Coffee | One Roast at a Time

🔎 Key Metrics — Roast 36 vs Roast 37 (Brazil Santos, 250g)

Roaster: Gene Café CBR-301
Ventilation: Oven hood extractor (not directly ducted)
Room temp: ~20–21°C

🟢 Roast 36 — Cold Start (20°C Drum)

  • 200°C: 04:46
  • 230°C: 07:49
  • Cool Initiated: 13:11
  • Weight Out: 210.4g
  • Weight Loss: 15.84%
  • Behaviour: Textbook progression, stable development

🔵 Roast 37 — Warm Start (71°C Drum)

  • 200°C: 03:06
  • 230°C: 05:50
  • Cool Initiated: 12:41
  • Weight Out: 208.1g
  • Weight Loss: 16.76%
  • Behaviour: Accelerated phases, deeper colour, higher WL despite shorter TRT

📌 Immediate Lessons

  • Cooling to 60°C display does not equal thermal neutrality.
  • Drum start temperature materially affects phase timing and WL.
  • Warm/hot starts require larger TRT reductions than previously assumed.
  • Visible smoke aligned with late Maillard/development phase, not early roast.

This session wasn’t about flavour.

It was about heat, airflow, and what “cooling complete” actually means on the Gene Café CBR-301.

Two back-to-back Santos roasts.
Same settings.
Very different thermal starting conditions.

The difference was instructive.

Setup & Fixed Parameters

  • Roaster: Gene Café CBR-301
  • Ventilation: Oven hood extractor (not directly ducted)
  • Bean: Brazil Santos (natural)
  • Batch size: 250g
  • Set temperature: 250°C
  • Fan: 2 → 3 at ~4:00
  • No preheat

Room temperature: ~20–21°C.

Roast 36 — Cold Start (Control)

Drum start temp: 20°C
200°C: 04:46
230°C: 07:49
Cool initiated: 13:11
Weight out: 210.4g
Weight loss: 15.84%

This roast was textbook.

Phase progression was steady and predictable.
Development felt controlled.
Colour aligned with previous Santos reference roasts.

The 13:11 total roast time (TRT) landed precisely in the validated cold-start guardrail (13:10–13:20), producing 15.8% WL — right in the target 15.8–16.2% comfort window.

Conclusion:
Cold-start TRT control remains reliable at ~20–21°C ambient.

Roast 37 — Warm Start (Drum 71°C)

Drum start temp: 71°C
200°C: 03:06
230°C: 05:50
Cool initiated: 12:41
Weight out: 208.1g
Weight loss: 16.76%

This roast began shortly after Roast 36 completed cooling.
Although the machine had reached its standard 60°C cooling threshold, the glass drum temperature was still 71°C.

That difference mattered.

Phase Acceleration

Compared to Roast 36:

  • 200°C was reached 1:40 earlier
  • 230°C was reached 1:59 earlier

That is significant acceleration.

Despite shortening TRT by ~30 seconds (13:11 → 12:41), the roast still finished at 16.76% WL — nearly 1% higher than the control roast.

Visually, beans were already deeper brown at 12:37–12:40.
The roast was pulled 4 seconds earlier than planned based on colour progression.

Key Lesson 1 — “Cooling Complete” ≠ Thermally Neutral

The machine reaching 60°C does not mean the drum has returned to a cold-start state.

A 71°C drum start:

  • Accelerates early phase development
  • Compresses Maillard progression
  • Increases total energy delivered
  • Raises weight loss even with shortened TRT

The internal thermal mass matters more than the display cooling threshold suggests.

Key Lesson 2 — Warm Start TRT Correction Must Scale With Start Temp

Previous calibration suggested:

  • Cold start (~20°C): 13:10–13:20 → ~15.8–16.2% WL
  • Warm start (~60°C): ~12:40 → similar WL

However, at 71°C start, a 30-second TRT reduction was insufficient.

Future guardrail adjustment:

  • Cold start (~20°C): ~13:15
  • Warm start (~60°C): ~12:40
  • Hot start (~70°C+): likely ~12:20–12:30

This will need repeat confirmation, but the directional pattern is clear.

Smoke Behaviour — Observation Only

Smoke was documented during both roasts under a non-optimal exhaust configuration:

  • Exhaust pipe currently damaged
  • Chaff box gasket temporarily sealed using heat-tolerant mastic
  • Oven hood extractor active but not directly ducted

Observed Pattern

Smoke became visible:

  • 10+ minutes into roast
  • Coinciding with:
    • Toasty / hot bread aroma
    • Peanut-brown colour shift
    • Late Maillard into development phase

Smoke did not appear early in roast.

This suggests phase-aligned volatile release rather than simple airflow leakage.

The extractor significantly reduced visible smoke wafting into the room during active roasting.

A mild roast aroma was still detectable in the room the following day. At this stage, it is not possible to isolate whether this is:

  • Residual accumulation from previous roasts
  • Minor airflow leakage
  • Normal post-roast off-gassing behaviour

Replacement chaff box and exhaust components are on the way, allowing controlled comparison under optimal airflow conditions in a future session.

Key Lesson 3 — Smoke Correlates With Development Intensity

The strongest visible smoke coincided with:

  • Colour deepening
  • Increased roast aroma intensity
  • Development ramp

This aligns with normal roasting physics:

Late Maillard and early caramelisation release more volatile compounds.

This reinforces an important distinction:

Active smoke during development is not inherently a defect.
It is phase-related.

Structural Outcomes From This Session

  1. TRT remains the primary WL control lever.
  2. Drum start temperature must now be explicitly recorded and considered.
  3. Cooling-to-60°C is not a sufficient reset for batch neutrality.
  4. Smoke observation should always be correlated with roast phase, not assumed to be airflow failure.

Closing Reflection

Roast 36 confirmed baseline repeatability.

Roast 37 exposed an assumption.

The Gene Café is consistent — but only when starting conditions are consistent.

Thermal mass carries forward.

Energy compounds.

And a 10°C difference at drum start meaningfully alters outcome.

This session was less about coffee.

More about system awareness.