Hmm, Mexican Altura sure roasts fast.

[Musicphan]

I would encourage you to try the A/B comparison. Claus's book is almost 20 years old and we have learned a TON about roasting since then. Back in those days, you were lucky to be able to source coffee from a specific region let alone a farm. Tom/Sweet Marias was the only place pushing boundaries. I only state this because in modern roasting we start at high temp and lower during the process (reverse of what you do... and reverse of what early Behmor roast profiles are as well). During the last 5-10 years we have discovered the time for the drying process (up to mailard / browning) has little impact on flavor (which is probably why the dehydration method works)... but the timing between drying and first crack has a huge impact on the body, and the time from FC to drop has a huge impact on acidity/fruit flavors coming through. Just food for thought... try diff methods and cup/taste to your preferences.

 

[addertooth]

Musicphan,

You gave me a lot to digest there.
Just two basic questions:

1. Is the goal to keep the time between Drying and first crack short, or long? (I am guessing short as possible).
2. Is the goal to keep the time between first crack to drop short, or long? (I am guessing stretched out).

Thank You for your input. Any results you get are only as good as the information (or suppositions) you start with.

Addertooth.

 

[Musicphan]

Musicphan,

You gave me a lot to digest there.
Just two basic questions:

1. Is the goal to keep the time between Drying and first crack short, or long? (I am guessing short as possible).
2. Is the goal to keep the time between first crack to drop short, or long? (I am guessing stretched out).

Thank You for your input. Any results you get are only as good as the information (or suppositions) you start with.

Addertooth.

It's a bit of a personal decision and what your taste buds dictate... there are no hard/fast rules IMO. However, we have discovered there are some generalizations or milestones people try to hit when roasting. Generally hit browning by 5 min (just when the beans start turning from light green to yellow/brown), another 5 minutes to FC, 1.5-2 minutes to finalize drop the temp. Keep in mind that for a fairly light roast just out of FC when finished. I take most of my coffees for drip into 2.5-3 minutes. Two books to look at when time/budget allows - Scott Rao's book on roasting and Rob Hoos's book on flavor development. Now forewarning, since you're not roasting on a traditional drum roaster some/most of the info won't apply but good technical info on the science of roasting.

 

[addertooth]

It's a bit of a personal decision and what your taste buds dictate... there are no hard/fast rules IMO. However, we have discovered there are some generalizations or milestones people try to hit when roasting. Generally hit browning by 5 min (just when the beans start turning from light green to yellow/brown), another 5 minutes to FC, 1.5-2 minutes to finalize drop the temp. Keep in mind that for a fairly light roast just out of FC when finished. I take most of my coffees for drip into 2.5-3 minutes. Two books to look at when time/budget allows - Scott Rao's book on roasting and Rob Hoos's book on flavor development. Now forewarning, since you're not roasting on a traditional drum roaster some/most of the info won't apply but good technical info on the science of roasting.

Thank You for those details. Now let me break out my slide-rule and figure out how I will approach this. I will probably do another Ethiopia Yirgacheffe Misty Mountain with a profile that matches what you describe. This way I can do a side-by-side with the two coffees in a couple days after resting.

I suspect this will also allow some of my peaberry roasts to be improved as well.

You are spot-on about how the literal temperatures from Drum Roasting won't directly apply, but they will at least let me know what kind of curve in bean temperature I should be looking for. I have been logging bean temperature change as shown by the optical pyrometer. The pyrometer has been showing 350f when the very first pop is heard on First crack... but (upon review of my notes) the pops aren't really rolling until the beans hit about 385... so maybe the pyrometer isn't as far off as I originally thought.

 

[addertooth]

MusicPhan,
Thank You for your advice, I ordered these three books:

Modulating the Flavor Profile of Coffee
by Rob Hoos

The Coffee Roaster's Companion
by Scott Rao

Coffee Roasting Best Practices
by Scott Rao

While doing some roasts, I played with the Optical Pyrometer to better understand its readings.
What was learned:
The Gene Cafe uses a glass drum, with a metal separation plate in it. The Pyrometer ONLY measures the Surface temperature of the Glass, and not what is inside the glass. This means that the glass which was not making contact with the beans reads higher than the portions where it was making almost constant contact with the cooler beans. Until I figured this out, it was confusing why such high-variations were seen in temperatures. The pyrometer does NOT measure the temperature of the beans inside. However, for the portions of the glass in heavy contact with the beans (which read cooler than other parts), it seemed to "scale" with the likely temperature of the beans (minus the loss in the glass temperature due to contact with ambient air). As a test for this, when the pyrometer read 300 degrees on the glass, the roasting chamber was removed and the lid opened. The metal in the interior was scanned, and showed a much higher temperature than the glass. Optical pyrometers only read the first surface they make contact with. Anything inside of the glass IS at a different temperature.

