We have built 6 FEDs using the new STL files and done some quantitative error testing on them.  Specifically, we dispensed 200 pellets from each manually and recorded 4 types of errors.  These included:

- Pellet "jumps", where the dispensed pellet jumped out of the pellet well, leading to a 2nd pellet dispensed.  n=28/1200 dispenses, 2.3%

- Double pellets, where the dispensed pellet was not detected, resulting in a 2nd pellet dispensed, n=23/1200 dispenses, 1.9%

-Jams (self-cleared), where the FED took longer than 10 tries to dispense, activating its self-clearing motor movements, n=9, 0.8%

-Jams (not cleared), where the FED required experimenter intervention to clear, this did not occur in these tests, n=0!  

Our next tests will involve mice in cages, stay tuned!