For the umpteenth time, someone has asked for profiling advice; and for the umpteenth time, the more experienced home roasters have given the conventional wisdom: that there are some profiling generalities, but that exact profiles are roaster and coffee specific.

Can we do better? Can we put together a good set of profiling information?

This is a brainstorming session, so all ideas from people at all experience levels are welcome.

I'll kick the session off; and you can all tell me how full of it I am.

Seems to me that each roaster (machine) needs a basic profile, the one to use on a new coffee, and the one to tweak to get the right roast for a known coffee. So maybe the article should focus on how to find that basic profile, and how to tweak it. We can do this in general, "this is how you do it for all roasters" and people can share their results for specific roasters, particularly the off the shelf ones.

For instance:

My general advice is that the profile has
-- a drying phase to 300F bean temperature (BT),
-- a first ramp to the first crack at around 390 BT,
-- and a finishing ramp to the end of the roast.

A basic profile would tell people how long each of these phases should be for their roaster. I would advise people to set up their basic profile so it produces a really good light roast, pulled just after the first crack ends. The acid test is to get some green and roasted (for brewing) beans from Terroir (the best light roasters in the country) and see how your equally light roast stacks up.
-- It should have no grassy flavors (too short a drying phase), or dusty, leathery, or choking ones (too long a drying phase),
-- it should not taste bready or like cheerios (too long a first ramp), or have no toasty flavors at all (too short a first ramp),
-- it should not taste overly acidic (too short a finishing ramp), nor should it taste flat and caramelly (too long a finishing ramp)

I would advise people to keep the length of all three roast phases mostly the same for any degree of roast (light or dark), but to add a minute or two to the finishing ramp when going from brewing to espresso roasts, since you want the flavor balance to have more caramel and less acids.

Here's how to get some very basic profiling on a completely unmodified popcorn popper:

Drying phase: Run the roaster one minute, then turn it off for two minutes, then restart the roast. Increase or decrease the turnoff time to change your drying time.
First ramp: There is not much that can be done here
Finishing ramp: Use a lower weight of beans to lengthen the time, a higher weight to shorten the time.

I dunno, Jim. I live in L.A. fairly close to sea level (under 500' above) with desert-like humidity levels (frequently single digit during dry periods and it rarely gets about 60% except for the infrequent occasions when it's raining). How helpful is a profile that I've developed (to the extent that you can vary profiles on the Behmor) going to be for somebody who lives at altitude or in a humid place? Unless we're both storing our greens in a climate controlled environment (which is never likely to be a luxury I can afford), we're going to have very different starting points, so running the same profile would likely end up with very different results, even if we're both using the same roaster.

Lots of folks around here have been roasting longer than I have, but after 6 years at it and well in excess of 1000 roasts, I'm not so sure that there's a 'magic formula' of precise steps to follow that will work for everyone.

edit: Although I'm more than willing to be convinced otherwise.

another_jim wrote:For the umpteenth time, someone has asked for profiling advice

Who's that noob ? Assuming he can join this session, here are his 0.02 USD worth of thoughts:

GO JIM! Seriously, though, all I can think of for this to work is an article per, available & profile-able, home roaster. On that roaster, 3-5 different "classes" of beans are profiled. With such an article written as a thread, we, the masses, reply with our own beans/results to reconfirm/clarify to other readers.

This is a brainstorming(i.e. dreaming) session, right?

George, no magic formula indeed. But espresso extraction suffers from a similar problem; but we have 9bar, 25secs, brew temp, etc as guidelines that made decent home espresso possible. Cuz of the involvement of roasting, go one more level specific, and BINGO! now you have the equivalent of headspace, dose, steaming style, etc recommendation. The good thing with roasting is that there are very few different accessible roasters when compared to the number of espresso machines.

It's all about starting somewhere reasonable while having access to good advice...

another_jim wrote: ... Terroir (the best light roasters in the country) ...

Agree with this praise of Terroir, but would say that PTs and Barefoot are also among the top light roasters. Barefoot, I think, has been actually the lightest of the three, then PTs, then Terroir.

Novo could cram into the phone booth too.

