ethorson wrote:Here is a link to the SM roasting guide. I am definitely going to slow things down on my next roast.
https://legacy.sweetmarias.com/library/ ... gree-roast

If you read what is written at the beginning of the "pictorial guide" this is what is written:

I was doing a faster roast in this case anyway, a bad roast profile overall (don't emulate it), of an unintentional green coffee blend we had made here a while back. Ignore the times, and take the temperatures as a ballpark figure. The important thing is here is to see the transformation the coffee goes through as it roasts and what look, color, bean size and surface texture, corresponds to the degree of roast.

He (I presume Tom) says the profile is "fast," and "a bad roast profile overall (don't emulate it)."

Also, the time interval between the BEGINNING of 1st crack and the beginning of 2nd crack is 11:50 minus 9:20, or 2 and a half minutes. This is not very different than what you can "accomplish" with a drum when the roast "gets away from you." Most drums unless they have a huge heat source in relationship to the bean load, would be hard pressed to compress the cracks much less than this. They simply don't have the heat capacity to go any faster.

But again, this set of pictures was made to show what beans look like at various stages of the roasting process, not to give a profile that anyone should try to copy.

One thing I did not emphasize in my post above was that in order to expand the interval between the onset of 1st crack and the end of the roast, one will usually have to try to reduce heat input before first crack starts, so as to avoid entering first crack with too much heat. If you do go into first crack with "too much momentum," e.g. too much heat, this can result in the roast "getting away from you," and once you have lost control of a roast at this phase of the roast, the die is probably cast and your roast product will be suboptimal.

ken

Ken Fox wrote:One thing I did not emphasize in my post above was that in order to expand the interval between the onset of 1st crack and the end of the roast, one will usually have to try to reduce heat input before first crack starts, so as to avoid entering first crack with too much heat. If you do go into first crack with "too much momentum," e.g. too much heat, this can result in the roast "getting away from you," and once you have lost control of a roast at this phase of the roast, the die is probably cast and your roast product will be suboptimal.

That is one bit of very interesting information. I guess a quick pulse (or increase) of the fan would do the trick.

Will you keep on editing the first post to add these tidbit? This could become very tedious for you but very informative for us sucking the knowledge (and life) out of you.

One thing that your text does not touch is the magnitude of the temperature. Is there a specific reason, such as it will vary from set-up to set-up (which I am sure it will). But are there ballpark figures to start experimenting around? You can guess that I am planning on incorporating Environment Temperature (ET) and Bean Temperature (BT) to my set-up.

Thanks again to Ken and the other contributors. I am now running when I was crawling before in the learning process.

Whale wrote:That is one bit of very interesting information. I guess a quick pulse (or increase) of the fan would do the trick.

That would depend on the design and how you use your particular roaster.

Whale wrote:Will you keep on editing the first post to add these tidbit? This could become very tedious for you but very informative for us sucking the knowledge (and life) out of you.

I have added a couple of sentences to the original post.

Whale wrote:One thing that your text does not touch is the magnitude of the temperature. Is there a specific reason, such as it will vary from set-up to set-up (which I am sure it will). But are there ballpark figures to start experimenting around? You can guess that I am planning on incorporating Environment Temperature (ET) and Bean Temperature (BT) to my set-up.

This is intentional. My purpose in writing this post was to try to help people to think while they roast and to be able to reason out, for themselves, how to continually improve as home roasters. Any roaster specific suggestions would be confusing for those not owning the type of roaster being addressed. As to specific roast temperatures that occur at different points during the roast, these are enormously variable depending on the type of roaster and where and how the roast temperature is being monitored. I have personally roasted on machines that have had the onset of First Crack begin anywhere from 375F to 425F. That's a 50 degree spread! Neither temperature is wrong, they are merely different measurements taken in different places in different pieces of equipment. What's important is consistency, and being able to correlate a given temperature that you can measure, with what is going on, in your roaster, with your beans.

ken

Ken Fox wrote:My purpose in writing this post was to try to help people to think while they roast and to be able to reason out, for themselves, how to continually improve as home roasters.

Oh! I am not trying to get a formula or recipe to follow (never followed a recipe in my life and I am not about to start now!). Even if you gave me one I probably would not follow it...

Ken Fox wrote:I have personally roasted on machines that have had the onset of First Crack begin anywhere from 375F to 425F. That's a 50 degree spread! Neither temperature is wrong, they are merely different measurements taken in different places in different pieces of equipment.

I hear your answer but can't help to think that there has to be an "actual" temperature range at which the coffee starts first crack and other milestones. Not the temperature that I will, or anyone will measure with their own set-up, but a scientific value. Like water boils at about 100°C. I have not come across that value just yet.

Or I should say that I have come across to many different temperatures to make me think that the value given are just measured not "actual". I have seen people posting a wide range of measured values but as you say it really depends on what and how you measure.

I have ordered a book on roasting (did not receive it yet) maybe it will be in there. Or is it that the range of varying coffees and conditions (humidity, process and such...) will indeed have that much of an effect on the "actual" temperature for each milestone?

The true temperature of the second crack is variable, since it's the breakdown of cellulose which depends on time and temperature (longer roasts have lower 2nd crack temperatures). But the literature gives the first crack temperature as pretty definitive, 395f on the outside, and maybe 380 on the bean inside. This is fixed, because the first crack is the temperature at which the remaining water turns to steam under the roughly 30 bar pressures that develop inside the coffee cells.

However, this is moot, since the temperature that is being measured is an air temperature that is influenced by the beans, but also by the hotter heat source of the roaster, and the cooler temperature outside the roaster. The goal is consistent readings, not scientifically accurate ones.

