Archive for February, 2009

Buried in Samples

Posted in coffee, coffee roasting, home roasting with tags , , , , , , , , , , , , on February 28, 2009 by Chad

As  I was finishing up the last of six sample roasts today, I started wondering how much of my coffee I’ve been roasting lately, as I coudn’t recall the last batch.

Check the roasting log.  Over the last two weeks, I’ve only roasted a pound of my own coffee, and 8 pounds of samples (16 roasts).  And to think, only ten more samples to go!

Here are some of the highlights from the last 16 samples:

Yemen Hufashi

Spice,  leather and toasted malt in the dry fragrance and wet aroma.  The dry grounds also hinted at tobacco and dark fruit, while the break held a touch of pepper and a semi-sweet citrus note.  Full, chewy body matched with medium body with that citrus hint.  Semi-sweet, rich, complex flavor filled with malt, tobacco, leather, a hint of citrus, vanilla, toasted nuts (pecans), and chocolate.  The cup sweetens as it cools.  All capped off by a lingering, smooth, rich, semi-sweet finish with malt, a hint of fruit along with tobacco and leather.  Some would call this a haunting finish.   This coffee, like most Yemens I’ve cupped, came together best about 7 days off the roast.  Roast level: City+

By the numbers:

Dry Fragrance: 3.6
Wet Aroma/Break: 3.6
Body/Mouthfeel: 7.7
Acidity/Brightness: 7.6
Flavor/Depth: 7.8
Finish: 8
Total Score: 38.5 + 50 = 88.3

Ethiopia Natural Yirgacheffe, Worka

The dry grounds burst and fill the room with tropical fruit dominated by bananas and pineapple, accompanied by a hint of spice.  The fruit and spice remain with the wet aroma, and mango pops out on the break.  Snappy, fruit filled brightness, and a creamy, light body.  The tropical fruit continues, along with a touch of coconut cream and caramal as the cup cools.  Piña Coladas, anyone?  Finishes smooth and creamy, with a touch of spice.  Overall an outstanding natural Yirgacheffe, with nice balance between the high-end fruits along with the cream and spice.  Roast level: City+

By the numbers:

Dry Fragrance: 3.8
Wet Aroma/Break: 3.8
Body/Mouthfeel: 7.4
Acidity/Brightness: 7.8
Flavor/Depth: 7.9
Finish: 7.8
Total Score: 38.5 + 50 = 88.5

Espresso Shots

After the cupping, I pulled single origin double shots, as well as a 60/40 Yirg/Yemen blend.

The Yirgacheffe came off a bit bright, though the tropical fruit sweetness popped as the cup cooled.  The Yemen had an excellent balance of the rich dark fruit and tobacco/leather flavors with a sweet fruit note.  The blend came together well, as the deep and rich flavors of the Yemen complemented and balanced the high-end fruit and brightness of the Yirgacheffe.  When I get the bags of these, I’ll be blending them with a chocolate-bomb Brazil, most likely the Moreninha Formosa.

Coffee Fest Chicago 2008 – Quick Thoughts

Posted in coffee on February 24, 2009 by Chad

I recently returned from Coffee Fest Chicago, the (mostly) coffee shop-focused trade show held three or four times every year.  My first experience with Coffee Fest was in Washington DC last February, when I was just beginning the green coffee sourcing adventure.

The retail coffee-shop focus can be seen in the aisle stacked with wholesale vendors offering sweet goods, CHOCOLATE!, small wares, gelato, smoothie systems, espresso and brewed coffee equipment, and retail business consulting.  Moreover, the majority of fee seminars accompanying the Fest focus on the economics of retail coffee shops.

Sprinkled in the maze of non-coffee wares were coffee roasters and green coffee vendors, which was the focus of my trip, along with the chance to meet up with fellow green coffee and home-roasting folks.  We all met up and chatted like old friends, enjoyed cupping together, and ended our day with some wonderful Armenian food at Savat Nova.  We all got the chance to meet some of the wonderful green coffee importers and roasters and connect with folks previously known only by email or telephone.

To finish the motley combination of vendor booths, the Fest hosted the Millrock Free Pour Latte Art Championship, and the Great Lakes Regional Barista Competition.  I’m not (yet) a complete espresso junkie, but both competitions involve extensive training and noteworthy barista skill.  I’ll post some latte art photos shortly.

