Bolivia Cup of Excellence: San Ignacio Juana Mamani Huanca.
Not my best effort at the roast. Rushed the warm, and pushed 1rst crack up about 2+ minutes. Shit. And it shows in the cup. Green and tight on the tongue, florals and fruits aren’t well developed. It still has some real nice sweetness that comes out as it cools, but it should be more delicate, balanced than this.
Well, still tweaking the roaster, but damn, wish i’d done this with a coffee that I didn’t have so little of, or that cost so much ($25/lb). Oh well, at least it was only 125g. And I still some.
But, on the bright side of things, I fiddled with the air/fuel ratio on the pinhalense tonight and have a much better flame now, one that has a slight hiss and a good tightly formed flame - not the laziness that it was before. Seems it needed more air (O2). I hope that this means that I’ll be able to mve at least a full load through the times I’m expecting… till now, 300g has been somewhere about a slow 30min. ugh.
I found where the chaff has been going… just under the cooling vents, and before the entry to the impeller.. that is the hole in on the left side of the floor.”
I can hardly stand the slowness that designing and building this roaster is. I know that it is my doing, but that doesn’t make it any easier not being able to really control my roasts. And it’s becoming increasingly painful as I become ever more eager to sink deep into cupping coffee and reaching balance with a roast. I’m not currently exploring as I know I am able… and truly return to my craft.
Well, In fact I couldn’t let this go any longer, and so purchased a sample roaster from Pinhalense.. I just got word at the end of last week that it is essentially ready in the factory as of tomorrow, so all it needs is some crating, an airplane, one long flight from Sao Paulo to Portland. Some gentle customs agents, and a van to get it home. If all goes well, by next weekend, I should be tearing through some junky green that I’ve been stashing away over the past couple months to get it nice and seasoned.
I was trying to edit a batch of images to Flickr with their batch tool, but the commenting was duplicating anything that I wrote (several times), so I gave up for now and simplified the set I was putting together. I am intending to post all my older designs for the drum—sections, elevations as well as these perspectives—but that will have to wait…
I decided not to go with this design pretty much because I changed my mind about the importance of air control with this batch size. I had convinced myself that at this batch size—80g to 500g—that air flow was less important. but while walking over to a friends home for dinner on night a while back, I re-thought my decision. So I am shifting over to a closed system so that I can adjust air flow much like larger roasters. I am not sure if (or how) I will keep the rest of this design. The dimensions of the chamber itself still feels correct to me, and the vane size and placements as well… I had considered possibly moving to something that agitates the beans in a way to keep them moving off the hot surface of the drum. I am leaning though, toward keeping this setup since I like how it maintains a relatively consistent mass - which my theory (somewhat supported by a scientific article (which I will post a link soon, promise) that the coffee will roast more evenly when it does this. Too much agitation will cause it to roast less evenly. so without much adoo, here are the drawings.
(BTW, notice the new fancy image viewer I found! It’s Pictobrowser a flash plugin written by a couple NYC interaction designers that pulls flickr sets into your site—all contained in this neat little package. The wordpress plugin was written by Kumara Sastry. I altered it a bit to get the original size images to load and so you could zoom them and read the notes… I like it a whole lot. It’s nice)
UPDATE_ I finally got all my drawings from sketchup posted and annotated - well, okay some of the annotations are cut and pasted from other annotations, but whatever… - so they are all in the set now. And if you care, now have a flickr account just for Deft.. you can see what’s there, uhh, here.
Okay, three last images. They simple show the difference in air flow for low and high air flow. Noticeably more chaff gets around the corner with high air flow.Here they are:
Low air flow:
High air flow:
Comparison of high and low air flow - the magenta overlay shows the difference when the high airflow was done (second):
I think that I have a bit more figured out and will post when I have a chance to scan a couple drawings… But I think that what I’ve landed on is a relatively simple air path that draws air via two different paths—one entering through the bean chute, through the drum and out the back toward the chaff collection box; the other entering through the bean tray (cooler) and hooking up with the chaff collection box, where both paths meet and out the door…
What is nice about this is I can essentially seal off the chamber where the heat source is located and around the exterior of the drum so that I can make this temperature stable and so controllable and responsive.
I am figuring out the design of the air flow—mostly to allow for control of the air flow, but will need to catch the chaff as well. I based what I am doing on what I know of older Gothot sample roaster —why reinvent this? I don’t have one to play with, so am reverse engineering based on my memory and images I have found on the internet.This video is of the easy portion—where the air exits the roaster and enters / moves through a wind tunnel of sorts, separating the chaff from the airflow. you can see where the chaff is caught in the dead zone below the hole it enters through. Still a couple refinements to figure out, but it is working.The Cheesecloth was there cause it is a normal fan and not a n impeller, and I wasn’t sure it would work (well, it didn’t at first, which is why it is version 1.3 -)