****Post Script*** A later batch was roasted, and the actual temperature of the beans was taken immediately as they were dumped. There was a large temperature difference between the glass surface temperature and the beans. The Beans were 45-50 degrees hotter than the glass surface temperature measurement taken seconds before dumping. This makes sense, as the exterior of the glass drum is in direct contact with the ambient air.

Summary: the glass is a "general indicator of interior temperature, but will always read cooler than the actual interior temperature". The glass temperature on both batches read about 340ish degrees F, when the beans were in the middle of first crack. So, a scaling factor would need to be applied to get a "better" guess of external bean temperature.

Still, this has proven to be a far better indicator of roast temperature than the temperature of the air leaving the chamber (which is what the Gene Cafe reports on its digital display). The air always seems to exit at a much higher temperature than the current roast temperature.

Next posting: The cupping of the Ethiopia Yirgacheffe Misty Valley (first batch).

 

[addertooth]

Cupping of the Ethiopia Yirgacheffe Misty Valley (first batch):

In a word, disappointing.
The roast flavor and body were fine, but the blueberry with floral flavor mixed, was not as pleasant as I hoped.
It was almost like if you tried to aproximate blueberry flavor by mixing chemicals. The flavor was quite recognizable (my wife spotted it without even being told what the brew was). The floral was also strongly present (in an old-lady floral cologne kind of way).

This is after two days of rest. Perhaps it will be less Sharp with two more days of rest.

Next posting: Round two roasting of the Ethiopia Yirgacheffe Misty Valley beans, listening to advice from MusicPhan.

 

[addertooth]

Round two roasting of the Ethiopia Yirgacheffe Misty Valley beans, applying advice from MusicPhan:

Ethiopia Yirgacheffe Misty Valley

Roasting notes for:

Ethiopia Yirgacheffe Misty Valley GR1 beans from Coffee Bean Corral. (Round two, ding ding!)

Roasting machine: Gene Cafe CBR101 2016 production model, stock.

Measure 8 ounces or 227 Grams of green Ethiopian Coffee beans and set aside. (final roasted weight was 7 ounces)

Place the empty roasting cyinder in the roaster, Pre-Heat Gene Cafe Roaster to 392 degrees F or 200 degrees C.

Wait 5 minutes for the temperature to stabilize, and the chamber to uniformly warm up. Optical pyrometer reads the exterior glass drum temperature as 300 F or 149 C. Interior temprature was hotter than the exterior scan shows.

Zero and start timer, place the green beans in the roaster, roast at 392 F for 5 minutes.
(Rougly one minute elapsed between when the pre-heated drum was pulled, loaded with the beans, and roast started.)

After the 5 minutes at 392 F, raise the temperature to 464 degrees F, or 240 degrees C. Continue roasting.

With this batch, the very first crack was heard at 10 minutes 10 seconds. Full rolling crack was at 10 minute 40 seconds.

1 minute and 30 seconds after first crack begins (11 minutes 40 seconds on a kitchen timer), lower the temperature to 437 degrees F, or 225 degrees C. (writing roasting notes really interferes with doing some things quickly).

Continue to roast until desired level of roast, Stopped and force cooled at 13 minutes and 30 seconds.
First crack had fully completed before this time.

For this roast, the Roast duration used was 13 minutes and 30 seconds, this roast did not use a separate dehydration phase.

Notes taken during roasting:

Exterior of the glass drum temperature plus observations at the time.

