iRoast Sample Roasts --
Tune Up & Learning Roasts
Spring, 2005

Roast1-CoffeeFinish.jpg
Roast1-CoffeeFinish
 Roast1-KProbe-AfterRoast.jpg
Roast1-KProbe-AfterRoast

Information on setup


You can see my K-Type thermocouple setup, above. It does not read bean temperature, as it's set in the bean bed. Based on roast color, it may be 5dF-20dF high, depending on degree of roast.

The following roasts are not yet fully cupped, I'll be adding information to this page later. These roasts are experiments to more fully understand iRoast functions, so aren't recommendations. But maybe some of the info is useful.


After Jim Schulman took a look at the temperature charts for Roasts 1-7 below, he offered some advice, which I have begun putting into practice in Roast 8, my first decent roast. Please take a look at the notes of Jim's advice and a visual roast-profile interpretation on this page.


ROASTS 1 - 7
below are informational -- showing some roast experiments with profiles, voltage, bean weights and such. None were successful roasts.Goals and observations are noted on each separate roast page.

SWEET MARIAS MONKEY BLEND

START with the best roast: Roast 8
See Roast 8 for an 11min. 40sec. Full City+ roast, 150grams, variable voltage setting.

Roast 1 is OCE "House" espresso blend, roasts 2-8 are all SM Monkey Blend.
Of the first 7 roasts, Roasts 2 and Roast 4 produced acceptable espresso. Roasts 1, 3 and 5 did not produce good results, but were helpful for experimentation.

Roasts 6, 7 are more 'tuned in' runs to find better profiles and temperature curves for Monkey Blend in the iRoast -- but these roasts run too long and have other problems (see Jim Schulman's advice, above).

See Roast 1 for an 11min. Viennese roast, 150grams, higher voltage setting.


See Roast 2 for a 12min. Full City+(+) roast, 120grams, lower voltage setting.


See Roast 3 for a 10min. Medium French roast, 120 grams, higher voltage setting.


See Roast 4 for a 14min. 10sec. Viennese roast, 120 grams, lower voltage setting.


See Roast 5 for a [damaged] 14min. 10sec. City+-to-Vienna roast, 150 grams, lower voltage setting.


See Roast 6 for a 14min. 30sec. City+ roast, 120 grams, lower voltage setting.


See Roast 7 for a 15min. Viennese roast, 120 grams, lower voltage setting.


ROAST 1, ROAST 2 & ROAST 3 use the same iRoast Profile:

Stage 1: 325F for 5 minutes
Stage 2: 390F for 7 minutes
Stage 3: 460F for 3 minutes


MY LATEST ROAST PROFILE:

150gram weight
Stage 1: 320F for 4 minutes, at 100V (measured while running)
Stage 2: 420F-430F for 2.25 minutes (at 110V-115V)
Stage 3: 390F 8.75 minutes (around 110V, or lowest reasonable)


iROAST FINDINGS
  1. Deficiency 1: Like most air roasters, the iRoast ramp warms and dries the beans too quickly. Lately, been using the lowest temp 320F at 100V (volts under load, controlled by 20 amp variac) to slow the rise to 300F to about 3.5 minutes (best I can do).

  2. Deficiency 2: End-stage of roast problem. From second crack to the end, a slower heat ramp is often called for, but the iRoast has a problem here - a setpoint of 385F (high speed fan) can cause a stall, or takes too long to get from 400F to say 455F. But 390F uses a medium-speed fan, and roasts too fast, from 400F-455F. Voltage control can help -- bringing down the voltage using the 390F setpoint -- but fan speed is affected beyond a certain point, resulting in poor bean circulation.

  3. Running-load voltage has a major effect on roast temperature curves.

  4. The weight/density of individual beans in different types of blends/varietals will have a major influence. For instance, Monkey Blend arrived with larger, heavier, denser (higher moisture content) beans than OCEx beans. What happened: the lowest-speed fan setting (e.g. a +430 high temp end-roast setting) does not circulate the beans properly -- a 120gram Monkey roast was ruined at the end stage (Roast 5).

  5. The above point is related to roast weight, which plays a large factor.

  6. Chaff -- Monkey blend contains over 200% the chaff of the OCEx blend, there is a 'chaff effect:' the temperature ramps up toward the end of the roast.

  7. A thermocouple is a must, the iRoast readout is innaccurate and inconsistent--also it will often not register important temp changes at critical times.

 Richard G
comments welcome. email
gilbert(at)iyume(dot)com

Go to Roast 1



iRoast Roasting links

tech and deep tech
Sweet Marias - best iRoast page to start
Coffee Geek iRoast Forum - engineering

chaff issue
Coffee Geek iRoast Forum - chaff issue

some profile discussions
Coffee Geek iRoast Forum
Roast Magazine (May/June) Roasting Fundamentals

bean density
Coffee Geek iRoast Forum