I've been pondering this exact question for a couple of months now, and have come to the following conclusions:

*crickets chirping*

The best that I can offer so far is to learn how to accurately measure temperatures for your roasting method. The foundation for profiling is accurate and CONSISTENT temperature measurements. I believe that this holds more weight than sight or smell, as it is more objective and easily measurable.

Off the top of my head, I believe that JimG has hit the nail on the head for dogbowl/heatgun (DB/HG) measurements. IMAWriter might want to chime in for the Stir Crazy/Turbo Oven crowd, and
another_jim and others who are old pros at Poppery roasting can provide information regarding probe placement for the air roasters.

Know your roaster and which temperature measurement is applicable. From what I've read, there are 3 basic measurements that have been shown to be of significance - environment temperature, bean mass temperature, and exhaust temperature. The basis for profiling as it relates to your roaster is to know which of these three can be: A) most accurately measured, B) most easily repeatable, and C) whose change has a significant effect on the final product.

More rambling to come,
-s.

Don't understate the importance of the envirornmental temp. Below freezing, i.e. all winter long I have to move from the garage into the house under the hood which is over the stove. and adjust for temps in 30's degF.

Marvin

It refers to the temperature inside the roaster, the beans' environment, not the roaster's.

Very interesting thread. I think that Marvin was stating that ambient temp (outside roaster) is important as well. Possibly humidity too. I noticed earlier this year that every time I went to roast it seemed to be raining. When I was roasting in poppers, I was huddled under the garage door trying to keep dry and keep the smoke under control. Now that I have a Behmor, I'm roasting on my bench and more comfortable. I never did correlate any effect of rain or humidity on my roasts, though.

I moved from poppers to a Behmor this year and have a thermocouple taped to the ceiling of the roaster (to the right of the light, angled down slightly and slightly behind the drum) and I've been logging my roasts with BehmorThing to try and temperature surf the Behmor beyond the built-in profiles. This is an ET probe, I have no BT probe.

I think that the conventional wisdom with the Behmor is that for most beans you start with a known load size and then run a P1 roast and note the time to first crack so that you can run the next roast on P2 or another profile with a temperature cut at or prior to first crack to slow down the bean momentum and extend the break between first and second crack. In addition, you can do things like open the door for 5 seconds every 15 seconds between first and second crack to try and get more time in this phase and open the door after a couple of minutes of cool cycle to cool the beans a little faster. (my shop-vac has a carpet attachment the width of the door that I use once the ET has dropped below 350 F or so to accomplish this while keeping the chaff in the roaster or vacuum)

My ET probe has shown me that there's more going on in the Behmor's programming then the charts indicate. There is a significant (40 degrees Fahrenheit or more) temperature drop programmed into the heating elements right before the afterburner is programmed to kick in, and this temperature drop occurs at different times based on the 1/4#, 1/2# or 1# weight settings on the roaster. This temperature drop occurs whether or not your afterburner is working.

The temperature drop occurs at about 3:30 (three minutes and thirty seconds) for the 1/4# setting, 5:00 for the 1/2# setting, and 7:30 for the 1# setting. You can see and hear this back-off on the heating elements, and should be able to see a change in the current/power draw on your electrical circuit as well (I monitor voltage but don't have a kill-a-watt to monitor current draw).

I have been trying to use these timings and bean load size to allow a nice ramp for the drying phase with the dip in temperature occurring at an opportune time before first crack to cut back on the bean's momentum allowing me to stretch out the time between first and second crack. I do a two minute pre-roast (run the Behmor for up to two minutes at P1 or P3 and hit stop then quickly enter the desired profile) to push the temperature drop closer to first crack, but you risk more smoke this way by delaying the afterburner kick-in (note, at some point the internal temperature of the Behmor will exceed a threshold that won't allow you to stop and then immediately re-start the Behmor, usually this is longer then two minutes, but for some cases it can be less then two minutes).

I've uploaded some roast charts showing these timings. These roasts have a two-minute pre-roast, but the temp drop will occur at the same time relative to the last time you hit start no matter if you pre-roast or not. The roast charts show different time scales which makes a side-by-side comparison more difficult, but they illustrate the point, I think.

