I started with about 100 grains in the trickler, as Paul suggests in his setup and use videos. By the time I stopped testing last night there was probably only 20-30 grains left in the cup and it was still having occasional issues.
Did some more experimenting last night and found there are 2 different binding cases.
Case 1 is a "soft" bind where the wheel will still rotate but becomes difficult to rotate, and the added friction usually causes the drive wheels to slip. From what I can see this happens when a kernel or kernels orient themselves long-wise while sitting against the back side of a notch as the wheel rotates (i.e. the long axis of the kernel ends up facing the center of rotation of the wheel); in this case the kernel wedges itself between the bottom of the notch in the wheel and the outer circumference of the cup. The more kernels that do this at any given time, the greater the friction on the wheel. Part of this is because N133 is a very small/short extruded powder. A notched wheel with shallower notches than the two wheels that are provided with the trickler would probably keep the kernels from turning sideways and wedging and prevent this issue, but there are no extra notched wheels available yet. Just for fun I tried using the ball powder disc without the wiper/deflector installed (if the wiper/deflector was installed it would hit the kernels and jam the disc) and while the ball powder disc never had the friction issue like the notched discs, eventually the small N133 kernels clogged about 3/4 of the holes in the ball powder disc and the trickling rate became very slow.
Case 2 is a "hard" bind where the wheel seemingly hits a hard stop and won't rotate at all. I tracked this down to kernels sticking to the notches and not dropping down the output chute. When this happens the kernel gets pinched between the notch on the wheel and the side of the housing over the output chute and binds the disc. See the picture below, the red arrow points to the stuck kernel that stopped the disc.
View attachment 7823201
Note that this did happen with the wiper/deflector installed to the correct "notched" position with the ball that helps to knock powder down into the output chute. This issue seems to have been solved by washing the wheel in solvent again-- this time using brake cleaner (much more aggressive than the ether I used the first time), then using a toothbrush soaked in brake cleaner to scrub each individual notch, then spraying it off again with brake cleaner. This seems to have eliminated the kernels occasionally sticking to the wheel causing the "hard" binding/stop-- at least for the tests yesterday.
I still really like the trickler and feel it's a big improvement in precision over the standard rotary trickler without sacrificing speed-- but it is certainly more sensitive to powder kernel size than the simple rotary trickler. Additional notched wheel configurations are probably going to be needed to handle all powders without malfunctions. I'd like to try a wheel with notches about 1/2 the depth of the included 2 wheels to see if that helps with the "soft" binding/drag issue that I'm currently experiencing with N133.
I also need to try it with some other very small kernel extruded powders like AR Comp or H322 and see if they have similar issues to N133.