Roast Control

[farmroast]

The more control I've added to my roaster the more I'm thinking that the word "control" and coffee roasting really don't go together. I do believe that the abilities to adjust and monitor I've added have improved my roasts but by no means will ever allow me to totally control my roast results and repeatability. The extremely complex makeup of the coffee bean has much to do with the problem. The slightest changes of just a couple degrees and/or seconds at many moments in the roast can result in rather profound changes. I've often wondered what a perfectionist pro roasters frank opinion on this would be. I had a bean from Costa Rica that in one roast gave me a distinctly wonderful chocolate covered cherries flavor that I just could never quite repeat. I still would encourage all the added roaster improvements possible but have lowered my expectations in terms of roast control perfection.
farm

[another_jim]

I think there might be a two part answer here; at least in my experience.

1. Fine tuned control like the programmable ramp soak controllers I use are mainly a convenience feature.

2. Changes that fall somewhere between control and design -- for instance, on my air roasting: going to the P1 with it's "pizza stone" ceramic buffer between the heat and the beans, splitting off the heat and adding the ability to dial it down, insulating the roast chamber to lower supply and environmental temperatures, detuning the PID so it doesn't cycle the supply temperatures up and down -- all these changes made more of a positive difference than anything I get from varying the exact profile I chose.

The problem with these changes is that they are bound to the technology one uses. The changes I describe all seem to work as advertised for most air roasters and maybe also for high flow convection drums, but they seem far less useful for unventilated drums or heat gun roasting.

The big design variable seems to be heat flux. The rate at which heat enters the bean is entirely dependent on airflow. For an unventilated drum or a heat gun roast, it is minute. And with these roasting techniques, high environmental temperatures and longer roast times have no deleterious effect one can notice. When the heat flux gets high, as in convection roasters, it becomes a lot easier to not just to control the roast, but also to screw it up.

Spikes or sharp drops in environmental temperatures that would hardly affect beans wearing an anorak of still air in an old unventilated drum will kill them when they are running around naked in a high tech convection or air roaster.

[farmroast]

Jim
Nice summation. My next tests are going to be on the conduction side for comparison. Up til now my convection heater with adjustable temp. and air flow along with bean bat speed has been my main focus with my lower adjustable conduction element being used for supplement and balance. Next I want to try a slightly less charge load with the conduction heater used more as the primary source with the upper element, fan and bean bat speed used to aide transitions. The questions then become at what points does high flux with lower relative ET vs lower flux with higher ET or raising or lowering flux without changing ET become beneficial? And can a mix of the 2 be worked into a single roast with improved results and stability? The other thought I've had is adding an adjustable nano carbon fiber element vertically in the middle of my roast pot to add IR and radiant heating to the mix. I don't have the situation of held heat in a heavy drum that air flow seems aide in an improved drum roaster.
Ed

[another_jim]

You're going to end up with the ultimate test bed roaster; so you can invite food science grad students over to run the experiments for you

IR is used in the mass market world to quickly change a medium convection oven roast to a dark roast. Probat implies that this produces a higher quality dark roast or espresso roast, but they only say so in the advertising for the roasters that have the feature. I've never read anything about it otherwise.

[farmroast]

I played with a Behmor for a bit. I find some merit in the IR/radiant heat and as you mention might be a nice way to quickly roast the surface a bit for a darker roast edge with a lighter inner.
The thoughts of a control board for this thing and my mind turns to mush. Jim, do you have a link with pics of your present roaster?
Ed

[another_jim]

My controllers are very primitive by the standards I see here. I use an inverted loaf pan as the heat sink for the SSR, and mount the project box above it. The controller is "plug and play," with a regular 110 volt plug and receptacle to bring in the power and plug in the heater. My last effort, built for a friend, uses the 1/32 din Fuji ramp-soak PID controller for the supply air, and a $30 Chinese Ebay PID just for its alarm function and relay output, to shut down the roast via the SSR's switching side when the target bean temp is reached. The two thermocouples are plugged into the box using the standard Omega connectors.

I'm a complete klutz when it comes to building these things, so they are very amateurish and I dislike posting pictures.

[Martin]

The more control I've added to my roaster the more I'm thinking that the word "control" and coffee roasting really don't go together.

Well, it they don't "go together," at least they travel in the same circles.

I don't post much anymore about heatgun roasting because I've already said about all that I can on the matter (something that I haven't said is that the 5th year of HG/DB was better than the fourth).

That doesn't mean discussions like this one have no relevance to me; they do, and they continue to inform my roasting, if only in the most diffuse and circular ways. IMO, there are four domains that enter into the home roaster's quest: The discussion here is mostly about thermometric and environmental control. Surprisingly, this may be the easiest domain to achieve and master. There is even some transference to someone like myself who, on the surface, has least temp and environment control.

Tied to this is repeatability. Actually, "anticipated results" is a term I'd prefer. It's the ability to create a roast that matches what the roaster has in mind--could be that the roaster wants to repeat ----that is, to repeat a previous roast, but more likely he wants to imagine to a high degree what would happen if, based on prior knowledge, he tried something new----a blend, a bean, a profile, and so forth. Anticipating results is a characteristic of artisans and craftmen in many fields. (I recommend the book, The Craftsman by Richard Sennett, for an interesting discussion.)

And this brings up the third domain----small production, or at least, capacity of the home roaster. There are just so many trials we can run----just so much coffee we or our friends can drink----too many crops and varieties to try. Temp and environment are useful to control, precisely because they are controlable, and the other domains are less so.

Finally, there's the biggie. Cupping experience and memory. You have to be able to know what you are tasting. It's the ultimate "control." Fortunately, highly experienced cuppers can, to some degree, transfer this roasting domain to others of us. It's what happens when an esteemed commercial roaster establishes a benchmark roast that is commented on by posters here.