If you’ve never attended, create your coffee start-up business (it’s a trade show) and head on down to Coffee Fest when it’s in your neck of the woods.  I promise an intense sensory experience!

One last note–two of us embarked on a quest to support an upcoming auction for Coffee Kids by asking vendors whether they would be interested in donating some of their wares.  The response was overwhelming, and I’ll be providing more information in future posts.  The response, and the overall trip reinforced what I’ve learned in the past year–specialty coffee is full of generous, warm, caring, and helpful people.  Every day I jump further down the rabbit hole of specialty coffee connects me more with the type of people make a difference.  Thank you to all.

The trouble with…

Posted in coffee, coffee roasting, home roasting with tags , , , , , , on February 10, 2009 by Chad

The trouble with engineers is that they can’t leave anything alone and feel they have to try to “improve” everything.

I think the same goes for coffee roasters.  But there’s a difference with coffee roasting–engineering has a need to deliver a final product to that meets some deadline.  Coffee roasting allows for continuous improvement with a much lower incremental improvement–unless you need to invest in a new roaster, modify the roaster, or buy better coffee.   Not like that’s ever happened here, no sir.

There’s another thing, too, where there’s similarity between engineering and coffee roasting.  More data provides more ability to know what’s going on, but with that usually  comes the quick realization of where things must be improved, or where you need more data.

So, while being tickled with the ability to monitor and control roasts, I quickly realized that there more more areas for improvement, and more parts of the process that I need to tamp down.

The Good – Kenya Kevote Estate AA


Here’s a roast where everything looks good, a good full city roast with about 4-1/2 minutes between the start of first crack and the end of the roast.  The best part is that the stretch of the profile at the end of the roast did not come with a major drop in environmental temperature, which can cause problems by stopping the roasting reactions at the surface of the beans.

The worrisome – Peru FTO Sample


Though I felt good about this one during roasting, looking at the profile I am left to wonder whether the drop in environmental temperature following the start of first crack affected the roast quality.  About half of the second day’s roasts with the new setup ended up with this environmental temperature drop.  Two things to learn from this–one, I am not keeping good enough logs of the heater level and fan speed during the roast (to help see what’s affecting ET drops), and I need to perform some cupping comparisons with different end environmental temperature drops to determine how the profiles affect cup quality.

One final thought that’s more universal than anything specific from this lesson–every time we learn something new–remember and take joy in that knowledge and the remaining vastness of our ignorance.  There are so many lessons yet to be learned–take pleasure in the journey.

Roaster Upgrade, First Tests

Posted in coffee, home roasting with tags , , , , , , on February 6, 2009 by Chad

The last two posts discussed and illustrated modifications I made to a Hottop roaster to install environmental and bean mass temperature probes, which were to be inputs to a datalogger.

Today I completed five roasts with the new system, taking logs on an antiquated laptop running Windows 98.  As I ran the first two roasts, I noticed something wrong with the environmental temperature readings.  As I dropped the heater power at the onset of first crack, the environmental temperature dropped precipitously, eventually dropping below the bean mass temperature even though I was still pouring heat into the beans.

Plot to show Environmental Probe problem


You can see where the environment temperature drops below bean mass, and though there’s a slight stall, the overall roast continues to progress.  That’s not possible in this part of the roast (endothermic heating) if the environment temperature were actually below the bean temperature.

I opened up the front of the roaster, removed the drum and pulled each thermocouple wire out a bit.  I also bent the environmental wire to stick up closer to the bean drop chute.  That solved the problem–environmental temperature tracked where I expected for the next three roasts.

Roast Profiles

After getting the system and thermocouple details fixed, I ran three more roasts–two with Sumatra Mandheling TP Select, which was one of my first full bag coffee orders.  I targeted each roast for a Full City or Full City+ level, or no more than 5-10 seconds into 2nd crack.  I was also looking to have at least four minutes between the start of 1st crack and the start of 2nd crack.

First Roast – Sumatra Mandheling TP Select

230g, 325°F drop temperature.  Roast pulled ~441°F


Vanguard pop of first crack hit around 11:30, full on 1st crack just before 12:00.  Beans pulled at 16:00.

Second Roast – Sumatra Mandheling TP Select

230g, 250°F drop temperature.  Roast pulled ~440°F.  Premature drop–I got distracted and stopped thinking about the drop temperature.  I also didn’t start logging until about 15 seconds into the roast.


