Bean trap partial resolved, advice on final step

[runnerbrain]

I used to think that the trapped beans are the ones that get diverged into the exhaust pipe as I charge the drum, but now I think these are the beans which, as they lose water mass,  get sucked into the exhaust pipe as I ramp up air flow towards the end of the roast. I am sure a lot of you have come to the same conclusion.

[edtbjon]

I used to think that the trapped beans are the ones that get diverged into the exhaust pipe as I charge the drum, but now I think these are the beans which, as they lose water mass,  get sucked into the exhaust pipe as I ramp up air flow towards the end of the roast. I am sure a lot of you have come to the same conclusion.

You're very close to the "very probable" source of the problem. Hankua commented earlier somewhere about the construction of the front plate and the exhaust pipes. In essence he said that the construction is a way to keeps costs down. Mandrill bent pipes are cheap to produce, but they are in this case also part of the problem with stuck beans. In my opinion, the quite narrow front plate to exhaust connection creates a kind of "venturi", where the air speed increases in that first 80deg exhaust bend (at the front plate) with the effect of sucking beans into that part of the exhaust.
For starters, I always puff down the exhaust directly after charging my beans. That clears the exhaust at that point of the roast which is why I usually don't have any green or light yellow beans in my roast as I have the fan set to almost zero.
Apart from that, puffing down the exhaust a few times more during the roast, especially in the later stages, will reduce the number of beans getting sucked into the exhaust. That "puffing" is much easier to do with the T setup, but it can of course be done with the J too. (One could discuss the matter of blowing cold air into the drum during the Maillard and development stages, but that's another discussion.)
The key to getting very few beans in the exhaust is control of airflow and to not overwork the fan. It's no secret that the Huky fans are very powerful and I don't think all that power is needed. I personally almost never go above 50-60% on my Variac. I'm thinking of making some tests later with various fan settings, to see if and how it affects the taste of the coffee.

[MJ5150]

I don't have a Variac, never have.
Every so often I'll look down the exhaust pipe during a roast and see a few beans piled up at the bottom. I blow into the pipe and push them back down.
Any beans I find up in the 'T' are there from when I charged the drum. I have a process to clear those out right after I charge.

-Mike

[coffee_cat]

Using a Variac. End of 1c it's not uncommon to crank up the fan to 75%. Haven't experienced this issue.

Early on blew down the intake sometimes to dislodge beans stuck at the bend. After a couple of hot kisses from the sweet Huky, gave that up. Put some replacement leg screws in and biased the Huky to lean back maybe 2% to 3%. At the outset of the roast, give a little tilt back to dislodge any stragglers.

Another potential source of hangups, as you probably know, is the trier. Need to religiously verify that it's in the correct position throughout the roast, then, at drop, to withdraw and insert it a few times to knock any hangers-on out and into the drum where they're expelled.

Roasts are beautifully even with the odd quaker being the only sign of any unevenness, that attributable to the beans themselves, I believe, not the roaster.