1. A quarter pound roast of 4 oz of 2008 Brazil COE #1 "Faz Kaquend" on a 1/4# P1 A profile with a two minute pre-roast here.

2. A half pound roast of 7.4 oz of Sweet Maria's Mika Kadir Blend on a 1/2# P1 A profile with a two minute pre-roast here.

3. A full pound roast of 16 oz of Ethiopian Sidamo Organic Natural Process Korate on a 1# P1 B profile with a two minute pre-roast here.

All these roasts were logged with BehmorThing (manually) and I never left my seat in front of the Behmor or took my attention away from my roasts (no matter how much my Wife would try and distract me). If you pre-roast or use a 1# setting for a small load size, it's important to make sure that your prepared to attend the roast and be present for any emergent situations (chaff fires, etc.).

The attached graphs may not be my best examples of timing the temp drop to maximize the time between first and second crack, I posted them only to illustrate when the temperature drop occurs to make my point clear.

I'm thinking that this information could be helpful for others trying to maximize the time between first and second crack on a Behmor.



-Chris

My ET probe has shown me that there's more going on in the Behmor's programming then the charts indicate. There is a significant (40 degrees Fahrenheit or more) temperature drop programmed into the heating elements right before the afterburner is programmed to kick in, and this temperature drop occurs at different times based on the 1/4#, 1/2# or 1# weight settings on the roaster. This temperature drop occurs whether or not your afterburner is working.

Chris
I have minimal experience with the Behmor. I did some mods to one and then loaned it out. This is the thermostat kicking in at the MET for that stage of the preset. So you can have 2 things that can happen to change the temp., the thermostat and the chosen profile percentage change. A kill a watt meter set on watts (watts being used) will give you a good idea of what the roaster is doing and exactly when.

another_jim wrote:It refers to the temperature inside the roaster, the beans' environment, not the roaster's.

"It puts the lotion in the basket, or it gets the hose."

hehehehe....

For reference, here are some further thoughts (and better information, probably) from Ambex Roasters regarding probe placement. The information is specific to drum roasters, but the information can be extrapolated to your favorite roasting method.

I can comment on the bread machine side - I've tried several different methods, ranging from open-top with a cheap analog dial thermometer to drilling holes in various points for wild ideas about mounting bare wire K probes, and so far I'm getting the most consistent results by using a steel sheathed K probe punched into the side of the machine. I've posted some pics on Homeroasters.org, but if anyone's curious I can post them here as well.

-s.

yakster wrote: here.

These Et dips during the first crack can't be good. Is there any way of avoiding them?

These Et dips during the first crack can't be good. Is there any way of avoiding them?

With a not so sensitive thermostat, radiant heat, not much drum or chamber mass and no insulation, when the thermostat kicks off the temp. drops quickly.

http://www.bootcoffee.com/articles.html

These 'Ruling the Roast' articles by Boot really cover alot on the subject. He refers mostly to drum and sample roasters. I find most applies for my air roaster as well. (....I know some are looking for a 'process recipe' here, it is buried in the text)

* The time from start of first crack until end of roast must be at least 3 minutes. Espresso roasts should (almost...) always be longer. I have found this to be a brick wall rule. Still I see alot of 'slow' home roasts that are too fast on the finish. This time is critical for aroma and flavor development.

* For my roaster, a reasonable range of ramp from 300F (drying) to first crack is 20-33F/min. (3-5 minutes) The final few degrees approach can be slower. This is the range that tends to produce even browning and good bean expansion (internal pressure). Softer beans slow, harder beans can run faster. Air roasters tend to run faster than drum roasters here.

* Drying times can be most variable depending on preheat and roaster heat transfer characteristics. I tend to ignore(subtract) time the BT is under 200F when looking at profiles. The BT response when the roaster is charged with the bean mass will give a signature of the roaster's heat transfer characteristics. Air roasters tend to run faster than drum here.