03:00 255 F
04:00 263 F
05:00 275 F Temperature raised to 464 F
06:00 280 F Beans beginning to get paler, and yellow.
07:00 305 F Beans solidly yellow with some light browning on the silver skin.
08:00 311 F Browning becoming more pronounced.
09:00 320 F
10:00 321 F Browning more noticable, puffing of the beans obvious, seam apparently spreading, chaff starting to flake off the beans.
10:30 no-temp the very first pop is heard.
11:00 340 F smoke is clearly seen, chaff is abundently separating from the beans.
12:00 no-notes, busy listening for when first crack is ending.
13:00 340 F (very quick reading, so it was likely a higher temperature).
13:30 no-temp beans pulled and force cooled.

Nose notes: These freshly roasted beans had a less astringent nose than the first batch.

Color notes: These freshly roasted beans are very slightly less dark, but were roasted shorter than first batch.

Bean appearance after roast: Ever so slight crinkling on the surface. When compared to previous batch the first batch which were smoother and had a more pronounced outward bulge at the seam. The more recent batch has better defined cracks probably due to a higher moisture content before the roasting temperatures were reached.

Bean flavor/texture. These beans still have a fair amount of cellulose integrity. When compared to the first batch, they are less "brittle" on bite through. The berry flavor seems "cleaner/less-artificial" with the batch roasted today, when compared to the former batch.

Moisture loss was roughly 12.5 percent. It would have been more precise if grams were used, as versus ounces.

 

[addertooth]

Round two roasting of the Kenya AA beans, applying advice from MusicPhan:

Kenya AA beans Select Plus Washed

Roasting notes for:

Kenya AA beans Select Plus Washed beans from Coffee Bean Corral. (also Round two, ding ding!)
First round with this bean was quite good.

Roasting machine: Gene Cafe CBR101 2016 production model, stock.

Measure 8 ounces or 227 Grams of green Kenya Coffee beans and set aside. (final roasted weight was 6.9 ounces)

Place the empty roasting cyinder in the roaster, Pre-Heat Gene Cafe Roaster to 392 degrees F or 200 degrees C.

Wait 5 minutes for the temperature to stabilize, and the chamber to uniformly warm up. Optical pyrometer reads the exterior glass drum temperature as 300 F or 149 C. Interior temprature was hotter than the exterior scan shows.

Zero and start timer, place the green beans in the roaster, roast at 392 F for 5 minutes.
(Rougly one minute elapsed between when the pre-heated drum was pulled, loaded with the beans, and roast started.)

After the 5 minutes at 392 F, raise the temperature to 464 degrees F, or 240 degrees C. Continue roasting.

With this batch, the very first crack was heard at 10 minutes 20 seconds. Full rolling crack was at 10 minute 40 seconds.

1 minute and 10 seconds after first crack begins (11 minutes 30 seconds on a kitchen timer), lower the temperature to 437 degrees F, or 225 degrees C. (writing roasting notes really interferes with doing some things quickly).

Continue to roast until desired level of roast, Stopped and force cooled at 13 minutes and 30 seconds.
First crack had fully completed before this time.

For this roast, the Roast duration used was 13 minutes and 30 seconds, this roast did not use a separate dehydration phase.

Notes taken during roasting:

Exterior of the glass drum temperature plus observations at the time.
01:00 239 F
02:00 240 F
03:00 248 F Silver Skin shows signs of drying (whiter color on silver skin)
04:00 250 F Beans are getting a paler color
05:00 264 F Temperature raised to 464 F
06:00 270 F Beans beginning to get paler, and yellow.
07:00 311 F Yellowing complete, browning clearly beginning.
08:00 320 F Browning becoming more pronounced.
09:00 327 F Silver skin is beginning to flake.
10:00 340 F Browning more noticable, puffing of the beans obvious, seam apparently spreading, chaff clearly flaking off the beans.
10:20 no-temp the very first pop is heard. Beginning of FC, but not a full rolling pop yet.
11:00 343 F smoke is clearly seen, chaff is abundently separating from the beans.
11:20 no-temp Middle of first crack, full crack is rolling with frequent pops.
12:00 343 F busy listening for when first crack is ending.
12:30 no-temp First crack is complete.
13:00 no-temp preparing to pull and cool the beans.
13:30 no-temp beans pulled and force cooled.

Nose notes: These freshly roasted beans are very clean smelling, one could easily mistake them for beans ready to brew with.

Color notes: These freshly roasted beans are roughly a full-city level of roast (cracks in beans). Nice uniform color.

