Dear all,

This is my fist post on home-barista (big fan!!). Before this I was an information leecher. Many thanks for that! I though it was time to contribute. For a year now I have been home-roasting with my Genecafe. I'm very happy with it. This year I decided to put some systematic in my information gathering. I started recording temperature, time and cracking sounds. My first goal is to get an answer to 3 questions:

- Does pre-heating the roaster influence the roasting process, judged by the moment in time of the 1e and 2e crack?
- Can I influence the heating ramp, by setting the temp. to max.
- Is the cooling process of the Genecafe sufficient.

The first question seems answered: see my blog post, http://koffietent.wordpress.com/2010/01/25/science-of-roasting-home-roasting-that-is/

The second question seems also answered: see my blog post,
http://koffietent.wordpress.com/2010/01/30/temperature-profiling-with-the-genecafe-sort-of/

The third question is more difficult for me. I used to start the cooling process at the start of the second crack. I supposed the beans would surf right towards the end of the crack.
My last roast I kept on roasting for a minute (after the first audible snaps of the second crack). This made a big difference (from full city to Vienna - Light French). This means the final result is very critical to the starting of the cooling process.
I know about the 'stopping the machine and using air to cool' method. But I don't like having to handle the hot drum.

I would like to have a more quantifiable way off getting to the desired roasting stage.
Do anyone of you have an idea on how to answer this.

What do you think of my roasting curves in general? Also feel free to comment on my blog. I just started the blog and want to learn so.....

Edwin

ps: I cross-posted some of this info on coffeegeek. Sorry for that.

Hi Edwin,

Welcome to HB.

From your profile graphics, it looks like the Gene operates more like an air roaster than a drum roaster. In which case preheating may be pointless.

The very convex profile is also typical of air roasters blowing the same temperature air throughout the roast. Tom's tip of starting with a hotter setting and turning it down at the first crack exaggerates this effect even more. My advice is to stat with a low temperature (around 200C) and turn it up gradually to to max (260C) at the first crack, then leave it there. This will straighten out your profile.

Once you have that shape, you can reduce the weight to get faster roasts.

I do not know the Gene, so I do not know if the result will taste better, but if you want to explore new profiles, that is the way to go. I suspect, based on my experience profiling air roasters, that this will give you better results for lighter roasts, stopped before the 2nd crack starts.

To do this, try stopping a roast by smell. As the first crack winds down, sniff the vapors coming out of the roaster. There will still be grassy and vinegary smells. For the lightest possible brewing roast, stop the roast when the grassy, vinegary aromas end. For the lightest possible espresso, wait a little longer, until you start sniffing toasted almonds or marzipan.

another_jim wrote:Tom's tip of starting with a hotter setting and turning it down at the first crack exaggerates this effect even more. My advice is to stat with a low temperature (around 200C) and turn it up gradually to to max (260C) at the first crack, then leave it there. This will straighten out your profile.

Jim, I thought that to optimize the sweetness of the roast, the goal was to get to 1st crack quickly , and then slow the roast down to 2nd crack (to minimize the carmelization of sugars). So going 482F to FC, and then turning down to say 450F might be a good strategy. I'll let you know how it compares to Jim's profile (similar to what I had been doing) in a couple of days.

Hi Jim,

Thanks for the input!

another_jim wrote:From your profile graphics, it looks like the Gene operates more like an air roaster than a drum roaster. In which case preheating may be pointless.

It took me some time to understand the physics behind the GeneCafe. I think it is a combination of an air roaster and a drum roaster. It has a rotating drum.....with hot air passing trough it.



There are a couple issues I'd like to address at the moment.

First: The "strange" convex profile of the graph.



The temperature measured by the roaster it self (and displayed), is actually the air temperature at the exhaust of the roaster. When the roast is started at ambient temp the roaster itself has to heat up. The air at the input of the drum is hot, the roaster itself is cold. Energy which is put into heating the roaster can not be put into heating the bean. Thus pre-heating the roaster should help.

Second: What is the bean temperature?

The temp. displayed at the GeneCafe has definitely nothing to do with the bean temp. So I guess the best thing to do with this roaster is to take the internal and surface bean temperature form literature. Sure it's different for every coffee and it's not very accurate, but it is definitely more useful then the air temp at the exhaust of the roaster. Now all you need is the time the first crack starts and you can calculate an average bean temp slope.

To see if I could influence this slope I did an experiment. You can read about it by clicking here


Basically I did 2 roasts. One starting at ambient temp and setting the temp at 235 deg (455 F) and one pre-heating the roaster at 250 deg (482 F) and keeping it at this temp until the start of the first crack. Then lowering is to 235 deg (455 F).
In the first roast the first crack started at 11 min 30 s and in the second at 7 min.
I assume the internal bean temp at the start of the first crack is 185 deg (365 F) and surface temp 205 deg. For the first roast the internal bean temp slope is 16 deg/min (61 F/min), the surface bean temp slope is 18 deg/min (64 F/min).



