometry at the upper corner resists the ball's sideward motion into the exit. If the ball comes in with enough speed it might force its way past the upper corner, but not always. 3Dthee released a fixed version of this tile which shifted the trap slot to the side to make angle between the two balls less shallow. This makes issue less prevalent but doesn't eliminate it entirely. Additionally, by shifting the trap slot, it makes the horizontal orientation less consistent. If a ball comes in at slow speed, it won't be able to jump the exit channel and will fail to become trapped.

My design improves on both of these by widening the trap slot and angling it so that the trapped ball sits differently depending on the orientation of the tile. In the vertical orientation, the trapped ball sits deeper in the trap slot so that when the passthrough ball comes in, the two balls will have an extreme enough angle so that the passthrough ball will be pushed to the side out the exit. In the horizontal orientation, the shallower position of the trapped ball allows optimizing the two criss-cross paths for allowing a ball to jump across even at low speed.

This tile can still fail very rarely. (About one in one hundred times.) In the vertical orientation, if a passthrough ball comes in at very high speed, it can sometimes bounce off the trapped ball at a weird angle and land in the other trap slot. But I have yet to see this tile fail a single time in the other ways I mentioned above.

Pinecone Trap: This was inspired by nostep's Snowflake Trap tile. I locked it to one orientation and removed the three lower trap slots in order to give the design more breathing room. So it has less synergy with the gear tile, but otherwise functions similarly.

The main problem I wanted to address was a passthrough ball getting hung up on the trapped ball in the top entrance. I solved this by flattening out the channel in that area so there is nothing for the passthrough ball to get stuck on. It invariably will get pushed to one side or the other with no resistance. I also raised the trap slot a little bit to prevent passthrough balls from jumping over the trapped ball at high speed.

The other problem this solves is a trapped ball coming in in the top at high speed sometimes pushing out one of the trapped balls on the side. Since the traps no longer overlap with each other this is no longer a problem.

Other notes about the design: I widened the side trap slots at the top since in earlier designs, balls coming at high speed could bounce off the upper corner of the trap slot and not go in. The depth of the side trap slots was iterated on a lot. Too deep and very slow balls (and especially two balls coming out of a reservoir tile) could get hung up on a trapped ball in the slot. Too shallow and very fast balls could do a very weird collision with the trapped ball and bounce up into the top trap slot!

I have not seen this tile fail a single time in its current iteration.

Triple Trap Pro: The triple trap has all the same problems I mentioned above with the double trap but widening the channel isn't an option. 3Dthee's fixed version fixes the passthrough ball getting hung up in the vertical orientation by flattening the geometry around the trap (the same fix I did for the upper channel on the Pinecone Trap) but then this causes serious problems for the horizontal orientation. If the ball doesn't come in at really fast speed, the trapped ball will fail to jump the exit channel and just exit out the bottom. I wasn't able to come up with a solution for this so I decided to fully redesign it using a design similar to the Pinecone Trap. The tile only works in two orientations rather than all six, so it loses synergy with the gear tile, but now it works every time.

For the top trap in the vertical orientation, I used the same design as with the Pinecone tile. The slot is widened to make sure the ball gets caught correctly and heightened so that a fast incoming ball doesn't jump over. For the side channels in the horizontal orientation, I added extra space at the top of the trap slot since this seems to stop balls from bouncing out if they come in too fast.

Pinecone Trap Alt: Same as the Pinecone Trap but with different exit pathing. The trap geometry is identical so this works just as consistently.