Bean appearance after roast: Ever so slight crinkling on the surface. The seam is well opened with the paler silver skin visible in each seam. Cracks are well defined.

Bean flavor/texture. Even though the profile was similar to the Ethiopia beans, they have a bit more brittle-crunch on bite-through. Flavor is very clean, and what has been previously seen with the other Kenya AA roast. There is a reason why I like this bean.

Moisture loss was roughly 13.75 percent. It would have been more precise if grams were used, as versus ounces.

 

[addertooth]

A/B side-by-side comparison of first and second roast of the "Ethiopia Yirgacheffe Misty Valley beans"

Spoiler alert: The second batch was much better.

The second batch had higher acid (brightness). This rounded out the Blueberry flavor and made it cleaner/clearer in flavor. The blueberry flavor was more "balanced (berry, acid, sweetness)" and less muddied with the floral; it was separate and distinct.

As both cups cooled, the chocolate flavor was easier to perceive. At that point, it reminded me of a blueberry chocolate bon-bon. The second batch was still superior cold.

Now, I finally understand the cupping comments from others for this bean.

 

[addertooth]

I went for "round three" roast with the Brazil Bourbon Peaberry Daterra estate beans.
The results continue to improve with each batch.

I had used a similar series of temperature steps with the Brazil Peaberry as both the Ethiopia Misty Valley and the Kenya AA. Only time was varied based upon the "needs of the brean".

The Brazil Peaberry had the shortest overall roast at 12 minutes and 00 seconds. The Kenya AA roast completed at 13 minutes and 30 seconds. Both are rather enjoyable.

I decided to do a side-by-side tasting between the Brazil Peaberry and the Kenya AA coffees. My discernment skills of flavors and smells seems to work better if two different cups are being sampled. It helps me define the unique characteristics with better clarity.

Both coffees had good body, and a solid roast flavor for their given level of roast (full-city, with cracks at the tip of the beans, pulled after first crack but well before second crack).

The Kenya also had well defined earthy and black tea notes to it.
The Brazil had chocolate, buttery and nutty notes; it reminded me of a reese's peanut butter cup.

Even though both of these coffees lacked fruity or floral notes, I found them more enjoyable than the Ethiopia Misty Valley coffee. The Ethiopia coffee was certainly more novel, and as a tasting experience was very worthwhile.

(Post script note: The flavor of the Ethiopia Misty valley continues to improve with time. I find I enjoy the coffee more than I did the day before, perhaps I was a bit pre-mature in my judgement of this coffee. It seems to benefit more from a longer rest than my other coffees.)

But, what this also tells me is that both the Brazil and the Kenya beans would likely need a faster initial rate of rise to preserve those fruity notes which the seller listed in the green bean's descriptions. This is backed up by the fact the beans were not "full yellow, and beginning to brown" until about 7 minutes into the roast.

The Ethiopia bean seems to retain those fruity notes, even when the same temperature steps are used. I am a bit torn on whether I want to increase the initial Rate of Rise, as both the Kenya and Brazil coffee turned out so enjoyable with this roast profile. But then, the goal is to learn.

 

[addertooth]

I ran some more roasts, looking for what "glass drum temperature" was read when first crack occurred. On all of them the answer was about 340 F. This is good news, as it will help me with some of the more quiet beans (decaf and peaberries). I realize this is well below the 385 F which is the typical Bean Temperature, but the glass is making contact with the outside air, which cools the surface to a temperature lower than the interior temperature. For roasts which were taken to second crack, I tried to get some comparitive bean temperatures as they were dumped from the glass drum, but their temperature dropped so fast (when in contact with the outside air) I felt the numbers were unreliable.

I also played with a slightly higher pre-charge warm up temperature as well (402 F as versus 392 F). The beans will need to rest another day before I can give then a fair cupping assessment. The beans did reach "yellow" about 30 seconds faster than previous batches which were started at 392. This is good because the early rate of rise on a Gene Cafe roaster seemed a little lackluster. Getting to a yellow color faster was a sure sign the rate of rise was improved.

I also tried the watchmaker/clockmaker microphone clipped on the chaff collector, to see if it could pick up the Pops of the first crack. It was very useless. It was easier to hear with my ears, than amplified through a speaker. It would likely require significant digital filtering to get rid of all the machine noise the Gene Cafe makes.