* I'm with Sherman on the temp probes; you need to know and control the temperature to profile roast.
The bean temp probe is essential, MET or Environment Temp is important as well; A measure of the heat applied to the bean mass. After running the popper with a MET probe for a few months, I no longer can work without it. It is a consistent method of controlling the heater power and a reliable predictor on where the bean probe will be in the near future. It has become more important than BT for controlling the heater and fan.

I've learned a lot about roasting in the 6 or 7 years I've been home roasting, especially in the last 5 or so years when I've had a real one pound commercial sample roaster. One thing I've learned during this process is that whatever you learned on one type of roaster probably has little ability to be transferred to another one. Another thing I've learned is that unless you have the funds to purchase a "dream" commercial roaster, it is unlikely that you will be able to obtain consistent and excellent results with anything that a home roaster is likely to purchase, unless whatever it is that you purchased has been substantially modified.

In order to teach yourself how to roast, you are going to have to expend considerable energy, considerable time, and at least some money. This is not to say that you can't produce acceptable results without a lot of energy or time, however if you want to achieve results that come close to what a really good commercial roaster can achieve, then time and effort and expense are all necessary. If all you want to do is to produce decent results that are better than what most people can easily buy locally, most anyone can learn how to do that with a popcorn popper or a very cheap air roaster.

Let's assume for the moment that you want to produce really good to outstanding results. If that is the case, there is no roasting device out there for less than a few thousand dollars that will produce excellent results without at least some modifications and effort on your part. The modifications and efforts you will need to expend will differ by the roaster, although no matter what you have it will be necessary to get feedback from people whose taste you respect, or you will be left with your own judgment which can be limiting (unless you are really truly experienced with evaluating coffees, in which case you would not even be reading this post or this thread in the first place).

Herewith, below, is a list of things I have learned about roasting that might be useful and perhaps transferable across various roasting devices.

(1) You can't shine sh*t. By this I mean, garbage in, garbage out. If you don't buy good beans you will not get good results, period.

(2) You need to be able to monitor your roast temperatures in a repeatable way, with the monitoring giving you some sort of information that you can act upon. Knowing what is going on with your temperature monitoring device when you have no way of modifying said temperature will be of no benefit, so in addition to having a way to measure temperature, you will need to have a way to modify heat input so that what knowledge you are gaining of the "on the fly" temperatures can be effected by something you can do with the heat source or ventilation or both.

(3) Don't introduce beans into a roasting apparatus that is already too hot or you may damage your beans before you have even roasted them. I can't be any more specific than this if my suggestion is to be useful for a wide variety of different roasting devices.

(4) A drying phase is necessary, but if it is dragged out too long you will end up with a dull lifeless "baked" roast. For me, this means about 3 minutes initially at a lower heat input until about 300F is reached on my bean probe, however I would hesitate to recommend any particular length of time or temperature that would work across multiple roasting devices.

(5) Getting to the onset of first crack after the drying phase must be done in a "reasonable" period of time. If you go too slow, you will "bake" your beans, but if you go too fast you will burn them.

(6) the time interval between the onset of first crack and the end of the roast is really really really important. I try to get around 4.5 to 5 minutes in during this period, during my (typically) 14-15 minute roast batches, which seldom go as far as to initiate 2nd crack. Jim Schulman and I did a very interesting little experiment a couple of years ago with some beans I'd intentionally roasted to have an interval of either 2.5 or 4+ minutes in between the onset of 1st and the end of the roast, with the same beans and taken to the same final temperature. Unlike most simultaneous blind tasting experiments we have done, the espresso shots made from the coffee with the shorter interval were GROSSLY inferior to those which had the longer interval. On every single blinded pair of shots we compared, it was very very easy to pick a "winner" and a "loser;" the loser in every instance was the coffee that had only 2.5 minutes of interval in between the onset of 1st and the end of the roast. We terminated the experiment after only about 5 shot pairs each, since we could distinguish them every time and the differences were so obvious.

(7) Rapid cooling after the roast is a good thing. My roaster cools the beans down to close to room temperature after 2 minutes outside of the roaster. I don't know what the "magic" number is as far as how long it can take you to cool down the beans, but longer is clearly not better.