Vanguard pop of first around 12:15, full on 1st crack at 12:40.  Roast terminated at 440°F, 17:35.  I went for a longer stretch on this, dropping the heater more and running the fan at full speed. You can see on the environment temperature about where I cut the heat (just before 12:00 on the graph, 385° bean temperature).  Environmental temperature stayed above bean mass throughout the roast.

Third Roast – Ethiopia Organic Natural Sidamo

After two darker roasts, I wanted to try stretching with a lighter roast.  Target was a city+ roast.  230g, 300°F drop temperature (bean mass TC).


I dropped the heater to 50% (then 30%) at 385° bean mass, and cranked up the fan to full.  Vanguard pops of 1st at 10:45 (390°F), full on at 11:00 (396°F).  Roast pulled ~423°F, 15:15 roast time.

I’m tickled with the manual controls and dual channel plotting–though the final proof will be in the cup.   I think we’ve fallen down another rabbit hole, Alice.

Roaster Upgrade, Part 2

Posted in coffee with tags , , , on February 5, 2009 by Chad

I recently completed an upgrade to a Hottop roaster consisting of two main modifications:

1) Replacement of the Model “P” control panel with the Model “B” control panel

2) Installation of environmental and bean mass temperature probes

Panel Replacement

I  “downgraded” to the “B” control panel for more manual control during the roast.  The Hottop site has a clear write-up of the differences between the control systems here.   I say “downgrade” because the “P” model retails for $200 more than the “B” model.

The procedure to change out the control panels was straightforward; the Hottop site had excellent instructions.  Control Panel change-out procedure–note that the “B” and “P” panels have three wire connectors vice the two shown in the photos.

Temperature Probe Installation

To install the temperature probes, I disassembled the rear portion of the roaster by first removing the fan and then removing the rear panel.  To provide access to drill through the rear portion of the roaster, I  removed the drum motor by removing the four motor mounting screws as shown below.  I left all the wiring and motor capacitor in place and provided temporary support to the motor.

I drilled and reamed out the temperature probe holes with a 1/8″ bit, so that I could insert 0.134″ OD stainless steel tubing to hold the thermocouples.  I located both tubes were located to be inside the drum circumference–the bean mass probe tube was installed into the rear of the drum about one inch, while the environmental probe tube was barely inserted into the chamber.   Both thermocouples are type K beaded wire, to provide the fastest response to temperature changes.

Environmental Probe


The red arrow above points to the environmental temperature installation; you can also see the bean mass probe installation on the left side, just below the two wires for the standard Hottop rear wall thermocouple.

Bean Mass Probe


A closer look at the bean mass probe location, identified by the red arrow.  You can also see the bean chute solenoid at the bottom-center of the photo.

Roasting Chamber


Looking into the roasting chamber, where the red arrow points to the bean mass probe, and the blue arrow to the environmental probe.  Note that the environmental probe is inside the drum circumference–I chose that location to shield the thermocouple from the element’s radiant heat and more closely monitor the temperature of the environment right near the beans.  You can also see by the clean area the location of the bean mass on the back wall during roasting.  For the next installation, I’ll probably locate the bean mass probe a bit closer to the ejection chute to better center it in the bean mass.

Upgrade Complete!


The completed setup, including the Omega HH506RA Dual Input, High Accuracy Datalogger/Thermometer and the variac I use to maintain an even input voltage to the roaster.

In the next post, I’ll show some completed roast curves after I run the machine through its paces.

Roaster Upgrade, Part 1

Posted in coffee with tags , , , on February 5, 2009 by Chad

It’s late, so this one will be short.

I just completed installing two thermocouples into my Hottop, one for environmental temperature and one for bean mass temperature.  During the same installation, I swapped out the Model “P” control panel for the Model “B”, which provides the ability for full manual control of the heater and fan during the roast.  The “P” allows for 8 profile segments that you can set target temperature, segment time and fan level, but its only manual control during the roast is the segment target temperature.

My thermocouples are attached to the Omega HH506RA Dual Input, High Accuracy Datalogger/Thermometer with USB output.

I roasted a sample of Colombia Huila Excelso from Royal Coffee NY.  I’ll need to get more familiar with the full manual control, but first roast wasn’t bad.  Full City+, with about 3-1/2 minutes between the start of first crack and roast completion.  I usually like to stretch that last part out another 30 seconds to a minute or so, but the beans looked good.  I’ll cup them over the weekend.

I’ll post moreshortly , including pictures and (hopefully) a datalogging graph.