There are still three types of beans which are still resting in the bags waiting first roast. The first is a Kona Peaberry, the second is a Panama Geisha bean, the third is a Jamaica Blue Mountain bean. All of these beans should be well worth a bit more understanding and prep before I tackle them.

Some favorite beans have emerged out of what has been tried. Kenya AA is a very reliable bean that seems to be rather accepting of a learning curve. It has never produced a bad batch, even when it was still being dialed in. It is consistently one of my top 3 favorites. The Colombia decaf turns out surprisingly well even when heavilly roasted. It still hangs onto it's chocolate notes. The Kona Extra Fancy seems surprisingly unfussy as well. It has yet to throw a bad batch (alas, I only ordered enough for two batches and the bag is empty). The Ethiopia Yirgacheffe (Misty Valley GR1) was disappointing until the roast was dialed in, but then it really brought forward the flavors which were written in the bean description by the seller. The Costa Rica decaf is also quite solid, but not at the level of the Colombia or Brazil decaf. A second Ethiopia bean was tried early on (Gadeb Worka Sakaro GR1), but at the time I didn't understand the beans from Ethiopia, and it deserved a different roasting profile (more hot and fast). It will probably be re-visited with the knowledge gained over the past couple weeks. Despite a rough start with the Brazil Pearl Bourbon Peaberry, the profile is fairly well dialed in (in spite of the fact that first crack cannot be detected). It is one of my top 4 favorites to drink. This bean was selected to be "training wheels" to get a better understanding of peaberries before I attempted the Kona Peaberry beans. A Mexican Altura bean reamains a clear low performer. That bean was picked up locally (when I was in a rush to find green beans, and I do mean ANY green beans which did not have to be mail ordered). Don't get me wrong, it is much better than Folgers, but it does not shine like the other beans.

 

[addertooth]

The Kona Peaberries were roasted, and I must say they give the Kona Extra Fancy (more than) a run for their money. I would say they are slightly preferable. I used the experience gained from roasting the Brazil Peaberry beans to better understand how to roast peaberry beans. The Brazil Peaberry beans were far less expensive. A good roast was achieved on the first attempt Kona Peaberry, which leaves a half-pound for a later roast.

The Kona Peaberry gave a clearly discernable first crack (which was far less clear with the Brazil Peaberry beans). The clear first crack on the Kona Peaberry gave a real leg-up on knowing how the roast was advancing. Upon roast completion, the Kona Peaberry beans "opened up" more than the Brazil Peaberry beans did.

I also ordered some beans from Sweet Maria's tonight. They are from Rwanda, Kenya and Ethiopia. They were selected for their respective "high scores" (89-91) and a sweet spot in the City to full-City roasting range. They were purchased in 5 pound bags with the intent of being used for half-pound "gift coffee" batches after they are roasted.

Below: picture of Kona Peaberry while cooling.

 

[addertooth]

Today was a big roasting day. Four different beans were roasted (with one of them done twice, as human error messed up the first attempt).

The first three were:
Yemen Mocca Khulani Natural
Bolivia Apolo Washed
Bali Kintamani Natural

The first attempt of the Bali was botched, I was too busy hand logging temperature/smell/appearance and blew right past the time I was supposed to raise the temperature from 392 F to 464 F. It sat at that lower step an extra 5 minutes before I noted my rate of rise was too feeble. That half-pound batch is likely trash, but I will sample it in a couple days against the follow-on batch (which was done according to plan). This will acquaint me to what an extended dehydration period does to the flavor profile. This is not the way I wanted to learn this, but it will be a useful experience. This bean was dropped just as second crack was beginning.

The last bean roasted was done after the roaster had cooled back down to ambient temperature. This bean kept getting pushed back on the "roast list", as a good strategy to roast it was not coming easilly. I wanted to start cold with this roast to constrain as many varables as possible. This bean was:

Panama Finca Lerida Geisha washed

I had read about some people who had used their Gene Cafe to roast Geisha beans, but they were all doing lower altitude beans. This bean is grown at the 1600 to 2000 meters above sea level, which made it more akin to an Ethiopian higher altutude bean than many of the other Geisha beans. A profile which was closer to an Ethiopia Yirgacheffe was used. With that profile, the fruit flavor found in some Yirgacheffe beans had been properly revealed.

