Artisan Documentation

Visual scope for coffee roasters

Graph Area

This is a complete plot graph area. Starting at the top and moving down:

Title: Costa Rica / East Timor P3 2
The convention used here is Beans, Program 3, roast 2

Graph lines legend: Red Environment Temperature, Purple Bean Temperature, Orange Delta Environment Temperature and Blue Delta Bean Temperature

Summary of the roast:
7:43 spent in drying phase for 38% of the roast. The temperature climbed at 25.5 degrees per minute on average.
8:01 spend to first crack or 39% of the roast. Temperature climb averaged 11.8 degrees per minute.
From start of first crack to end of roast was 4:29 or 22% of the roast. Temperature climb was 8.8 degrees per minute.

Four graph lines: Between ET/BT and DeltaET/DeltaBT are horizontal bars. These are where Power, Damper and Fan events are plotted.

Power was 10 at 4:00 and lowered to 7 at 6:10. Fan started at 0 and went to 1 at 6:00 and 2 at 8:45.

Boxed lines from end of the ET/BT lines are the HUD projections.

Left Y axis is the temperature while the X axis is the time.

Roast Phases
At the top of the graph is a section where the roast is broken into three phases and a description is attached. The phases include: drying, mid-phase and finish phase. Here are the program rules on how the description of the phase is calculated. This is based on work by Jim Schulman.

The roast is divided in three phases (lengths of time) as follows.

  • Dry phase: from [CHARGE] to [DRY END]
  • Mid Phase: from [DRY END} to [FIRST CRACK]
  • Finish phase: from [FIRST CRACK] to [DROP]

The [Buttons] symbolize the time you press each button.

The times listed are the presets. You can change those in the statistics dialog (Conf/Statistics).

Furthermore, since 0.3.5 you might see evaluations like “Acidic/OK” (eg. combinations of OK with one of the other two in that category) indicating that in principle the evaluation result is OK but it’s at the border to “Acidic”. This is computed by splitting each phase in three parts. Only the middle part gets a full OK, while the other two parts are reported as such “border cases”.

  • Dry Phase
    • if dry phase time less than 3 mins or less than 26% of the total time then Short Drying Phase “Grassy”
    • if dry phase time greater then 6 mins or more than 40% of the total time then Long Drying Phase “Leathery”
    • else “OK”
  • Mid Phase
    • if mid phase time less than 5 minutes then Short To 1C Phase “Toasty”
    • if mid phase time greater then 10 minutes then Long To 1C Phase “Bready”
    • else “OK”
  • Finish Phase
    • if finish phase is less than 3 mins then Short Finish Phase “Acidic”
    • if finish phase is over 6 minutes then Long Finish Phase “Flat”
    • else “OK”

Tags: ,

5 Responses to “Graph Area”

  • Rick Shaw says:

    First and foremost I would like to thank you for a great application. It is incredibly useful.

    Now for the feature request: Is it possible to start the phase timers after the turning point? For example, if I charge at 325°F BT, as the beans reach equilibrium, the BT passes through the dry phase region but this technically isn’t part of the dry phase until after the turning point and BT begins to rise.

    A related feature would be a “TP” button between “charge” and “dry end” which would 1) mark the turning point as a roast landmark and 2) start the phase timers.

    By starting the phase timers at the turning point instead of the charge point, you’ll more accurately reflect how close to three equal phases you’re getting in your roast.

    You can see in the image linked below that the roast turned at about 1:15 and started the dry phase around 2:50 and ended at 6:12, but the “phase indicator” shows a 6:12 dry phase instead of a 3:22 dry phase.

    http://www.fourpointsroastery.com/wp-content/uploads/2011/11/roast-42.png

  • Barrie says:

    I notice that on my graph of a roast, at the “Beans In” point, Delta Bt starts to decrease nearly a minute before Bt decrease. As these are very sharp “inflection” points, the difference is very clear. What could account for this?

  • MAKOMO says:

    Barrie on all channels;) The earlier drop on the graph on DeltaBT is caused by the smoothing. Every mathematical smoothing causes some error and looses information. If you turn of smoothing you see just the pure data. However, also the raw DeltaBT is not “recorded”, but just computed. It is an approximation of the first derivation of the BT. That approximation also causes errors. Remember, we do not have the mathematical precise BT function, but just a series of samples thereof.

  • Barrie Fairley says:

    One more question re the “graph area” if I may? At the bottom there is a series of calculations, the first four are clear, but the fourth (ETBTa) is less so (to me :-)). Is this the area between the ET and BT curves and, if so, how is the range derived? What interpretation may be attached to ETBTa?

  • Barrie Fairley says:

    Sorry, a typo. The fifth is ETBTa.