It would be nice if there was a method for calibrating and converting the logs captured by different roasters to something close to true bean temperatures, so they could be compared. The first crack is a good fixed point, but a second point is also required, and I can't think of a good one. So far, the only truly communicable constants are the length of roast and the shape* of the profile (more convex for fast start/slow finish air roasts, almost straight for slow start/fast finish drum roasts).

*The slope of a curve changes when you stretch or move axes, but it's basic shape doesn't, so for gauging the shape, it is doesn't matter what precise bean temperature you measure

Can the end of drying stage be used as a reference point? When the beans turn yellow, it may be helpful to record the tempeature, because it is from this point on the "ramp up" starts. Of course, this assumes the beans can be visualized with a sight glass or trier.

another_jim wrote:The true temperature of the second crack is variable, since it's the breakdown of cellulose which depends on time and temperature (longer roasts have lower 2nd crack temperatures). But the literature gives the first crack temperature as pretty definitive, 395f on the outside, and maybe 380 on the bean inside.

Thank you Jim. That is what I was looking for.

another_jim wrote:However, this is moot,...

It may very well be moot to you and most people, but not to me. Remember, I am not trying to apply a recipe or reproduce someone else measurement. I am trying to learn and understand the coffee roasting process. Which is what I believe Ken is trying to achieve, thus my request/comment.

another_jim wrote:So far, the only truly communicable constants are the length of roast and the shape* of the profile (more convex for fast start/slow finish air roasts, almost straight for slow start/fast finish drum roasts).

*The slope of a curve changes when you stretch or move axes, but it's basic shape doesn't, so for gauging the shape, it is doesn't matter what precise bean temperature you measure

That is very interesting information. From a theoretical point of view again, is there a taste/effect relationship that can be attributed to these profiles? Assuming that both profiles occur within the same time period and the curve between the first and second cracks is the same?

The question about the best profile doesn't really have a meaningful general answer.

If you look at the specifications of all the commercial roasters that sell well and are respected, you see roast times of 8 to 12 minutes for systems that use a lot of airflow (air and convection roasters) and 10 to 15 minutes for drum roasters which use lower airflow. Drum roasters that do 20 minute roasts like the Diedrichs, or air roasters that do 6 minute roasts like the Sivetz have been less succcessful commercially, and are usually modified by their owners to get into the more normal time limits. Therefore, I would say these times represent the consensus of specialty coffee roasters on optimum roast times

Experienced home roasters with roasters in this style* also tend to modify their gear to achieve roasts in this time frame. So I would say the commercial consensus holds for the analogous home air and drum roasters.

This means your profiling range will be a four to five minute window covering both the depth of roast and the timing of the various roast stages. Within this window, you need to work by taste -- roast, taste, tweak, repeat. Professional roasters use cupping and cupping roasts to minimize the amount of trial and error, but all the good ones test their production roasts and tweak the profile after every roast. It's a lot like adjusting a grinder. You decide on an ball park dose and shot time after an initial tasting, and then you tweak after each shot.

One of Ken's big points is that this "theory of roasting" takes all of 5 minutes to learn; while the "local knowledge" of roasting consistently enough on your own roaster to do this tweaking takes months to years to learn. By the time you've gotten to that point, your tasting and roast diagnostic skills will be good enough to know what to tweak and how to interpret other people's reports.

* Home roasting devices are a lot more varied than commercial ones. Devices like heat gun/dog bowl combos or very solid BBQ drums can have extremely low airflow and much longer roasting times. I've had great coffee from people using these methods, so the timing consensus derived from the commercial roasting methods doesn't apply to them.

another_jim wrote:The question about the best profile doesn't really have a meaningful general answer.

Where and by whom was this question asked in this thread? This is not what I asked so I am not sure that your post is directed at me at all. AFAIC, there is no "best" anything in coffee.

Just to reiterate my query, is there a flavor/effect that can be attributed to a particular profile (temperature vs time curve?

Convex curve tend to make the coffee taste something like...
Concave-straight curve make coffee taste more ...
Flat/straight curve tend to produce.... flavors.
Flat-convex curve will ....
Concave-convex (S) curve ...

All assuming the same coffee beans and time period. And if you will, let assume a 10 minutes period to first crack for the sake of this discussion. Is there such heuristic that can be written down.

If there is little appreciable difference between the pre-first crack profile (shape of the curve) I would tend to think that to achieve Ken's recommendation to slow down the roast prior to first crack than a straight-convex curve would be the easiest to "program" (if someone has such luxury).

Do not get me wrong. I will try all the above profile with varying time period for myself to find which I like most, if I taste a difference at all. That is, I will try to the limit of the frankenroaster that I will end up building.

chang00 wrote:Can the end of drying stage be used as a reference point? When the beans turn yellow, it may be helpful to record the tempeature, because it is from this point on the "ramp up" starts. Of course, this assumes the beans can be visualized with a sight glass or trier.

Roasters behave differently. My 1lb sample roaster, for example, does not have a clear cut period when the beans "turn yellow." In my roaster, the beans don't all start to look the same until just before first crack starts.

I've tried to stay away from anything that is roaster specific in this post. I also want to avoid using jargon that may confuse people, giving them the impression that they are "missing something" that other people are seeing. The concept of a "drying phase," for example, is more a mental construct than it is a real, discrete, period during the roast that you can time precisely.

As people become more experienced they can explore more of the physical and chemical underpinnings of what goes on during a roast. At the beginning what they need most is to learn to use their basic observational skills and to correlate them with what is going on in the beans as they roast, and how this effects their results.

ken