This required pre-heating to 402 F, stopping and removing the drum, loading the beans and quickly restarting the roaster to give the roaster a small bit of a head start to have a decent rate of rise. The Gene needs a preheating cycle to achieve a decent rate of rise. Even so, the beans were not at full yellow until almost the 6 minute mark. The temperature setting had already been elevated to 473 F at the 5 minute mark.

First crack was heard at 9 minutes and 30 seconds. The temperature was dropped to 437 F about 45 seconds after first crack was heard. The beans were dropped and force cooled at 12 Minutes and 30 seconds. Moisture loss was at 12.3 % and the Development phase was 24 % of the entire roast time (I could probably get away with shortening it a bit, but that decision will have to wait for the cupping in a couple days).

The Yemen beans were a bit of a challenge too, as they had a LOT of varability in both size and type (there was a surprisingly high percentage of Peaberries in the bag). But, the overall numbers for this roast indicated things were fairly happy. Moisture loss was 13.66 % and Development was 23.1 %. This bean was dropped before second crack.

The Bolivia beans were pretty straight forward, with the only hitch being a longish development percentage of 26%. The Moisture loss on those beans was 13.22 %. This bean was dropped before second crack.

 

[addertooth]

Follow on concerning the Bali Kintamani Natural beans. This bean has a first crack which spread out, and flows into second crack. After resting a week, the beans from both batches have some oil on the surface (a sure sign of entering second crack for both batches). During the Roast there appears to be little sign of the transition from first to second crack, and the progression is rather quick as well. Future roasts with it will require special attention in this area.

The Panama Geisha beans produced a wonderful cup, more floral than a Kona, but with a lime/cherry finish that balanced it well. I finally understand the interest in Geisha beans.

The Yemen beans produced a solid hit of chocolate, and solid body. It is a great drinking coffee.

Today, something new (to me) will be roasted. It comes from Sweet Marias. The bean is a Kenya Kiriyaga Kamwangi Peaberry bean. It was selected due to the high (subjective) rating from Sweet Marias. They scored it at 90.2, and should make a solid cup of coffee. I can't seem to lay off the Peaberry beans, as they have pretty uniformly produced good results for me. It will be roasted pretty hot and agressivly due to the dense nature of these beans. Roast goal will be around a City+. This will be my first bean roasted from Sweet Marias. The outcome will let me know how much I can rely upon their descriptions and rating numbers.

Post Script Edit: It rained hard with blowing rain today. I roasted outside in a car port (to deal with smoke from roasting). The blowing rain added to the challenge, but it was dealt with. Three batches of Different Decaf beans were roasted for the wife (Brazil, Costa Rica, and Colombian), along with the Kenya bean. All of the batches were 227 Gram (1/2 US pound).

I played around with monitoring the Exhaust gas temperature from the Gene Cafe during the roast, (along with external drum temperature at the same time). These were manual measurments done during the roast in real-time. It was pretty time consuming. It is making me consider designing a data logger to simplify the process.

There appears to be a dip in exhaust temperature associated with first crack, and it may be useful in detecting first crack on beans which are more quiet.

 

[addertooth]

Quick note on the Kenya Kiriyaga Kamwangi Peaberry. It is a very solid cup with no detracting characteristics.
It lacks some of the wow factor of other beans which have been roasted, but just makes a clean and wonderful cup.
The Citrus, Tea and Spice notes are there, just not pronounced.

As I am sitting here and typing this, I am drinking my last cup from the half-pound batch; it went fast.

The Bali Kintamani Natural beans have not been dialed in yet. I need to catch them before second crack begins to give a fair evaluation. I did note that brewing method is very important with this bean, with some brewing methods bringing forward a "leathery" flavor which I do not enjoy.

This weekend a Kenya Nyeri Gatomboya Peaberry bean will be roasted. It is rated a 92.1 via Sweet Maria's website. I realize those numbers can be subjective. As a treat for my wife, an Ethiopia Guji Gorebeti SWP Decaf bean will be roasted as well. Her wheelhouse has been Colombia, Brazil, and Costa Rica beans. She is also willing to be "daring" and accept a lighter roasting on this coffee than she is accustomed to. She is normally a "into second crack" kind of gal. We shall see how she likes the baked fruits and cocoa flavors this bean may produce.