For the second roast the internal bean temp slope was 26 deg/min (79 F/min), the surface bean temp slope was 29 deg/min (85 F/min).



I know this is not very accurate information and it is an average temp slope, but I think it is the best you can do with an unmodified GeneCafe.

Question for me now is: What is better, the 'slow' slope, or the 'fast' slope. The answer to this is of course that I have to taste the difference. I tried a cupping session last weekend, but that was also my first and very difficult to taste the difference (I think because of inexperience). Maybe you have some advice on the temperature slope?

I don't think I am ready to stop a roast by smell. Sure I smell the smells. But it is still a bit difficult to distinguish them.

Again I appreciate any input. I consider my self as a roasting rookie and want to learn!!

Regards

Edwin

I've never used the Gene, and I hope somebody who has can chime in.

One point. Often, when roasting outdoors in the winter, the roasts stall. I can't tell, but yours look like they flattened at after the first crack. You will still get a second crack when a roast stalls with a bean temperature above 210C, but the taste will be odd, with lowered acidity but not much roast development. The easiest fix is to put a cardboard box around the roaster so it warms up its intake air.

The easiest method for procedural questions (is this profile better than that one? -- is this Brazil better than that one, etc.) is to triangle test first. Three unmarked cups, one from one option, the other two from the second option. If you cannot reliably pick out the odd cup; changing procedures is not too consequential. However, its important to realize that a lot of indistinguishable changes can eventually add up to one very distinct one; so it's always useful comparing the way you do things now to the way you did them a year or two back.

Jim,

I see your point on outdoors roasting. Thanks for the tip. I appreciate every input, even more knowing that you have no experience on the GeneCafe.

In general what do you think of the ramp to the first crack. Suppose the slow one is 16 deg/min (61 F/min) and the fastest I can produce is 26 deg/min (79 F/min). Both are for the bean surface temp.

Edwin

Caveat, I'm assuming you graphs show some approximation of bean temperature. It's fine to use a process variable peculiar to your roaster for controlling your roasts, but if you want to communicate with others, you need to use a measure that is true of all roasters, i.e. bean temperature. This can be gotten either by direct logging, or by an estimate using a conversion equation of whatever you are measuring.

It seems you are going too fast in the drying phase, to 150C, where you want to spend 3 to 4 minutes. You are slightly slow on the ramp from 150C to 200C, since this should take about 3 to 4 minutes as well. Your graph from 200C to roast finish looks completely stalled, usually around 3 to 4 minutes is what you want for espresso.

To some extent, the very convex air roast profile, with a fast start and slow finish, compensates for the lack of drying, but going 15 minute roasts on an air roaster is a recipe for a dead tasting roast.

The roaster is clearly underpowered for outdoor roasting in the winter, as are almost all roasters, and all the talk about profiles is moot until you can finish roasts in 8 to 9 minutes using full power. Profiling a roast, by definition, means running at something less than full power. If your full power roasts are too slow, you cannot profile. Do some roasts with a box, and then see if you have enough power in reserve.

So if I understand correctly:

First drying: bean temp from ambient to 150 C in 3 to 4 min. So about 37 C /min.
Then to first crack (+/- 200 C) also in 3 to 4 min. that is 14 C /min.

Or do you mean:

Bring the bean temp to 150 C. let them stay there (drying) for 3 to 4 min
Then to first crack in 3 to 4 min

The bean temp everybody is talking about is the bean surface temp? Or the internal bean temp?

None of these questions matter until you can do eight to nine minute light roasts. You need to power up your roasts. Insulate the roaster, crank up the power, and see if you can get to the second crack in ten minutes or less. Once you have that, you can start thinking about profiles.

With a year or so of experimentation behind me, I agree that preheating is necessary, and that keeping the power up is essential to getting a good lighter roast (I aim for just before SC in about 12 minutes).

It is actually possible to get to SC in less than 10 minutes -- I just did it. Leaving the roaster on full power after a 15 min 250C preheat got to FC in 7:35 and first snaps of SC at 9:45 where I stopped it. Admittedly not winter -- ambient is 20C at the moment.

A couple of observation. Edwin, I notice you did a couple of roasts at 100g. At one point I did some tests varying weight and found good behaviour in the 220 - 250 g range, reasonable down to 200g, but less than that even at full power the roast would be flat, like it had stalled. I suspect heat loss through the drum. I might be wrong as my tests weren't exhaustive but I suggest you cross check.

As a means of getting some idea of bean temp I log time and sensor reading at important colour changes -- yellow, tan, light brown and brown. With preheating and full power, yellow usually comes at around four minutes.