(8) Whatever you do and however you do it, we are talking about one process, from bean selection to roasting to drink production. I like to make single varietal espresso shots using what is regarded by many here as a low PF dose of coffee, e.g. around 14g. Everything I do from start to finish is done with this in mind. If on the other hand I wanted to make shots from blends using 20g in the double basket, my bean selection would be different as would the way that I roast. Likewise, if I buy blends from our friends the marquee roasters, excellent and well regarded pros, and if I take their blends designed to be dosed at 22g and instead I dose them at 14g, the espresso shots taste like dishwater. Everything is connected. You have to decide what it is that you want to produce, in the cup, before you can really choose the coffees you want to roast and how you want to roast them.

ken

another_jim wrote: the more experienced home roasters have given the conventional wisdom: that there are some profiling generalities, but that exact profiles are roaster and coffee specific.

Can we do better? Can we put together a good set of profiling information?

This is a brainstorming session, so all ideas from people at all experience levels are welcome.

My first thoughts were quickly followed by a flood of variables. Didn't seem to matter if I thought across the range of roasters or specific. So started to look from the other direction.
The better I can detect roast specific problems in the cup and then either know what I need to adjust or be able to communicate what I'm finding and to what degree would be how I would be then able to determine my sort of base profile. Now that I kinda know how a roast can be screwed up I've thought of intentionally accenting some specific mistakes in separate batches to get a clearer sense of what effects they have in the taste.
Another way being able to examine the roast. Looking for divots, cracks,scorching, tipping, color, oils,surface textures,cracking a bean open, chewing on one,etc.

Last week in Santa Fe, I went to see a evening of though-provoking conversation sponsored by the Santa Fe Institute. The two speakers were Nobel Laureate Gerald Edelman and Marvin Minsky, founder of the MIT Media Lab, EE, and AI theorist. The conversation was about brain, mind, computers and the internet. One of the things that Minsky said that really resonated with me, and I think applies very well here was this: "An expert is someone who knows all the horrible mistakes that can be made." As with many things, I think that, in roasting, knowing what not to do is probably the most important aspect of the learning process. I know that I've learned much more from my mistakes than from my successes, as they lead me further into exploring the hows and whys behind the process.

It strikes me that a roast profile could be well-described (using Jim's three segments) by the BT's at the segment boundaries, and the rate-of-change of BT within the segments. For instance, the most common profile I use on my Hottop roaster could be described this way:

• Drying phase: TP to 300F; 20F per minute
• Ramp to 1st: 300F to 390F; 20F per minute
• Finishing ramp: 390F to end of roast; 10F per minute

Jim

farmroast wrote:The better I can detect roast specific problems in the cup and then either know what I need to adjust or be able to communicate what I'm finding and to what degree would be how I would be then able to determine my sort of base profile. Now that I kinda know how a roast can be screwed up I've thought of intentionally accenting some specific mistakes in separate batches to get a clearer sense of what effects they have in the taste.
Another way being able to examine the roast. Looking for divots, cracks,scorching, tipping, color, oils,surface textures,cracking a bean open, chewing on one,etc.

Ed, would you (or others) mind expanding on this?

What are some of the specific errors that your familiar with, and what changes have been effective for you in compensating for those errors?

Noah brought up a thread about differences in bean gloss, and there were some useful nuggets provided by another_jim:

another_jim wrote:The luster of the beans just after the roast is based on how much oil is on the surface (matte, eggshell, semi-gloss, high-gloss, oildrops ).

When you roast to the second crack, the beans will be high gloss inside the roaster, but will reabsorb the oils and be matte as they cool. Then, depending on exactly how far you've gone, the oils can reappear after a few days.

The association of oils and second crack is not graven in stone. In a slow roast, the 2nd starts at a lower temperature than for a fast roast, and fewer oils appear.

These two factors, roast speed and reabsorption could explain your gloss difference.

For roast quality, it's better to examine the beans for wrinkles and cracks, rather than sheen. A really good roast to the first pops of the second should give you botox smooth beans.

and from farm_roast:

farmroast wrote:If your flatter/oilier roast might have been slower in the start it would seem that more oil could be drawn to the surface later in the roast and if there was a little less moisture available might have somewhat limited some flavor development

-s.